J S S
ISSN 1615-9306 · JSSCCJ 38 (12) 2007–2192 (2015) · Vol. 38 · No. 12 · June 2015 · D 10609
JOURNAL OF
SEPARATION SCIENCE
Methods Chromatography · Electroseparation Applications Biomedicine · Foods · Environment
12 15
www.jss-journal.com
2117
J. Sep. Sci. 2015, 38, 2117–2125
Jiawei Meng1,2 Wenpeng Zhang Tao Bao Zilin Chen1,2 1 Key
Laboratory of Combinatorial Biosynthesis and Drug Discovery (Wuhan University), Ministry of Education, and Wuhan University School of Pharmaceutical Sciences, Wuhan, China 2 State Key Laboratory of Transducer Technology, Chinese Academy of Sciences, Beijing, China Received March 6, 2015 Revised March 24, 2015 Accepted March 24, 2015
Research Article
Novel molecularly imprinted magnetic nanoparticles for the selective extraction of protoberberine alkaloids in herbs and rat plasma In this work, a novel magnetic nanomaterial functionalized with a molecularly imprinted polymer was prepared for the extraction of protoberberine alkaloids. Molecularly imprinted polymers were made on the surface of Fe3 O4 nanoparticles by using berberine as template, acetonitrile/water as porogen, acrylamide as functional monomer and ethylene glycol dimethacrylate as cross-linker. The optimized molar ratio of template/functional monomer was 1:7. The polymeric magnetic nanoparticles were characterized by transmission electron microscopy and Fourier transform infrared spectroscopy. The stability and adsorption capacity of the molecularly imprinted polymers were investigated. The molecularly imprinted polymers were used as a selective sorbent for the magnetic molecularly imprinted solid-phase extraction and determination of jatrorrhizine, palmatine, and berberine. Extraction parameters were studied including loading pH, sample volume, stirring speed, and extraction time. Finally, a magnetic molecularly imprinted solid-phase extraction coupled to high-performance liquid chromatography method was developed. Under the optimized conditions, the method showed good linear range of 0.1–150 ng/mL for berberine and 0.1– 100 ng/mL for jatrorrhizine and palmatine. The limit of detection was 0.01 ng/mL for berberine and 0.02 ng/mL for jatrorrhizine and palmatine. The proposed method has been applied to determine protoberberine alkaloids in Cortex phellodendri and rat plasma samples. The recoveries ranged from 87.33–102.43%, with relative standard deviation less than 4.54% in Cortex phellodendri and from 102.22–111.15% with relative standard deviation less than 4.59% in plasma. Keywords: Herbs / Magnetic molecularly imprinted polymers / Protoberberine alkaloids / Rat plasma / Solid-phase extraction DOI 10.1002/jssc.201500264
Additional supporting information may be found in the online version of this article at the publisher’s web-site
1 Introduction Protoberberine alkaloids are the main active components of many herbs and they have an isoquinoline structure (Supporting Information Fig. S1). The activity of the protoberberine Correspondence: Dr. Zilin Chen Luojia Chair Professor. Vice Dean and Institute Director School of Pharmaceutical Sciences, Wuhan University, Wuhan 430071, China Email: [email protected] Fax: 86-27-68759850
Abbreviations: AA, acrylamide; M-MIP, magnetic molecularly imprinted polymer; M-MI-SPE, magnetic molecularly imprinted solid phase extraction; M-NIP, magnetic non-imprinted polymer; MNP, magnetic nanoparticle; TEOS, tetraethylorthosilicate; MAPS, 3methacryloxypropyltrimethoxysilane; MAA, methacrylic acid; 4-VP, 4-vinylpyridine; AIBN, 2,2 -azobis (2-methylpropionitrile)
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
alkaloids depends on the oxygen substituents in rings A, C, D and the quaternary nitrogen atom [1, 2]. Protoberberine alkaloids possess activities such as anti-inflammatory [3], antibacterial [4, 5], antimalarial [6], antitumor [7, 8], and have been used for the treatment of gastrointestinal diseases and bacterial diarrhea. They can also affect the pharmacokinetics of other drugs [9] and the content of serum cholesterol [10,11]. Berberine, palmatine and jatrorrhizine are most commonly found in many herbs, such as Coptidischinensis [12], Cortex phellodendri, and Berberis vulgaris [13]. Protoberberine alkaloids also serve as the main active component of dozens of compound medicinal preparations such as Fufang Huangbai Ye and Huanglian Shangqing Wan. Generally, the matrix of herbs is very complex, leading to difficulty in the QC of the herbs. Besides, determination of the active alkaloids in biofluids such as plasma and urine is necessary for pharmacokinetic study. The quantitative analysis would be more difficult due to the complexity of matrix
www.jss-journal.com
2118
J. Sep. Sci. 2015, 38, 2117–2125
J. Meng et al.
from both herbs and biofluids. Therefore, the extraction of target analytes as well as clean-up of interferent is necessary for the determination at trace level in real samples. In recent years, magnetic nanoparticle-based techniques have attracted more and more attention in separation science. Magnetic nanoparticles (MNPs) which possess a large surface area can move, gather and disperse conveniently by the control of the magnetic field [14–16]. Besides, extraction materials could be modified onto the surface of nanoparticles to facilitate their application in extraction, such as polymer [15, 16], graphene [17], metal–organic frameworks [18], polyaniline [19], silica shell [20], and octadecylsilyl [14]. The variety of sorbents expands the application of MNPs in different kinds of samples. However, the selectivity of these sorbents is still limited, so that satisfactory clean-up efficiency would not be obtained in very complex matrix such as herbs and biofluids. Molecularly imprinted polymers (MIPs) are synthesized materials with functional recognition sites and complementary cavities in accordance with shape and functional groups of the template molecules. MIPs can be obtained by copolymerization of functional and cross-linking monomers in a mixture solution containing molecular template, initiator and porogen. After polymerization, the template is removed and the MIPs with the tailor-made binding sites and cavities are obtained [15, 16, 21–24]. MIPs possess similar selectivity with natural enzyme but have advantages of higher stability and less cost. MIPs can be used in a wide range of samples such as food [25], traditional Chinese medicines [26], biological and environmental samples [27,28]. In a previous work, we have successfully prepared a berberine-MIP monolith inside a micropipette tip for -SPE of berberine. However, to the best of our knowledge, MIP-based MNPs have not been reported for separation or extraction of protoberberine alkaloids. In this work, we aimed to fabricate berberine-MIP based MNPs and apply to extraction of protoberberine alkaloids in complex samples. The characterization, synthetic conditions, adsorption capacities and stability of the magnetic molecularly imprinted polymers (M-MIPs) were studied to obtain good selectivity. Finally, an M-MIP-based extraction-HPLC method was developed for the determination of protoberberine alkaloids in Cortex phellodendri and rat plasma samples.
2 Materials and Methods 2.1 Materials and Chemicals Berberine, jatrorrhizine, palmatine, tetraethylorthosilicate (TEOS), and 3-methacryloxypropyltrimethoxysilane (MAPS) were obtained from Aladdin-reagent (Shanghai, China). Ethyleneglyol dimethacrylate was obtained from Alfa-Aesar (Lancashire, UK). Acrylamide (AA), methacrylic acid (MAA), 4-vinylpyridine (4-VP), and 2,2 -azobis(2-methylpropionitrile) (AIBN) were obtained from Shanghai Reagent Factory (Shanghai, China). The above reagents were all analytical reagent grade. Methanol and acetonitrile were HPLC grade and purchased from Sigma companies (USA). C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
All experimental protocols involving animals were reviewed and approved by the Institutional Animal Experimentation Committee of Wuhan University. Sprague–Dawley (SD) rats weighing 150–200 g, were obtained by the Experimental Animal Center of Wuhan University (Wuhan, China). All the SD rats were maintained under standard conditions with normal access to food and water. 2.2 Instrumentation A Shimadzu HPLC system (Tokyo, Japan) consisted of two 20AD pumps, a 20A3 degasser, 20A UV detector, and thermostat controlled column compartment were used for chromatographic analysis. Data were acquired and processed using Shimadzu LC Solution software. The analytical column was a 250 mm × 4.6 mm, 5m C-18 column from GL Science (Tokyo, Japan). The mobile phase consisted of acetonitrile and KH2 PO4 (10 mM) buffer (32:48, v/v) with a flow rate of 0.8 mL/min. The detection wavelength was set at 345 nm and column temperature was 25 ⬚C. 2.3 Preparation of M-MIPs The synthesis protocol of M-MIPs is shown in Supporting Information Fig. S2. The Fe3 O4 nanoparticles were synthesized according to a reported procedure by Wang et al. [29]. Fe3 O4 @SiO2 -MAPS particles were fabricated according to the sol–gel method. MNPs were activated by HCl (0.1 M), collected and then rinsed with H2 O and ethanol for three times. The dried MNPs were dispersed in isopropanol (80 mL)/H2 O (6.4 mL) solution, followed by adding NH3 •H2 O (8 mL), TEOS (1.6 mL) and MAPS (1.6 mL) sequentially under stirring and reaction for 12 h. The synthesized nanoparticles were collected by a magnet and washed with H2 O and ethanol for several times. The polymerization mixture contained the template (berberine, 0.022 mmol), functional monomer (AA, 0.157 mmol), acetonitrile/H2 O (8 mL, 9:1, v/v), Fe3 O4 @SiO2 MAPS nanoparticles (80 mg), and cross-linker (ethyleneglyol dimethacrylate, 0.679 mmol). The solution was stirred for 2 h at room temperature, followed by adding the initiator AIBN (40 mg). After deoxygenation by ultrasonication for 15 min, purging with nitrogen for 15 min, the system was incubated at 60 ⬚C for 20 h. The M-MIPs were eluted with methanol/acetic acid (9:1, v/v) and methanol several times in an ultrasonic bath. The resulting particles were dried at 60 ⬚C overnight. As a comparison, magnetic non-imprinted polymers (M-NIPs) were prepared using the same method but without addition of berberine during the polymerization process. 2.4 Procedure of M-MI-SPE Jatrorrhizine, palmatine, and berberine were dissolved in 20 mL of phosphate buffer (pH 7) at concentration of www.jss-journal.com
Sample Preparation
J. Sep. Sci. 2015, 38, 2117–2125
2119
Figure 1. FTIR spectra of the Fe3 O4 (A), Fe3 O4 @SiO2 -MAPS (B), and M-MIPs (C).
mixed solution was extracted and analyzed as presented in Section 2.4. The equilibrium adsorption amount of berberine (Q, g) was calculated using the following equation: Q = CV
(1)
where C (g/mL) is the concentrations of berberine after extraction and V (mL) is the volume of eluent (0.2 mL).
2.6 Application of M-MIPs in Cortex phellodendri and rat plasma samples
Figure 2. TEM images of M-MIPs.
100 ng/mL. The prepared berberine M-MIPs (5 mg) were mixed with 20 mL of the standard solution at a stirring speed of 400 rpm. After loading for 2 h, the particles were separated from the solution and eluted with 0.2 mL acetonitrile by sonication for 5 min. Then the eluent (20 L) was analyzed by HPLC.
2.5 Adsorption experiment To evaluate the adsorption and desorption capacity of M-MIPs toward template molecule, the adsorption experiments were operated. M-MIPs and M-NIPs (5 mg) were suspended in 20 mL of phosphate buffer (pH 7) solution with berberine concentrations ranging from 0.01–100 g/mL. The C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Cortex phellodendri was sliced and ground into powder. Then 0.2 g of the powder was extracted with 5 mL of purified water for 40 min by sonication. The crude extraction solution was centrifuged at 10 000 rpm for 8 min and the supernatant was diluted to 10 times. Then 30 L of the resulting solution was added into 20 mL of phosphate buffer (pH 7), extracted and analyzed as represented in 2.4. The powder of Cortex phellodendri was sieved through a 75 m sieve. 1.0 g of the obtained particles was weighed accurately and extracted by 8 mL of water for 40 min in an ultrasonic bath. After the rats received a single oral dose of 4 mL of the above extract for 60 min, the medicated blood sample was collected in a heparinized Eppendorf tube and centrifuged at 10 000 rpm for 10 min. The harvested plasma was added with equal volume of 5% trichloroacetic acid solution to remove protein. After centrifugation for 10 min at 10 000 rpm, the supernatant was added slowly with NaOH solution (1 mol/L) to adjust the pH value to 7. Then 0.5 mL of the solution was added into 20 mL of phosphate buffer (pH 7), extracted and analyzed as presented in Section 2.4. www.jss-journal.com
2120
J. Meng et al.
J. Sep. Sci. 2015, 38, 2117–2125
Figure 3. The effects of different functional monomers (A), different ratios of berberine/ functional monomer (B), and different porogens (C).
3 Results and discussion 3.1 Characterization of magnetic nanoparticles The prepared M-MIPs were characterized by FTIR spectroscopy. The FTIR spectra of Fe3 O4 , Fe3 O4 @SiO2 -MAPS and M-MIPs are shown in Fig. 1. The peak at 630 cm−1 attributed to Fe–O stretching bond for Fe3 O4 nanoparticles (Fig. 1A). As shown in Fig. 1B, the peak at 1162 cm−1 corresponded to Si–O–Si stretching. The absorbance bands at 1729, 2946 cm−1 were caused by stretching of C=O and C–H, respectively, indicating that MAPS was modified onto Fe3 O4 successfully. Compared with Fe3 O4 @SiO2 -MAPS, the peaks at 1729 and 2946 cm−1 became stronger evidently (Fig. 1C), indicating MIP layer had been formed on the surface of Fe3 O4 @SiO2 -MAPS. The morphological structure of M-MIPs was characterized by TEM. As shown in Fig. 2, a smooth and uniform layer was coated on the Fe3 O4 @SiO2 -MAPS face, indicating that MIPs had been introduced. It can be seen that the diameter of pure MNPs was about 80 nm and the thickness of MIP layer was approximately 30 nm. There were connected MIP particles in the TEM view, while these MIPs will not affect the extraction capacity.
3.2 Optimization of synthesis conditions Firstly, Fe3 O4 particles were prepared by a co-precipitation method at 80 ⬚C. The pH value of solution was adjusted with NaOH to 9–10 to avoid the synthesis of iron hydroxide. When pH value was too low or high, most of the final product was probably iron hydroxide. After the reaction, the process of C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
stirring in oil bath for 30 min and ultrasonication for 20 min was repeated three times to promote particles dispersion. Secondly, the surface of MNPs was modified. As the common method, the MNPs were coated with silica firstly and then modified with MAPS. However in this study, double bonds were introduced onto the Fe3 O4 surface with MAPS by one step. MAPS participated in the silica surface when Fe3 O4 was coated with silica. The proposed method possessed advantages such as simple operation and freedom of anhydrous toluene. 3.2.1 Effect of functional monomers Functional monomers played an important role for the selectivity of M-MIPs because monomers provided functional groups to form accurate and selective binding sites with the templates. Different monomers including AA, MAA, and 4-VP were tested respectively to evaluate the specific recognition of imprinted polymer. As can be seen in Fig. 3A, the best selectivity and extraction efficiency were obtained when AA was used as functional monomer, therefore, AA was chosen for polymerization. The possible reason was that four hydrogen bonds can be formed between hydrogen atoms on amino group and four oxygen atoms of berberine, the electrostatic interaction can be formed between carbonyl group and quaternary ammonium of berberine. It was the hydrogen bonds and electrostatic force that lead to the self-assembled polymerization of berberine and AA. The molar ratio of template/functional monomer can influence the density and imprinting efficiency of M-MIPs. Over-high ratios diminish the selectivity due to higher non-specific binding sites in MIPs, but lower ones induce fewer template-monomer complexes resulting in less binding www.jss-journal.com
Sample Preparation
J. Sep. Sci. 2015, 38, 2117–2125
2121
Figure 4. The effects of sample pH (A), sample volume (B), stirring speed (C), and extraction time (D).
capacity. The ratios were studied from 1:3 to 1:8 to prepare M-MIPs with berberine. As shown in Fig. 3B, largest difference between MIPs and NIPs was observed when the molar ratio was 1:7, and it was selected for further studies. 3.2.2 Effect of porogens Porogens were not only used to generate cavities, but also served as solvent for polymerization. An ideal porogen should allow adequate dissolution of molecule template, functional monomer and cross-linker, and formation of appropriate tailor-made cavities. Berberine dissolved little in solvents with little polar organic, while higher polar solvents probably disturbed the main interaction of berberine-AA, such as hydrogen bond. In consideration of these factors, DMSO, DMF/water (9:1, v/v) and acetonitrile/water (9:1, v/v) were investigated. As can be seen in Fig. 3C, when acetonitrile/water (9:1, v/v) was used as the porogen, significant difference was observed between M-MIPs and M-NIPs.
Figure 5. Extraction yield curves of M-MIPs and M-NIPs for berberine.
solution of three protoberberine alkaloids (100 ng/mL for each) with 3 mg of polymers. 3.3 Optimization of extraction conditions 3.3.1 Effect of loading pH It is important to optimize extraction conditions to achieve high adsorption efficiency of M-MIPs. The conditions possible to affect extraction efficiency were studied in the mixed C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
pH of sample solution affected ionization degree of the target molecules, surface charge of M-MIP particles, and further www.jss-journal.com
2122
J. Sep. Sci. 2015, 38, 2117–2125
J. Meng et al.
Table 1. Analytical parameters of the M-MI-SPE-HPLC methods for analysis of protoberberine alkaloids
RSD% (n = 3) Analytes
Range (ng/mL)
Regression equation
R
LOD (ng/mL)
LOQ (ng/mL)
Intra-day
Inter-day
Jatrorrhizine Palmatine Berberine
0.1-100 0.1-100 0.1-150
y = 4066.90x + 2781.88 y = 4511.22x + 3011.72 y = 5808.02x + 23851.32
0.9996 0.9997 0.9991
0.02 0.02 0.01
0.08 0.07 0.03
0.94 2.00 1.22
4.43 1.54 1.59
Figure 6. Chromatograms of Cortex phellodendri: (A) with MI-SPE. (B) without extraction; 1 jatrorrhizine, 2 palmatine, 3 berberine. Table 2. Results and recoveries in Cortex phellodendri and rat plasma samples
Cortex phellodendri
Plasma
Samples
Added (mg/g)
Recovery (%)
RSD (%, n = 3)
Added (ng/mL)
Recovery (%)
RSD (%, n = 3)
Jatrorrhizine Palmatine Berberine
3.33 3.33 3.33
101.20 101.35 94.07
2.34 4.55 4.22
20 20 20
111.18 104.22 99.95
4.59 1.76 3.06
affected the formation of hydrogen bonds between adsorbed molecules and M-MIPs. Therefore, sample solution with pH value varied from 3 to 9 was investigated. As shown in Fig. 4A, the adsorption capacities of M-MIPs to the three protoberberine alkaloids increased gradually from pH 3 to 7, but the ones of jatrorrhizine and palmatine decreased obviously at pH 7. The optimal pH for berberine was pH 8, while the best pH for jatrorrhizine and palmatine were pH 7. Considering the good extraction efficiency of three components, pH 7 was selected. As described in Section 3.2.1, the reasons were probably that under acidic condition, the charged amino group of AA was difficult to form hydrogen bonds. Under basic conditions, OH¯ was easy to bind with quaternary ammonium resulting in difficulty for the formation of electrostatic force. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3.3.2 Effect of sample volume Sample volume determined the extraction efficiency to some extent with given initial polymer dosage, so a series of volumes from 5 to 25 mL were studied. As shown in Fig. 4B, extraction efficiency increased apparently with the increase of volume in range of 5 to 20 mL and was nearly invariable when larger volume was loaded. Sample volume of 20 mL was loaded for extraction for better extraction efficiency. 3.3.3 Effect of stirring speed Stirring played an important role during the loading process by accelerating and ensuring adequate dispersion of M-MIP particles in the sample solution. Extraction process www.jss-journal.com
Sample Preparation
J. Sep. Sci. 2015, 38, 2117–2125
2123
Figure 7. Chromatograms of rat plasma: (A) with MI-SPE. (B) without extraction; 1 jatrorrhizine, 2 palmatine, 3 berberine.
was conducted at the stirring speed of 100–400 rpm. As shown in Fig. 4C, the extraction capacities were low when the stirring rate was 100 rpm and nearly invariable at the rate from 200–400 rpm. To ensure the equilibrium of samples on the particle surface, the stirring rate of 400 rpm was chosen for further studies.
specific recognition sites. When there was an excess amount of berberine molecular in solution, more non-specific binding was reformed resulting in reduced selectivity.
3.3.4 Effect of extraction time
In the process of M-MI-SPE, the M-MIPs were immersed in solution all the time for loading with stirring and for desorption with ultrasonication. Therefore, there was the possibility that the coating of particles was destroyed. Besides, the residual of analytes in used particles would probably affect the reuse of M-MIPs. To study the stability, M-MIPs were used for five times continuously under the same conditions. The peak area in the first application was regarded as a unit and the ratio of the other peak areas with the first one was calculated. For the second time the ratios ranged from 100.12– 102.03% and 98.10–99.90% for the third time of the three alkaloids. The ratios of peak area were 99.84–107.18% and 98.97–104.72% for the fourth and fifth use respectively. The results indicated the good stability of M-MIP particles.
Extraction process needs enough time to reach adsorption equilibrium so that M-MIP particles can adsorb analytes as much as possible. Extraction time from 0.5 to 5 h was evaluated. The result (Fig. 4D) indicated that extraction efficiency increased obviously within a short time (0.5–1 h) and was almost the same from 2 to 5 h. In addition, extraction efficiency grew slowly between 1 and 2 h. In consideration of rapid extraction, 2 h was selected as the extraction time.
3.4 Adsorption experiment Adsorption experiment can evaluate the adsorption capacity of M-MIPs to berberine, as well as the selective difference between M-MIPs and M-NIPs. The result is shown in Fig. 5. The amounts of berberine bound to the M-MIP and M-NIP particles increased sharply and then slightly with the increment of berberine concentration. The adsorption equilibrium was achieved when berberine concentration exceeded 50 g/mL. The extraction capacity of M-MIPs was about twice that of M-NIPs within a low berberine concentration (0.01– 0.1 g/mL). However, when the concentration of berberine was over 0.1 g/mL, the difference of extraction yield between M-MIPs and M-NIPs was relatively small. The reason was probably that the M-MIP particles had a number of C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3.5 Stability of M-MIPs
3.6 Method validation By coupling with HPLC, the magnetic imprinted polymers were applied to extract the three protoberberine alkaloids in aqueous samples. This M-MI-SPE–HPLC method was validated and the analytical figures of merit including linearity, sensitivity and repeatability were listed in Table 1. Good linearity was obtained by the ratios of the HPLC peak areas (y) with corresponding concentrations (x, ng/mL). The correlation coefficients (R) were greater than 0.999 for all compounds in the range of 0.1–150 ng/mL for berberine, www.jss-journal.com
2124
J. Sep. Sci. 2015, 38, 2117–2125
J. Meng et al.
0.1–100 ng/mL for jatrorrhizine and palmatine. The LOD (S/N = 3) for berberine was 0.01 ng/mL, while that of jatrorrhizine and palmatine was 0.02 ng/mL. The LOQ (S/N = 10) was 0.08 ng/mL for jatrorrhizine, 0.07 ng/mL for palmatine and 0.03 ng/mL for berberine. Repeatability was investigated by loading the standard solution (100 ng/mL) for three times within the same day and the following three days. Intra-day and inter-day RSDs for the peak area were from 0.94–2.00% and 1.54–4.43%, respectively, indicating the good reproducibility of the method.
3.7 Application in real samples Under the optimum experimental conditions, the proposed method was used to determine jatrorrhizine, palmatine and berberine in Cortex phellodendri and rat plasma samples. The chromatograms of the initial Cortex phellodendri solution and extracted by M-MIPs are shown in Fig. 6. All three protoberberine alkaloids were well detected after extraction, the contents were calculated to be 0.26 mg/g for jatrorrhizine, 4.09 mg/g for palmatine and 7.66 mg/g for berberine. There were no interference peaks affecting the determination of the three compounds. To evaluate the accuracy of the proposed method, extraction recovery was tested by adding accurate amounts of mixed standards to the powder of Cortex phellodendri (Table 2). The recoveries are in range of 87.33–102.43%, with RSD less than 4.54%, indicating that the method can be used in traditional Chinese medicines. The method was further applied to determine the medicated rat plasma samples. Chromatograms of the plasma samples before and after extraction by M-MIPs are shown in Fig. 7. No peaks of the analytes were observed in plasma without M-MI-SPE. While the three components had obvious peaks after extraction, indicating the method had good extraction efficiency. Besides, the method had good clean-up efficiency by suppressing the interfering peaks arising from the plasma. The contents were calculated to be 90.00 ng/mL for jatrorrhizine, 45.76 ng/mL for palmatine and 171.60 ng/mL for berberine. The recovery was carried out in rat plasma by spiking with standard analytes and resulted in the range of 102.22–111.15%, with RSD less than 4.59% (Table 2). The results showed that the proposed method was effective for determination of the protoberberine alkaloids in plasma samples.
4 Concluding remarks In this work, we prepared novel magnetic molecularly imprinted polymers (M-MIPs) for selective extraction of protoberberine alkaloids. The M-MIPs showed high stability and good selectivity to protoberberine alkaloids. M-MIPs were used as the sorbent of SPE and the M-MI-SPE–HPLC method was developed for determination of protoberberine alkaloids in real samples such as in herb medicine of Cortex phellodendri and in rat plasma. By using M-MIPs as selective sorbent, C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
most of the matrix in the plasma sample was eliminated and three protoberberine alkaloids were selectively extracted and well determined. The results showed that M-MIPs have high clean-up efficiency and good extraction efficiency. This work was supported by the National Natural Science Foundation of China (Grant nos. 21375101 and 91417301), Natural Science Foundation of Hubei (No. 2014CFA077), Wuhan Science and Technology Bureau (No: 20140601010057), Innovation Seed Fund and Translational Medical Research Fund of Wuhan University School of Medicine, and the Experimental Research Project of Wuhan University Instrument and Facility Department. The authors have declared no conflict of interest.
5 References [1] Iwasa, K., Kim, H.-S., Wataya, Y., Lee, D.-U., Eur. J. Med. Chem. 1998, 33, 65–69. [2] Iwasa, K., Nishiyama, Y., Ichimaru, M., Moriyasu, M., Kim, H.-S., Wataya, Y., Yamori, T., Takashi, T., Lee, D.-U., Eur. J. Med. Chem. 1999, 34, 1077–1083. [3] Li, W., Huang, H., Zhang, Y., Fan, T., Liu, X., Xing, W., Niu, X., Eur. J. Pharmacol. 2013, 715, 62–71. [4] Cheng, P., Wang, B., Liu, X., Liu, W., Kang, W., Zhou, J., Zeng, J., Nat. Prod. Res. 2014, 28, 413–419. [5] Cushnie, T. T., Cushnie, B., Lamb, A. J., Int. J. Antimicrob. Ag. 2014, 26, 343–356. [6] Wangchuk, P., Keller, P. A., Pyne, S. G., Lie, W., Willis, A. C., Rattanajak, R., Kamchonwongpaisan, S., J. Ethnopharmacol. 2013, 150, 953–959. [7] Bhowmik, D., Hossain, M., Buzzetti, F., D’ Auria, R., Lombardi, P., Kumar, G. S., J. Phys.Chem. B. 2012, 116, 2314– 2324. [8] Kumari, S., Kaladhar, D., Sandeep Solmon, K., Malla, R., Kishore, G., Process Biochem. 2013, 48, 1565–1571. [9] Zhou, Y., He, P., Liu, A., Zhang, L., Liu, Y., Dai, R., Phytother. Res. 2012, 26, 772–777. [10] Kong, W., Wei, J., Abidi, P., Lin, M., Inaba, S., Li, C., Wang, Y., Wang, Z., Si, S., Pan, H., Wang, S., Wu, J., Wang, Y., Li, Z., Liu, J., Jiang, J. D., Nat. Med. 2004, 10, 1344–1351. [11] Kong, W. J., Wei, J., Zuo, Z. Y., Wang, Y. M., Song, D. Q., You, X. F., Zhao, L. X., Pan, H. N., Jiang, J. D., Metabolism 2008, 57, 1029–1037. [12] Yang, X., Wang, J., Luo, J., Kong, L., J. Liq. Chromatogr. R. T. 2012, 35, 1842–1852. ´ ´ [13] Suau, R., Rico, R., Lopez-Romero, J. M., Najera, F., Cuevas, A., Phytochemistry 1998, 49, 2545–2549. [14] Chu, B., Lou, D., Yu, P., Hu, S., Shen, S., J. Chromatogr. A. 2011, 1218, 7248–7253. [15] Chen, F. F., Xie, X. Y., Shi, Y. P., Talanta 2013, 115, 482–489. [16] Chen, F. F., Xie, X. Y., Shi, Y. P., J. Chromatogr. A. 2013, 1300, 112–118. [17] Li, J., Zhang, S., Chen, C., Zhao, G., Yang, X., Li, J., Wang, X., ACS. Appl. Mater. Interfaces 2012, 4, 4991–5000.
www.jss-journal.com
Sample Preparation
J. Sep. Sci. 2015, 38, 2117–2125
2125
[18] Hu, Y., Huang, Z., Liao, J., Li, G., Anal. Chem. 2013, 85, 6885–6893.
[24] Shi, S., Guo, J., You, Q., Chen, X., Zhang, Y., Chem. Eng. J. 2014, 243, 485–493.
[19] Mehdinia, A., Esfandiarnejad, R., Jabbari, A., J. Sep. Sci. 2015, 38, 141–147.
[25] Sadeghi, S., Jahani, M., Food. Chem. 2013, 141, 1242– 1251.
[20] Deng, Y., Qi, D., Deng, C., Zhang, X., Zhao, D., J. Am. Chem. Soc. 2008, 130, 28–29.
[26] Chen, F.-F., Wang, R., Shi, Y.-P., Talanta 2012, 89, 505–512.
[21] Yılmaz, H., Basan, H., J. Sep. Sci. 2015, 00, 1–7.
[27] Lai, J. P., Chen, F., Sun, H., Fan, L., Liu, G. L., J. Sep. Sci. 2014, 37, 1170–1176.
103–
[28] Shi, H. L., Peng, S. L., Sun, J., Liu, Y. M., Zhu, Y. T., Qing, L. S., Liao, X., J. Sep. Sci. 2014, 37, 704–710.
[23] Mou, X., Zhang, W., Chen, Z., J. Appl. Polym. Sci. 2013, 130, 1873–1878.
[29] Wang, Y., Wang, L., Tian, T., Hu, X., Yang, C., Xu, Q., Analyst 2012, 137, 2400–2405.
[22] Zhang, 109.
W.,
Chen,
Z.,
Talanta
2013,
103,
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
ISSN 1615-9306 · JSSCCJ 38 (12) 2007–2192 (2015) · Vol. 38 · No. 12 · June 2015 · D 10609
JOURNAL OF
SEPARATION SCIENCE
Methods Chromatography · Electroseparation Applications Biomedicine · Foods · Environment
12 15
www.jss-journal.com
2117
J. Sep. Sci. 2015, 38, 2117–2125
Jiawei Meng1,2 Wenpeng Zhang Tao Bao Zilin Chen1,2 1 Key
Laboratory of Combinatorial Biosynthesis and Drug Discovery (Wuhan University), Ministry of Education, and Wuhan University School of Pharmaceutical Sciences, Wuhan, China 2 State Key Laboratory of Transducer Technology, Chinese Academy of Sciences, Beijing, China Received March 6, 2015 Revised March 24, 2015 Accepted March 24, 2015
Research Article
Novel molecularly imprinted magnetic nanoparticles for the selective extraction of protoberberine alkaloids in herbs and rat plasma In this work, a novel magnetic nanomaterial functionalized with a molecularly imprinted polymer was prepared for the extraction of protoberberine alkaloids. Molecularly imprinted polymers were made on the surface of Fe3 O4 nanoparticles by using berberine as template, acetonitrile/water as porogen, acrylamide as functional monomer and ethylene glycol dimethacrylate as cross-linker. The optimized molar ratio of template/functional monomer was 1:7. The polymeric magnetic nanoparticles were characterized by transmission electron microscopy and Fourier transform infrared spectroscopy. The stability and adsorption capacity of the molecularly imprinted polymers were investigated. The molecularly imprinted polymers were used as a selective sorbent for the magnetic molecularly imprinted solid-phase extraction and determination of jatrorrhizine, palmatine, and berberine. Extraction parameters were studied including loading pH, sample volume, stirring speed, and extraction time. Finally, a magnetic molecularly imprinted solid-phase extraction coupled to high-performance liquid chromatography method was developed. Under the optimized conditions, the method showed good linear range of 0.1–150 ng/mL for berberine and 0.1– 100 ng/mL for jatrorrhizine and palmatine. The limit of detection was 0.01 ng/mL for berberine and 0.02 ng/mL for jatrorrhizine and palmatine. The proposed method has been applied to determine protoberberine alkaloids in Cortex phellodendri and rat plasma samples. The recoveries ranged from 87.33–102.43%, with relative standard deviation less than 4.54% in Cortex phellodendri and from 102.22–111.15% with relative standard deviation less than 4.59% in plasma. Keywords: Herbs / Magnetic molecularly imprinted polymers / Protoberberine alkaloids / Rat plasma / Solid-phase extraction DOI 10.1002/jssc.201500264
Additional supporting information may be found in the online version of this article at the publisher’s web-site
1 Introduction Protoberberine alkaloids are the main active components of many herbs and they have an isoquinoline structure (Supporting Information Fig. S1). The activity of the protoberberine Correspondence: Dr. Zilin Chen Luojia Chair Professor. Vice Dean and Institute Director School of Pharmaceutical Sciences, Wuhan University, Wuhan 430071, China Email: [email protected] Fax: 86-27-68759850
Abbreviations: AA, acrylamide; M-MIP, magnetic molecularly imprinted polymer; M-MI-SPE, magnetic molecularly imprinted solid phase extraction; M-NIP, magnetic non-imprinted polymer; MNP, magnetic nanoparticle; TEOS, tetraethylorthosilicate; MAPS, 3methacryloxypropyltrimethoxysilane; MAA, methacrylic acid; 4-VP, 4-vinylpyridine; AIBN, 2,2 -azobis (2-methylpropionitrile)
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
alkaloids depends on the oxygen substituents in rings A, C, D and the quaternary nitrogen atom [1, 2]. Protoberberine alkaloids possess activities such as anti-inflammatory [3], antibacterial [4, 5], antimalarial [6], antitumor [7, 8], and have been used for the treatment of gastrointestinal diseases and bacterial diarrhea. They can also affect the pharmacokinetics of other drugs [9] and the content of serum cholesterol [10,11]. Berberine, palmatine and jatrorrhizine are most commonly found in many herbs, such as Coptidischinensis [12], Cortex phellodendri, and Berberis vulgaris [13]. Protoberberine alkaloids also serve as the main active component of dozens of compound medicinal preparations such as Fufang Huangbai Ye and Huanglian Shangqing Wan. Generally, the matrix of herbs is very complex, leading to difficulty in the QC of the herbs. Besides, determination of the active alkaloids in biofluids such as plasma and urine is necessary for pharmacokinetic study. The quantitative analysis would be more difficult due to the complexity of matrix
www.jss-journal.com
2118
J. Sep. Sci. 2015, 38, 2117–2125
J. Meng et al.
from both herbs and biofluids. Therefore, the extraction of target analytes as well as clean-up of interferent is necessary for the determination at trace level in real samples. In recent years, magnetic nanoparticle-based techniques have attracted more and more attention in separation science. Magnetic nanoparticles (MNPs) which possess a large surface area can move, gather and disperse conveniently by the control of the magnetic field [14–16]. Besides, extraction materials could be modified onto the surface of nanoparticles to facilitate their application in extraction, such as polymer [15, 16], graphene [17], metal–organic frameworks [18], polyaniline [19], silica shell [20], and octadecylsilyl [14]. The variety of sorbents expands the application of MNPs in different kinds of samples. However, the selectivity of these sorbents is still limited, so that satisfactory clean-up efficiency would not be obtained in very complex matrix such as herbs and biofluids. Molecularly imprinted polymers (MIPs) are synthesized materials with functional recognition sites and complementary cavities in accordance with shape and functional groups of the template molecules. MIPs can be obtained by copolymerization of functional and cross-linking monomers in a mixture solution containing molecular template, initiator and porogen. After polymerization, the template is removed and the MIPs with the tailor-made binding sites and cavities are obtained [15, 16, 21–24]. MIPs possess similar selectivity with natural enzyme but have advantages of higher stability and less cost. MIPs can be used in a wide range of samples such as food [25], traditional Chinese medicines [26], biological and environmental samples [27,28]. In a previous work, we have successfully prepared a berberine-MIP monolith inside a micropipette tip for -SPE of berberine. However, to the best of our knowledge, MIP-based MNPs have not been reported for separation or extraction of protoberberine alkaloids. In this work, we aimed to fabricate berberine-MIP based MNPs and apply to extraction of protoberberine alkaloids in complex samples. The characterization, synthetic conditions, adsorption capacities and stability of the magnetic molecularly imprinted polymers (M-MIPs) were studied to obtain good selectivity. Finally, an M-MIP-based extraction-HPLC method was developed for the determination of protoberberine alkaloids in Cortex phellodendri and rat plasma samples.
2 Materials and Methods 2.1 Materials and Chemicals Berberine, jatrorrhizine, palmatine, tetraethylorthosilicate (TEOS), and 3-methacryloxypropyltrimethoxysilane (MAPS) were obtained from Aladdin-reagent (Shanghai, China). Ethyleneglyol dimethacrylate was obtained from Alfa-Aesar (Lancashire, UK). Acrylamide (AA), methacrylic acid (MAA), 4-vinylpyridine (4-VP), and 2,2 -azobis(2-methylpropionitrile) (AIBN) were obtained from Shanghai Reagent Factory (Shanghai, China). The above reagents were all analytical reagent grade. Methanol and acetonitrile were HPLC grade and purchased from Sigma companies (USA). C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
All experimental protocols involving animals were reviewed and approved by the Institutional Animal Experimentation Committee of Wuhan University. Sprague–Dawley (SD) rats weighing 150–200 g, were obtained by the Experimental Animal Center of Wuhan University (Wuhan, China). All the SD rats were maintained under standard conditions with normal access to food and water. 2.2 Instrumentation A Shimadzu HPLC system (Tokyo, Japan) consisted of two 20AD pumps, a 20A3 degasser, 20A UV detector, and thermostat controlled column compartment were used for chromatographic analysis. Data were acquired and processed using Shimadzu LC Solution software. The analytical column was a 250 mm × 4.6 mm, 5m C-18 column from GL Science (Tokyo, Japan). The mobile phase consisted of acetonitrile and KH2 PO4 (10 mM) buffer (32:48, v/v) with a flow rate of 0.8 mL/min. The detection wavelength was set at 345 nm and column temperature was 25 ⬚C. 2.3 Preparation of M-MIPs The synthesis protocol of M-MIPs is shown in Supporting Information Fig. S2. The Fe3 O4 nanoparticles were synthesized according to a reported procedure by Wang et al. [29]. Fe3 O4 @SiO2 -MAPS particles were fabricated according to the sol–gel method. MNPs were activated by HCl (0.1 M), collected and then rinsed with H2 O and ethanol for three times. The dried MNPs were dispersed in isopropanol (80 mL)/H2 O (6.4 mL) solution, followed by adding NH3 •H2 O (8 mL), TEOS (1.6 mL) and MAPS (1.6 mL) sequentially under stirring and reaction for 12 h. The synthesized nanoparticles were collected by a magnet and washed with H2 O and ethanol for several times. The polymerization mixture contained the template (berberine, 0.022 mmol), functional monomer (AA, 0.157 mmol), acetonitrile/H2 O (8 mL, 9:1, v/v), Fe3 O4 @SiO2 MAPS nanoparticles (80 mg), and cross-linker (ethyleneglyol dimethacrylate, 0.679 mmol). The solution was stirred for 2 h at room temperature, followed by adding the initiator AIBN (40 mg). After deoxygenation by ultrasonication for 15 min, purging with nitrogen for 15 min, the system was incubated at 60 ⬚C for 20 h. The M-MIPs were eluted with methanol/acetic acid (9:1, v/v) and methanol several times in an ultrasonic bath. The resulting particles were dried at 60 ⬚C overnight. As a comparison, magnetic non-imprinted polymers (M-NIPs) were prepared using the same method but without addition of berberine during the polymerization process. 2.4 Procedure of M-MI-SPE Jatrorrhizine, palmatine, and berberine were dissolved in 20 mL of phosphate buffer (pH 7) at concentration of www.jss-journal.com
Sample Preparation
J. Sep. Sci. 2015, 38, 2117–2125
2119
Figure 1. FTIR spectra of the Fe3 O4 (A), Fe3 O4 @SiO2 -MAPS (B), and M-MIPs (C).
mixed solution was extracted and analyzed as presented in Section 2.4. The equilibrium adsorption amount of berberine (Q, g) was calculated using the following equation: Q = CV
(1)
where C (g/mL) is the concentrations of berberine after extraction and V (mL) is the volume of eluent (0.2 mL).
2.6 Application of M-MIPs in Cortex phellodendri and rat plasma samples
Figure 2. TEM images of M-MIPs.
100 ng/mL. The prepared berberine M-MIPs (5 mg) were mixed with 20 mL of the standard solution at a stirring speed of 400 rpm. After loading for 2 h, the particles were separated from the solution and eluted with 0.2 mL acetonitrile by sonication for 5 min. Then the eluent (20 L) was analyzed by HPLC.
2.5 Adsorption experiment To evaluate the adsorption and desorption capacity of M-MIPs toward template molecule, the adsorption experiments were operated. M-MIPs and M-NIPs (5 mg) were suspended in 20 mL of phosphate buffer (pH 7) solution with berberine concentrations ranging from 0.01–100 g/mL. The C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Cortex phellodendri was sliced and ground into powder. Then 0.2 g of the powder was extracted with 5 mL of purified water for 40 min by sonication. The crude extraction solution was centrifuged at 10 000 rpm for 8 min and the supernatant was diluted to 10 times. Then 30 L of the resulting solution was added into 20 mL of phosphate buffer (pH 7), extracted and analyzed as represented in 2.4. The powder of Cortex phellodendri was sieved through a 75 m sieve. 1.0 g of the obtained particles was weighed accurately and extracted by 8 mL of water for 40 min in an ultrasonic bath. After the rats received a single oral dose of 4 mL of the above extract for 60 min, the medicated blood sample was collected in a heparinized Eppendorf tube and centrifuged at 10 000 rpm for 10 min. The harvested plasma was added with equal volume of 5% trichloroacetic acid solution to remove protein. After centrifugation for 10 min at 10 000 rpm, the supernatant was added slowly with NaOH solution (1 mol/L) to adjust the pH value to 7. Then 0.5 mL of the solution was added into 20 mL of phosphate buffer (pH 7), extracted and analyzed as presented in Section 2.4. www.jss-journal.com
2120
J. Meng et al.
J. Sep. Sci. 2015, 38, 2117–2125
Figure 3. The effects of different functional monomers (A), different ratios of berberine/ functional monomer (B), and different porogens (C).
3 Results and discussion 3.1 Characterization of magnetic nanoparticles The prepared M-MIPs were characterized by FTIR spectroscopy. The FTIR spectra of Fe3 O4 , Fe3 O4 @SiO2 -MAPS and M-MIPs are shown in Fig. 1. The peak at 630 cm−1 attributed to Fe–O stretching bond for Fe3 O4 nanoparticles (Fig. 1A). As shown in Fig. 1B, the peak at 1162 cm−1 corresponded to Si–O–Si stretching. The absorbance bands at 1729, 2946 cm−1 were caused by stretching of C=O and C–H, respectively, indicating that MAPS was modified onto Fe3 O4 successfully. Compared with Fe3 O4 @SiO2 -MAPS, the peaks at 1729 and 2946 cm−1 became stronger evidently (Fig. 1C), indicating MIP layer had been formed on the surface of Fe3 O4 @SiO2 -MAPS. The morphological structure of M-MIPs was characterized by TEM. As shown in Fig. 2, a smooth and uniform layer was coated on the Fe3 O4 @SiO2 -MAPS face, indicating that MIPs had been introduced. It can be seen that the diameter of pure MNPs was about 80 nm and the thickness of MIP layer was approximately 30 nm. There were connected MIP particles in the TEM view, while these MIPs will not affect the extraction capacity.
3.2 Optimization of synthesis conditions Firstly, Fe3 O4 particles were prepared by a co-precipitation method at 80 ⬚C. The pH value of solution was adjusted with NaOH to 9–10 to avoid the synthesis of iron hydroxide. When pH value was too low or high, most of the final product was probably iron hydroxide. After the reaction, the process of C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
stirring in oil bath for 30 min and ultrasonication for 20 min was repeated three times to promote particles dispersion. Secondly, the surface of MNPs was modified. As the common method, the MNPs were coated with silica firstly and then modified with MAPS. However in this study, double bonds were introduced onto the Fe3 O4 surface with MAPS by one step. MAPS participated in the silica surface when Fe3 O4 was coated with silica. The proposed method possessed advantages such as simple operation and freedom of anhydrous toluene. 3.2.1 Effect of functional monomers Functional monomers played an important role for the selectivity of M-MIPs because monomers provided functional groups to form accurate and selective binding sites with the templates. Different monomers including AA, MAA, and 4-VP were tested respectively to evaluate the specific recognition of imprinted polymer. As can be seen in Fig. 3A, the best selectivity and extraction efficiency were obtained when AA was used as functional monomer, therefore, AA was chosen for polymerization. The possible reason was that four hydrogen bonds can be formed between hydrogen atoms on amino group and four oxygen atoms of berberine, the electrostatic interaction can be formed between carbonyl group and quaternary ammonium of berberine. It was the hydrogen bonds and electrostatic force that lead to the self-assembled polymerization of berberine and AA. The molar ratio of template/functional monomer can influence the density and imprinting efficiency of M-MIPs. Over-high ratios diminish the selectivity due to higher non-specific binding sites in MIPs, but lower ones induce fewer template-monomer complexes resulting in less binding www.jss-journal.com
Sample Preparation
J. Sep. Sci. 2015, 38, 2117–2125
2121
Figure 4. The effects of sample pH (A), sample volume (B), stirring speed (C), and extraction time (D).
capacity. The ratios were studied from 1:3 to 1:8 to prepare M-MIPs with berberine. As shown in Fig. 3B, largest difference between MIPs and NIPs was observed when the molar ratio was 1:7, and it was selected for further studies. 3.2.2 Effect of porogens Porogens were not only used to generate cavities, but also served as solvent for polymerization. An ideal porogen should allow adequate dissolution of molecule template, functional monomer and cross-linker, and formation of appropriate tailor-made cavities. Berberine dissolved little in solvents with little polar organic, while higher polar solvents probably disturbed the main interaction of berberine-AA, such as hydrogen bond. In consideration of these factors, DMSO, DMF/water (9:1, v/v) and acetonitrile/water (9:1, v/v) were investigated. As can be seen in Fig. 3C, when acetonitrile/water (9:1, v/v) was used as the porogen, significant difference was observed between M-MIPs and M-NIPs.
Figure 5. Extraction yield curves of M-MIPs and M-NIPs for berberine.
solution of three protoberberine alkaloids (100 ng/mL for each) with 3 mg of polymers. 3.3 Optimization of extraction conditions 3.3.1 Effect of loading pH It is important to optimize extraction conditions to achieve high adsorption efficiency of M-MIPs. The conditions possible to affect extraction efficiency were studied in the mixed C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
pH of sample solution affected ionization degree of the target molecules, surface charge of M-MIP particles, and further www.jss-journal.com
2122
J. Sep. Sci. 2015, 38, 2117–2125
J. Meng et al.
Table 1. Analytical parameters of the M-MI-SPE-HPLC methods for analysis of protoberberine alkaloids
RSD% (n = 3) Analytes
Range (ng/mL)
Regression equation
R
LOD (ng/mL)
LOQ (ng/mL)
Intra-day
Inter-day
Jatrorrhizine Palmatine Berberine
0.1-100 0.1-100 0.1-150
y = 4066.90x + 2781.88 y = 4511.22x + 3011.72 y = 5808.02x + 23851.32
0.9996 0.9997 0.9991
0.02 0.02 0.01
0.08 0.07 0.03
0.94 2.00 1.22
4.43 1.54 1.59
Figure 6. Chromatograms of Cortex phellodendri: (A) with MI-SPE. (B) without extraction; 1 jatrorrhizine, 2 palmatine, 3 berberine. Table 2. Results and recoveries in Cortex phellodendri and rat plasma samples
Cortex phellodendri
Plasma
Samples
Added (mg/g)
Recovery (%)
RSD (%, n = 3)
Added (ng/mL)
Recovery (%)
RSD (%, n = 3)
Jatrorrhizine Palmatine Berberine
3.33 3.33 3.33
101.20 101.35 94.07
2.34 4.55 4.22
20 20 20
111.18 104.22 99.95
4.59 1.76 3.06
affected the formation of hydrogen bonds between adsorbed molecules and M-MIPs. Therefore, sample solution with pH value varied from 3 to 9 was investigated. As shown in Fig. 4A, the adsorption capacities of M-MIPs to the three protoberberine alkaloids increased gradually from pH 3 to 7, but the ones of jatrorrhizine and palmatine decreased obviously at pH 7. The optimal pH for berberine was pH 8, while the best pH for jatrorrhizine and palmatine were pH 7. Considering the good extraction efficiency of three components, pH 7 was selected. As described in Section 3.2.1, the reasons were probably that under acidic condition, the charged amino group of AA was difficult to form hydrogen bonds. Under basic conditions, OH¯ was easy to bind with quaternary ammonium resulting in difficulty for the formation of electrostatic force. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3.3.2 Effect of sample volume Sample volume determined the extraction efficiency to some extent with given initial polymer dosage, so a series of volumes from 5 to 25 mL were studied. As shown in Fig. 4B, extraction efficiency increased apparently with the increase of volume in range of 5 to 20 mL and was nearly invariable when larger volume was loaded. Sample volume of 20 mL was loaded for extraction for better extraction efficiency. 3.3.3 Effect of stirring speed Stirring played an important role during the loading process by accelerating and ensuring adequate dispersion of M-MIP particles in the sample solution. Extraction process www.jss-journal.com
Sample Preparation
J. Sep. Sci. 2015, 38, 2117–2125
2123
Figure 7. Chromatograms of rat plasma: (A) with MI-SPE. (B) without extraction; 1 jatrorrhizine, 2 palmatine, 3 berberine.
was conducted at the stirring speed of 100–400 rpm. As shown in Fig. 4C, the extraction capacities were low when the stirring rate was 100 rpm and nearly invariable at the rate from 200–400 rpm. To ensure the equilibrium of samples on the particle surface, the stirring rate of 400 rpm was chosen for further studies.
specific recognition sites. When there was an excess amount of berberine molecular in solution, more non-specific binding was reformed resulting in reduced selectivity.
3.3.4 Effect of extraction time
In the process of M-MI-SPE, the M-MIPs were immersed in solution all the time for loading with stirring and for desorption with ultrasonication. Therefore, there was the possibility that the coating of particles was destroyed. Besides, the residual of analytes in used particles would probably affect the reuse of M-MIPs. To study the stability, M-MIPs were used for five times continuously under the same conditions. The peak area in the first application was regarded as a unit and the ratio of the other peak areas with the first one was calculated. For the second time the ratios ranged from 100.12– 102.03% and 98.10–99.90% for the third time of the three alkaloids. The ratios of peak area were 99.84–107.18% and 98.97–104.72% for the fourth and fifth use respectively. The results indicated the good stability of M-MIP particles.
Extraction process needs enough time to reach adsorption equilibrium so that M-MIP particles can adsorb analytes as much as possible. Extraction time from 0.5 to 5 h was evaluated. The result (Fig. 4D) indicated that extraction efficiency increased obviously within a short time (0.5–1 h) and was almost the same from 2 to 5 h. In addition, extraction efficiency grew slowly between 1 and 2 h. In consideration of rapid extraction, 2 h was selected as the extraction time.
3.4 Adsorption experiment Adsorption experiment can evaluate the adsorption capacity of M-MIPs to berberine, as well as the selective difference between M-MIPs and M-NIPs. The result is shown in Fig. 5. The amounts of berberine bound to the M-MIP and M-NIP particles increased sharply and then slightly with the increment of berberine concentration. The adsorption equilibrium was achieved when berberine concentration exceeded 50 g/mL. The extraction capacity of M-MIPs was about twice that of M-NIPs within a low berberine concentration (0.01– 0.1 g/mL). However, when the concentration of berberine was over 0.1 g/mL, the difference of extraction yield between M-MIPs and M-NIPs was relatively small. The reason was probably that the M-MIP particles had a number of C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3.5 Stability of M-MIPs
3.6 Method validation By coupling with HPLC, the magnetic imprinted polymers were applied to extract the three protoberberine alkaloids in aqueous samples. This M-MI-SPE–HPLC method was validated and the analytical figures of merit including linearity, sensitivity and repeatability were listed in Table 1. Good linearity was obtained by the ratios of the HPLC peak areas (y) with corresponding concentrations (x, ng/mL). The correlation coefficients (R) were greater than 0.999 for all compounds in the range of 0.1–150 ng/mL for berberine, www.jss-journal.com
2124
J. Sep. Sci. 2015, 38, 2117–2125
J. Meng et al.
0.1–100 ng/mL for jatrorrhizine and palmatine. The LOD (S/N = 3) for berberine was 0.01 ng/mL, while that of jatrorrhizine and palmatine was 0.02 ng/mL. The LOQ (S/N = 10) was 0.08 ng/mL for jatrorrhizine, 0.07 ng/mL for palmatine and 0.03 ng/mL for berberine. Repeatability was investigated by loading the standard solution (100 ng/mL) for three times within the same day and the following three days. Intra-day and inter-day RSDs for the peak area were from 0.94–2.00% and 1.54–4.43%, respectively, indicating the good reproducibility of the method.
3.7 Application in real samples Under the optimum experimental conditions, the proposed method was used to determine jatrorrhizine, palmatine and berberine in Cortex phellodendri and rat plasma samples. The chromatograms of the initial Cortex phellodendri solution and extracted by M-MIPs are shown in Fig. 6. All three protoberberine alkaloids were well detected after extraction, the contents were calculated to be 0.26 mg/g for jatrorrhizine, 4.09 mg/g for palmatine and 7.66 mg/g for berberine. There were no interference peaks affecting the determination of the three compounds. To evaluate the accuracy of the proposed method, extraction recovery was tested by adding accurate amounts of mixed standards to the powder of Cortex phellodendri (Table 2). The recoveries are in range of 87.33–102.43%, with RSD less than 4.54%, indicating that the method can be used in traditional Chinese medicines. The method was further applied to determine the medicated rat plasma samples. Chromatograms of the plasma samples before and after extraction by M-MIPs are shown in Fig. 7. No peaks of the analytes were observed in plasma without M-MI-SPE. While the three components had obvious peaks after extraction, indicating the method had good extraction efficiency. Besides, the method had good clean-up efficiency by suppressing the interfering peaks arising from the plasma. The contents were calculated to be 90.00 ng/mL for jatrorrhizine, 45.76 ng/mL for palmatine and 171.60 ng/mL for berberine. The recovery was carried out in rat plasma by spiking with standard analytes and resulted in the range of 102.22–111.15%, with RSD less than 4.59% (Table 2). The results showed that the proposed method was effective for determination of the protoberberine alkaloids in plasma samples.
4 Concluding remarks In this work, we prepared novel magnetic molecularly imprinted polymers (M-MIPs) for selective extraction of protoberberine alkaloids. The M-MIPs showed high stability and good selectivity to protoberberine alkaloids. M-MIPs were used as the sorbent of SPE and the M-MI-SPE–HPLC method was developed for determination of protoberberine alkaloids in real samples such as in herb medicine of Cortex phellodendri and in rat plasma. By using M-MIPs as selective sorbent, C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
most of the matrix in the plasma sample was eliminated and three protoberberine alkaloids were selectively extracted and well determined. The results showed that M-MIPs have high clean-up efficiency and good extraction efficiency. This work was supported by the National Natural Science Foundation of China (Grant nos. 21375101 and 91417301), Natural Science Foundation of Hubei (No. 2014CFA077), Wuhan Science and Technology Bureau (No: 20140601010057), Innovation Seed Fund and Translational Medical Research Fund of Wuhan University School of Medicine, and the Experimental Research Project of Wuhan University Instrument and Facility Department. The authors have declared no conflict of interest.
5 References [1] Iwasa, K., Kim, H.-S., Wataya, Y., Lee, D.-U., Eur. J. Med. Chem. 1998, 33, 65–69. [2] Iwasa, K., Nishiyama, Y., Ichimaru, M., Moriyasu, M., Kim, H.-S., Wataya, Y., Yamori, T., Takashi, T., Lee, D.-U., Eur. J. Med. Chem. 1999, 34, 1077–1083. [3] Li, W., Huang, H., Zhang, Y., Fan, T., Liu, X., Xing, W., Niu, X., Eur. J. Pharmacol. 2013, 715, 62–71. [4] Cheng, P., Wang, B., Liu, X., Liu, W., Kang, W., Zhou, J., Zeng, J., Nat. Prod. Res. 2014, 28, 413–419. [5] Cushnie, T. T., Cushnie, B., Lamb, A. J., Int. J. Antimicrob. Ag. 2014, 26, 343–356. [6] Wangchuk, P., Keller, P. A., Pyne, S. G., Lie, W., Willis, A. C., Rattanajak, R., Kamchonwongpaisan, S., J. Ethnopharmacol. 2013, 150, 953–959. [7] Bhowmik, D., Hossain, M., Buzzetti, F., D’ Auria, R., Lombardi, P., Kumar, G. S., J. Phys.Chem. B. 2012, 116, 2314– 2324. [8] Kumari, S., Kaladhar, D., Sandeep Solmon, K., Malla, R., Kishore, G., Process Biochem. 2013, 48, 1565–1571. [9] Zhou, Y., He, P., Liu, A., Zhang, L., Liu, Y., Dai, R., Phytother. Res. 2012, 26, 772–777. [10] Kong, W., Wei, J., Abidi, P., Lin, M., Inaba, S., Li, C., Wang, Y., Wang, Z., Si, S., Pan, H., Wang, S., Wu, J., Wang, Y., Li, Z., Liu, J., Jiang, J. D., Nat. Med. 2004, 10, 1344–1351. [11] Kong, W. J., Wei, J., Zuo, Z. Y., Wang, Y. M., Song, D. Q., You, X. F., Zhao, L. X., Pan, H. N., Jiang, J. D., Metabolism 2008, 57, 1029–1037. [12] Yang, X., Wang, J., Luo, J., Kong, L., J. Liq. Chromatogr. R. T. 2012, 35, 1842–1852. ´ ´ [13] Suau, R., Rico, R., Lopez-Romero, J. M., Najera, F., Cuevas, A., Phytochemistry 1998, 49, 2545–2549. [14] Chu, B., Lou, D., Yu, P., Hu, S., Shen, S., J. Chromatogr. A. 2011, 1218, 7248–7253. [15] Chen, F. F., Xie, X. Y., Shi, Y. P., Talanta 2013, 115, 482–489. [16] Chen, F. F., Xie, X. Y., Shi, Y. P., J. Chromatogr. A. 2013, 1300, 112–118. [17] Li, J., Zhang, S., Chen, C., Zhao, G., Yang, X., Li, J., Wang, X., ACS. Appl. Mater. Interfaces 2012, 4, 4991–5000.
www.jss-journal.com
Sample Preparation
J. Sep. Sci. 2015, 38, 2117–2125
2125
[18] Hu, Y., Huang, Z., Liao, J., Li, G., Anal. Chem. 2013, 85, 6885–6893.
[24] Shi, S., Guo, J., You, Q., Chen, X., Zhang, Y., Chem. Eng. J. 2014, 243, 485–493.
[19] Mehdinia, A., Esfandiarnejad, R., Jabbari, A., J. Sep. Sci. 2015, 38, 141–147.
[25] Sadeghi, S., Jahani, M., Food. Chem. 2013, 141, 1242– 1251.
[20] Deng, Y., Qi, D., Deng, C., Zhang, X., Zhao, D., J. Am. Chem. Soc. 2008, 130, 28–29.
[26] Chen, F.-F., Wang, R., Shi, Y.-P., Talanta 2012, 89, 505–512.
[21] Yılmaz, H., Basan, H., J. Sep. Sci. 2015, 00, 1–7.
[27] Lai, J. P., Chen, F., Sun, H., Fan, L., Liu, G. L., J. Sep. Sci. 2014, 37, 1170–1176.
103–
[28] Shi, H. L., Peng, S. L., Sun, J., Liu, Y. M., Zhu, Y. T., Qing, L. S., Liao, X., J. Sep. Sci. 2014, 37, 704–710.
[23] Mou, X., Zhang, W., Chen, Z., J. Appl. Polym. Sci. 2013, 130, 1873–1878.
[29] Wang, Y., Wang, L., Tian, T., Hu, X., Yang, C., Xu, Q., Analyst 2012, 137, 2400–2405.
[22] Zhang, 109.
W.,
Chen,
Z.,
Talanta
2013,
103,
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
