Anal Bioanal Chem DOI 10.1007/s00216-015-8764-5


Screening bioactive compounds from Ligusticum chuanxiong by high density immobilized human umbilical vein endothelial cells Qian Li 1 & Jing Wang 1 & Guangxin Liu 1 & Huanmei Sun 1 & Liujiao Bian 1 & Xinfeng Zhao 1 & Xiaohui Zheng 1

Received: 25 January 2015 / Revised: 28 April 2015 / Accepted: 5 May 2015 # Springer-Verlag Berlin Heidelberg 2015

Abstract High throughput screening methodologies play a very important role in screening bioactive compounds from complex media. In this work, a new strategy for attaching cells onto amino microspheres using human umbilical vein endothelial cells (HUVECs) as a probe was developed. The immobilization depended on the specific affinity between integrin on the cells and the RGD peptide, which was coated on poly[oligo (ethylene glycol) methacrylate] by atom transfer radical polymerization. Validated application of the stationary phase was performed in the analysis of Ligusticum chuanxiong extraction by high performance affinity chromatography-mass spectrometry. Three compounds were screened as the bioactive compounds of Ligusticum chuanxiong. Two of them were identified as 3-butylhexahydroisobenzofuran-1(3H)-one and tetramethylpyrazine (TMP), whereas the other one remains indistinct. The association constant of vascular endothelial growth factor (VEGF) and TMP binding to VEGF receptor (VEGFR) on HUVECs were calculated to be (1.04±0.08)×1011 M−1 and (9.84± 1.11)×108 M−1 by zonal elution. Molecular docking showed that one hydrogen bond was formed between N atom of TMP and 3-N atom of imidazole group in histidine223 of VEGFR. Both zonal elution and molecular docking indicated that TMP and VEGF bind to the same site of VEGFR on HUVECs. It is possible to become a promising tool for high throughput screening of the bioactive compounds binding to HUVECs through broad application of the stationary phase.

* Xinfeng Zhao z[email protected] 1

Key Laboratory of Resource Biology and Biotechnology in Western China, Ministry of Education, College of Life Sciences, Northwest University, Xi’an 710069, China

Keywords Poly[oligo (ethylene glycol) methacrylate] . Affinity chromatography . RGD peptide . Vascular endothelial growth factor receptor . Human umbilical vein endothelial cells . Surface-initiated atom transfer radical polymerization

Introduction Traditional medicines have been the primary means to manage health of both animals and humans for thousands of years in America, Europe, China, Korea, and Japan [1, 2]. The rational use of natural products is believed to be the material basis of the curative effect of these therapies; it has also been utilized as the main resource for drug discovery in the last two decades [3]. However, thousands of components are included in traditional medicines, which produces a very complex issue when it comes to the recognition, separation, and purification of bioactive components from the medicines [4]. High throughput screening technologies have proven to be the potential assays to address this issue [5]. The major strategy for drug development from traditional medicines [6] is sequential activity-guided fractionation of the target ingredients and their derivatives and subsequent standardized evaluation according to the guidelines of US Food and Drug Administration. The disadvantages of this strategy are that it is labor intensive, time consuming, and costly. Several modern techniques, such as computer-aided drug discovery and development [7], high throughput screening [8], and fragment-based drug discovery [9], have the potential of improving the efficacy of drug discovery. These techniques need to be improved because of a few issues. Computational methods often generate false positives, requiring further in vitro or in vivo experimental verification. High throughput screening is expensive to set up and run, and requires much time and effort to generate hits with any potential.

Q. Li et al.

Fragment-based drug discovery assays are challenging because of the difficulty in detecting weak binding interaction at equilibrium. High performance affinity chromatography (HPAC) works well for analyzing compound mixtures, and can generate quantitative binding affinity estimates for a variety of proteins [10]. This provides the method with a great advantage in screening bioactive compounds from traditional medicines. This technique often utilizes the immobilized cell membranes containing functional proteins such as receptors and transporters as stationary phases for screening target compounds from complex media. Wainer and his colleagues [11–15] have constructed a series of novel immobilized artificial membrane-based receptor stationary phases using physical adsorption for screening receptor binding ligands. Other research [16] reports on adsorbed cell membrane on the surface of silica gel to establish chromatographic methods for screening bioactive compounds from traditional medicines. The main issue with these two types of assays is the stability of the stationary phase, which is synthesized by physical adsorption. This work was designed to improve the stability of cell membrane-based stationary phase by immobilizing the human umbilical vascular endothelial cells (HUVECs) on the surface of amino microspheres. The immobilization was performed through the specific affinity interaction between integrin on the cell membrane and arginine-glycine-aspartic acid (RGD) peptide-functionalized polymer brushes. The work also aimed to increase the density of the cells on the stationary phase by atom transfer radical polymerization (ATRP). Validated application of the proposed method in real samples was finally performed using Ligusticum chuanxiong Hort (L. chuanxiong) as a probe.

Experimental Instruments and materials The chromatographic system applied in this work was an Agilent Technologies 1100 Series of HPLC apparatus (Waldbronn, Germany), equipped with a vacuum degasser, a binary pump, a column thermostat, a UV detector, and a ChemStation 5.2 software installation for data acquisition and processing. The packing machine was from Dalian Yilite Analytic Instruments Company Ltd. (ZZXT-A; Dalian, China). Amino microspheres (30 μm, 100 Å) with amino group density of 2.0 mmol/g were purchased from Suzhou Knowledge and Benefit Sphere Tech. Co., Ltd. EGM-2 BulletKit (Endothelial Cell Growth Medium-2 BulletKit, CC-3162) was purchased from Lonza (San Francisco, CA, USA). Oligo (ethylene glycol) methacrylates [~360 g/mol (OEGMA6)] were obtained from Sigma-Aldrich (Shanghai, China).

Arg-Gly-Asp acid (RGD) peptide was from China Peptides Co., Ltd. VEGF (vascular endothelial growth factor, MDL number: MFCD00286542) was purchased from SigmaAldrich (Shanghai, China). Standards of ligustrazine hydrochloride (110817-200305), ferulic acid (11073-0611), and puerarin (110752-200209) were obtained from the National Institutes for Food and Drug Control (Beijing, China).

Cell culture HUVECs obtaining HUVECs (>70 % viability) preserved at the end of the primary culture were obtained from Thermo Fisher Scientific Inc. (Waltham, MA, USA) and was stored in cryogenic tubes (500, 000 viable cells/vial) under liquid nitrogen environment. According to the guidelines, the cells were positive for von Willebrand factor, CD 31, and dil-Ac-LDL uptake in the corresponding test. Mycoplasma, hepatitis B, hepatitis C, HIV-1, and fungi such as bacteria and yeast were not detected in the cells. HUVECs thawing One cryogenic tube containing 500,000 HUVECs was removed from liquid nitrogen and thawed immediately in a 37 °C water bath. The thawed HUVECs were suctioned into a 15.0 mL tube in the presence of 2.0 mL EGM-2 medium. This cell suspension was centrifuged for 5.0 min at a speed of 1000 rpm. The supernatant was discarded and the cell pellet was resuspended in 2.0 mL EGM-2. To break up cell clumps, the cell suspension was triturated by a yellow Gilson tip connected to P100 pipette. The cell suspension was then transferred into a 75-cm2 vented tissue flask (T-flask) with 10.0 mL EGM-2 medium. Subsequent incubation was performed at 37 °C in a humidified incubator under 5 % CO2 and 95 % air until 80 %–90 % confluence was reached (about 48 h). HUVECs passaging The cultured HUVECs were passaged when 80 %–90 % confluence was reached. The old EGM-2 medium in the flask was aspirated and any residual serum (that may interfere with the proteolytic action of trypsin) was removed by adding and aspirating 10.0 mL prewarmed PBS. Two mL trypsin-EDTA solution was pipetted into the flask prior to incubation at 37 °C for 3.0 min to allow trypsin to detach the monolayer of cells. The flask was tapped sharply to dislodge the remaining adherent cells. The resulting suspension was transferred to a 15 mL centrifuge tube and 6.0 mL D-PBS was added to remove any remaining cells from the flask. The pellet was centrifuged for 5 min at 1000 rpm, and then resuspended in 6.0 mL EGM-2

Screening bioactive compounds from Ligusticum chuanxiong

for further passage. This procedure was repeated until three population doublings were reached. Finally, the cells were detached by 0.1 % trypsin/0.02 % EDTA solution after assessing the growth and cellular phenotype within a time frame. The cells (1×105 cells/ T flask) were washed twice by PBS relevant to immediate use.

Immobilization of HUVECs Synthesis of POEGMA6 polymer brushes The poly OEGMA6 (POEGMA6) brushes were synthesized by surface-initiated atom-transfer radical polymerization(SIATRP) [17]. As depicted in Fig. 1, 1 g of amino microspheres (30 μm, 100 Å) was sonicated for 10 min in dry dichloromethane (30 mL). Using 4-dimethylaminopyridine (DMAP, 0.08 g) as a catalyst, a mixture of triethylamine (TEA, 1.25 mL) and 2-bromoisobutyryl bromide (2-BIB, 1 mL, 8 mmol) was added slowly to the amino microspheres. The resulting suspension was stirred with a magnetic stirrer in an ice bath. Two hours later, the temperature of the reaction system was changed to ambient temperature followed by an extra 12 h stirring. The acquired microspheres were rinsed thoroughly with dichloromethane, acetone, and methanol, and finally dried in a vacuum oven at 60 °C for 12 h. The dried microspheres were suspended in a solution consisting of POEGMA6 (17.5 mL, 63 mmol), the solvent (V(water):V(methanol) =1:4, 17.5 mL), and bipyridine (490 mg, 3.15 mmol), and degassed by two freeze-pump-thaw cycles. This suspension was further degassed after the addition of CuBr (180 mg, 1.26 mmol) and CuBr2 (14 mg, 0.063 mmol). The reaction was purged in nitrogen and thermo stated (60 °C) environment. Polymerization was stopped after 3 h by exposing the system to air. The POEGMA6-modified amino microspheres were rinsed with deionized water and methanol for the Fig. 1 Schematic showing the synthesis method for RGD peptide-functionalized polymer brushes coated amino microspheres

removal of unreacted polymer and copper catalyst. The cleaned microspheres were dried under a stream of nitrogen. POEGMA6-modified amino microspheres were activated by N,N-carbonyl diimidazole (CDI). In this case, the POEGMA6-modified amino microspheres were suspended in 50 mL dry dimethylsulfoxide (DMSO) containing 0.32 g CDI and 0.24 g 4-dimethylaminopyridine (DMAP). The suspension was stirred for 6 h at room temperature under nitrogen atmosphere. The resulting microspheres were thoroughly washed with dichloromethane and DMSO consecutively. The RGD peptide functionalization progressed by adding the activated microspheres to anhydrous dimethylformamide containing 1 mmol RGD peptide and 2.5 mM of DMAP. This reaction was allowed for 18 h in the dark at room temperature. The residual adsorbed peptide was removed by washing the obtained microspheres with DMF and PBS (20 mM, pH 7.4) three times. Two flasks of the harvested HUVECs were mixed with the RGD peptide functionalized microspheres and reacted for 30 min at 4 °C under gentle shaking. The HUVECs coated microspheres were rinsed with PBS until no cells were detected in PBS elution [18, 19]. The immobilized HUVECs were packed into a stainless steel column (2.1 mm×5.0 cm) using PBS (20 mM, pH 7.4) as slurry and replacing agent at a pressure of 4.0×107 Pa. The control column containing RGD peptide functionalized microspheres and exposed amino microspheres were prepared by the same method. Characterization of the immobilized HUVECs The exposed amino microspheres, the RGD peptide modified microspheres, and the immobilized HUVECs were dried and sprayed with a 5- to 10-nm layer of gold under vacuum. The surface morphologies of these mircospheres were examined under scanning electron microscopy (SEM). SEM analysis was performed by a Hitachi SU8010 SEM (Tokyo, Japan) with 2.0 kV primary beam energy.

Q. Li et al.

Sodium nitrite (10 μg/mL), VEGF (10 ng/mL), ligustrazine hydrochloride (tetramethylpyrazine, TMP, 220 μg/mL), and puerarin (5 μg/mL) were used to characterize the chromatographic behaviors of the columns. The injection volume was set at 5.0 μL. The mobile phase was phosphate buffer (50 mM, pH 7.4) and the flow rate was 0.2 mL/min. The detection wavelengths for the four compounds were 254, 280, 295, and 252 nm.

Preparation of L. chuanxiong extraction The extraction of L. chuanxiong was prepared by percolation method described in the Pharmacopoeia of People’s Republic of China (2005 edition). The dry roots of L. chuanxiong were collected from Tongrentang. Ten grams of the roots was powdered using an electric disintegrator. One gram of the powder was soaked with 100 mL ethanol-water (7:3, v:v) mixture for 24 h. Further extraction was performed using a percolation flow rate of 15 drops/min and a time of 8.0 h. The extract was filtered and the filtrate was vaporized to 100 mL using rotary evaporator. The resulting solution was filtrated by 0.45 polytetrafluoroethene membrane (size: 0.45 μm) for chromatographic analysis. A high performance liquid chromatographic (HPLC) method described in the Pharmacopoeia of People’s Republic of China was used to determine the reference standard (ferulic acid) in the extract.

Molecular docking The crystal structure of domain 2 of the Flt-1 Receptor (Flt1D2, Protein Data Bank code: 1FLT) at a resolution of 1.7 Å was obtained from the RCSB Protein Data Bank (PDB, http:// The pretreatment of the receptor structure was performed by Molecular Operating Environment (MOE), which included deletion of the water and addition of the hydrogen atoms. The three-dimensional structure of TMP was built by ChemDraw Ultra 8.0. The resulting geometry was read in AutoDock 4.2 software in compatible file format, from which the PDBQT style files were generated. Molecular docking was performed by AutoDock 4.2 suite of programs, which utilizes the Lamarckian genetic algorithm (LGA) [20]. The grid maps were constructed by a grid box size of 34 Å× 42 Å×38 Å points and a grid point space of 0.375 Å. The center of the grid box was set at the point with x, y, and z coordinates of 23.500, 0.501, and 6.403. The docking parameters included a LGA population size of 150, a maximum number of energy evaluations of 250,000, and a LGA crossover mode of two points. The lowest binding energy conformer was searched out from 50 different conformations for each docking simulation for further analysis. The PYMOL and MOE software package were used for visualizing the docking conformations [21].

Results and discussion Selection of L. chuanxiong

Screening the bioactive compounds in L. chuanxiong The bioactive compounds in L. chuanxiong were screened by the immobilized HUVECs column using HPAC. The mobile phase was PBS (50 mM, pH 7.4) and the flow rate was set at 0.2 mL/min. The detection wavelength was 295 nm and the oven temperature was 25 °C. The retention peak on the column containing immobilized HUVECs was collected for further separation by an Inertsil/ Wondasil C18 column (5 μm, 150 mm×4.6 mm i.d.) and identification by electrospray ionization-quadrupole time of flightmass spectrometry (ESI-QTOF-MS). The mobile phase was a solution of methanol and 0.2 % formic acid water (60:40, v:v) with a flow rate of 0.6 mL/min. Mass spectrum was recorded in the positive ionization mode over a mass range of m/z 20– 1000 with a scan rate of one spectra per s. The capillary voltage was 4000 V using nitrogen as nebulizer with a pressure of 35 psi. The flow rate and temperature of drying gas (nitrogen) were 8.0 L/min and 350 °C. The collision energies for the MS/ MS mode were set at 5, 15, 25, and 40 V for all the detection. The MS and MS/MS data were processed by Masshunter Workstation software (ver. B.04.00; Agilent Technologies, Waldbronn, Germany).

The rhizome of L. chuanxiong Hort. (LC) belongs to Umbelliferae family. It is widely used to cure various cardiovascular and cerebrovascular diseases [22] in practice. The herb consists of many kinds of chemical constituents, including TMP, volatile oil, and ferulic acid. A survey of literatures shows that TMP [23, 24] and some lactones [25] in the extract of L. chuanxiong have effects of anti-atherosclerosis and antiangiogenesis. The major pathologic mechanisms of the two ailments involve capillary endothelial cell proliferation, migration, and tissue infiltration [26, 27]. HUVECs were the classic cell model to demonstrate the effect of drugs on cell proliferation, migration, and tube formation [28]. In this point, L. chuanxiong was expected to contain bioactive components specifically binding to HUVECs. Immobilization of HUVECs In previous reports, many molecules such as RGD peptide [29, 30], epidermal growth factor (EGF) [31] and VEGF [32] have been utilized to synthesize bio-functional supports. Successful preparation of these supports ascribes to the signaling domains in the molecules, which can specifically bind

Screening bioactive compounds from Ligusticum chuanxiong

to integrin on cell membranes [33]. Various methods consisting of covalent bonding, layer-by-layer self-assembly, plasma treatment, photo-initiated grafting, and γ-irradiationinduced grafting have been utilized to immobilize the biomolecules onto the solid matrix [34, 35]. These strategies have several limitations, including low grafting density because of steric hindrance, uncontrollable grafting yield of polymer brushes, and undesired reactions between reactive groups of the brushes and the surface [36]. ATRP method is one of the recently published strategies for attaching bio-molecules on the surface of solid matrix. It enables precision control over the bio-molecular structure, order, composition, and functionality during the immobilization. The surface-initiated ATRP method provides an alternative for the preparation of well-defined polymer brushes containing pendent reactive groups for post-polymerization of bio-molecules [37]. Inspired by the advantages of abovementioned techniques, we combined the specificity of RGD peptide binding to integrin with the density-controllable property of ATRP method to propose a new strategy for cell immobilization. Brush of POEGMA6 with the layer thickness of 100 nm was prepared according to the literature [17] where CuBr/CuBr2/bipy was used as the catalyst in water/methanol at 60 °C with a polymerization time of 3 h. The surface concentration of peptides were 12 pmol/cm 2 for 100 nm PPEGMA 10 brush, 14 pmol/cm2 for 20 nm PPEGMA10 brush, and 18 pmol/cm2 for 20 nm PPEGMA6 brush. Based on these results, the surface concentration of peptide for 100 nm PPEGMA6 brush was calculated to be 15.4 pmol/cm2. In this work, the specific surface area of the amino microsphere was 700 m2/g, and 1.0 mg RGD peptide was utilized in the immobilization.

Assuming a complete peptide attachment, the surface concentration of the peptide was estimated to be 143 pmol/cm2. This concentration was far greater than the suggested threshold of 1.0–5.3 pmol/cm2 by Tugulu et al [17]. It indicated that HUVECs were capable of effectively adhering and spreading well on the surface of the brush. On the other aspect, even when 2×105 HUVECs were utilized to react with RGD peptide modified microspheres, no residual cells were detected in the filtrate. This result showed that all the HUVECs specifically bond to RGD peptide on the microsphere surface. The number of HUVECs was far lower than the attached RGD peptide (calculated to be 9.89×1013 by 143 pmol/cm2 and 700 m2/g of the microsphere surface), which guaranteed the entire immobilization of the cells. Surface morphology The SEM images of naked amino microspheres, the RGD peptide functionalized microspheres, and the immobilized HUVECs are presented in Fig. 2. The naked amino microspheres showed uniform smooth particles with an average diameter of 30 μm. After being sequentially modified by POEGMA and RGD peptides, the surface of the particles became rough, wrinkled, and non-uniform. Compared with the morphology of RGD peptide modified microspheres, the surface of immobilized HUVECs was altered greatly and the shriveled ball bumps were covered by HUVECs. The SEM images of amino microspheres, RGD peptide modified microspheres, and immobilized HUVECs at higher magnifications (Fig. 2d, e, f) indicated that HUVECs were successfully immobilized on amino microspheres.

Fig. 2 The SEM image of amino microspheres (a, d), RGD peptide midified microspheres (b, e), and HUVECs coated microspheres (c, f)

Q. Li et al.

Characterization of immobilized HUVECs The chromatograms of sodium nitrite, VEGF, TMP, and puerarin on immobilized HUVECs are shown in Fig. 3. Sodium nitrite gave no retention on the column, whereas VEGF, TMP, and puerarin presented good retentions with retention times of 9.6, 13.2, and 4.6 min, respectively. In addition, the four compounds showed no retention on the control columns containing RGD peptide modified microspheres and naked amino microspheres. These results indicated that the HUVECs column had the specificity and bioactivity for recognizing specific ligands of the receptors on the cell membrane. The stability of the column containing immobilized HUVECs was studied by detecting the retention times of VEGF, TMP, and puerarin for 14 consecutive d. The relative standard deviations (RSD) of the retention times during the 2 wk were 1.5 %, 2.2 %, and 2.0 %. The peak profiles of the four compounds displayed little difference during the testing time. These results showed that immobilized HUVECs had a good stability in at least 2 wk.

The chromatogram of the extract of L. chuanxiong on HUVECs column is displayed in Fig. 4. An obvious peak with retention time of 13.4 min was found and annotated as peak III. This peak was believed to contain bioactive compounds specifically binding to HUVECs attributed to the much longer retention time than the void time (determined by sodium nitrite) of the chromatographic system. Further separation of peak III was performed by HPLC-ESI-QTOF-MS. The corresponding total ion current is shown in Fig. 5. It was found that peak III on HUVECs column was separated into three peaks. Peak 1 showed an intensive ion of m/z 226.952 [M+H]+, which gave a MS/MS pattern of m/z 158.964 and 90.977. Peak 2 presented a father ion of m/z 196.962 [M+H]+ and daughter ions of m/z 98.985, 80.9743, and 62.964. Peak 3 showed an intensive ion of m/z 137.107 [M+H]+, which generated no daughter ions even though we used four collision energies in fragmentation experiments. In this experiment, the errors for all the ions were less than 1.4 ppm. With reference to the mass spectrometric data in previous reports [38], peaks 2 and 3 were identified as 3-butyl-hexahydroisobenzofuran1(3H)-one and TMP, whereas the structure of peak 1 was indistinct to the scope of our knowledge.

Screening the bioactive compounds in L. chuanxiong Comparison with other HPAC screening methods As a reference standard for quality control of L. chuanxiong, the content of ferulic acid was recommended to be higher than 0.1 % of the weight of the medical material according to the Pharmacopoeia of People’s Republic of China. In this work, the concentration of ferulic acid in the extract of L. chuanxiong was determined to be (2.85±0.47) mg/mL. This concentration equals to a content of (0.3 %±0.05 %) per gram medical material, which indicated a high quality of L. chuanxiong in this work and confirmed the validation of percolation method for preparing the extract of the herb.

HPAC has received a keen interest as one of the means for screening bioactive compounds from complex media. The interactions that occur in HPAC are similar to those taking place in many biological systems, including lectin/sugar, enzyme/inhibitor, protein/protein, DNA/protein, and drug/ protein [39]. This property, in turn, requires one of a pair of two interacting agents attached to a support and used as the stationary phase. In most cases, proteins, nucleotides, cells, drugs, or other biologically related molecules are used as

Fig. 3 Superimposed chromatograms of sodium nitrite, VEGF, TMP, and puerarin on the HUVECs column (2.1 mm×5.0 cm). The mobile phase was PBS (20 mM, pH 7.4), and the flow rate was 0.2 mL/min. All the experiments were performed at 25 °C

Fig. 4 Chromatogram of extract of Chuanxiong on HUVECs column (2.1 mm×5.0 cm). The mobile phase was 20 mM PBS with pH value of 7.4, and the flow rate was set at 0.2 mL/min. All the experiments were performed at 25 °C

Screening bioactive compounds from Ligusticum chuanxiong Fig. 5 HPLC-Q/TOF-MS/MS chromatogram of the detected compounds. The retention peaks 2 and 3 were identified as 3-butylhexahydroisobenzofuran-1(3H)one and TMP, respectively

affinity ligands. Many researches have focused on immobilization of transporter proteins with varying functions such as complex system screening, drug–protein interaction, and chiral separation [40, 41]. The major issue of these researches is the relative lack of specificity. Receptors, the main target for drugs, are also immobilized on silica gel to construct more specific receptor-based stationary phase for screening bioactive compounds from traditional medicines [42–44]. Although the specificity improved greatly, the receptor-based stationary phase has limitations because it is difficult to maintain the natural conformation of the receptor without a bilayer membrane system. Wainer et al [16, 45] and He and colleagues [47] have adsorbed cell membrane on silica gel and developed a series of cell membrane supports; however, the adsorbed cell membrane inevitably disassociates from the stationary phase after a long period of chromatographic analysis. In previous reports, the stability of cell membrane base stationary phase prepared by Wainer’s method was a 10-d period for continuous use [46]. The stationary phase synthesized by He’s approach was only stable for a 48-h period when routine analysis was performed [47]. In this work, HUVECs were immobilized on the surface of amino microspheres through high affinity between RGD peptide and integrin. Compared with physical adsorption method, this high affinity resulted in a relatively long stability of 14-d period for intensive sampling. Compared with receptor-based stationary phase, the method in this work has the capacity to retain the native conformation of the receptors because of the presence of cell membrane. Moreover, it is effortless to remove the immobilized cells from RGD peptide modified matrix by flushing the stationary phase with RGD peptide solution. This enables the reversible utilization of RGD peptide modified matrix for immobilization of other cells. It is concluded that the proposed method has the advantage of high specificity, good stability, and broad application for any cells. Binding of VEGF and TMP to VEGFR TMP is the main bioactive alkaloid in L. chuanxiong. It has multiple pharmacologic activities such as antioxidant, anti-inflammatory, and anti-cancer [25]. Previous researches [44]

show that TMP has obvious effects on the binding affinity between VEGF and VEGFR. In this work, TMP presented good retention on the column containing immobilized HUVECs. This indicated that TMP was capable of binding to the effectors on the cell membrane. According to the pharmacologic action of TMP and VEGF, we hypothesized that both TMP and VEGF bound to the same target protein (i.e., VEGFR). To confirm our hypothesis, zonal elution was carried out to achieve the binding behaviors of VEGF and TMP to VEGFR. In this theory, a known series of competing agent (I) concentrations continuously pass through an immobilized ligand (L). The injecting volume of the analyte (A) is negligible compared with the concentrations of I in the mobile phase. Assuming a direct competition between I and A at a single binding site on L and a disregard of association/dissociation kinetics, Eq. 1 can be used to describe the binding of A and I to L [48], 1 k 0 ‐X


V m K I ½I  Vm þ K A mL K A mL


where k' is the overall capacity factors of A ascribed to all the types of sites, Vm is the volume of the mobile phase in the chromatographic column, and X is the capacity factor for A attributable to the existence of a second type of binding site on the column. KA is the association equilibrium constant for the binding of A to L, and KI is the association equilibrium constant for the interaction of I at the same site. Zonal elution was performed where VEGF served as injection analyte and TMP was used as competing agent. The injection volume of VEGF was 20 μL of 10 ng/mL solution. The concentrations of TMP in the mobile phase were 6.5, 7.5, 9.0, 10.0, 12.0, 14.0, 16.0, and 18.0 nM. Triplicate injections were performed for each concentration of TMP in the mobile phase. All the experiments were performed at 25 °C. The capacity factors of VEGF on HUVECs column are summarized in Table 1. A capacity factor decrease was found when TMP mobile phase concentration increased. The adsorption properties of VEGF were in line with Langmuir model since the column was not saturated by VEGF during the testing concentrations. According to Eq. 1, plotting the curve of 1/(k′-X) against the concentrations of TMP in the mobile

Q. Li et al. Table 1 The capacity factors of VEGF on HUVECs column in the presence of series concentrations of TMP in mobile phase at 25 °C

Concentrations of TMP in mobile phase (×10−9 M)

Capacity factors (k′)

6.5 7.5 9.0 10.0 12.0

1.115±0.050 0.949±0.011 0.846±0.046 0.742±0.019 0.642±0.027

14.0 16.0 18.0

0.571±0.032 0.514±0.036 0.436±0.008

phase, a good linear relationship was obtained with a regression equation of y=(1.232±0.066)×108x+(0.125±0.009) and a correlation coefficient of 0.9932. The value of X was calculated to be 0.021±0.003 by nonlinear regression. Dividing the intercept by the slope of the regression equation, we calculated the association constant of TMP of (9.84±1.11)× 108 M−1. It was reported that VEGFR distributed in endothelial cells were 4 × 10 4 receptors/cell [49]. The total numbers of HUVECs immobilized on the column were considered to be 2×105. With the use of column volume (1.73×10−4 L), the concentrations of binding site on the column were calculated to be 7.69×10−11 M. Substituting this value into the regression equation of y=(1.232±0.066)×108x+(0.125±0.009), the association constant of VEGF binding to HUVECs in the column was (1.04±0.08)×1011 M−1. von Tiedemann et al [50] have determined the association and dissociation rate constants of VEGF/VEGFR and gave ka of (4±1.2)×106 M−1/ S−1 and kd of (3±0.8)×105 S−1 by optical affinity sensor method. This means a KA value of (1.20±0.96)×1012 M−1. Feng et al [44] have investigated the binding of VEGF to VEGFR in the presence of TMP by radio-ligand binding assay and showed that the dissociation constant ranged from (179.07± 25.65)×10−12 M to (343.3±36.64)×10−12 M. Our data was in good agreement with these reported results, which indicated that immobilized HUVECs based HPAC method was a valid

To reveal the reliability of zonal elution, molecular docking was used to examine the binding of TMP to VEGFR. The biological function of VEGF depended on its binding to two receptors. The two receptors were localized on the cell surface of various endothelial cell types [51] and confirmed to be FMS-like tyrosine kinase (Flt-1, VEGFR-1) and the kinase domain receptor (KDR; murine flk-1, VEGFR-2). Flt-1 binds to VEGF in pM level and has a 10-fold higher affinity for VEGF than KDR [52]. Wiesmann and colleagues [53] have reported that domain 2 of Flt-1 (Flt-1D2) is able to represent the receptor for examining the binding of VEGF to Flt-1. In this work, Flt-1D2 was chosen as the receptor to achieve the docking for exploring the binding of TMP to VEGFR. The docking conformation (Fig. 6a) with minimized binding energy (−2.41 kJ/mol) showed that one hydrogen bond was formed between N atom of TMP and 3-N atom of imidazole group in histidine223. The 2D structure of TMP-Flt-1D2 complex (Fig. 6b) showed that TMP was surrounded by the amino acids of Ile202, Gly203, Leu221, Thr222, His223, and Arg224. These amino acids were reported to have the ability to form the binding interface of VEGF to Flt-1D2. It indicated

Fig. 6 Details of the complex between domain 2 of the FMS-like tyrosine kinase domain receptor (Flt-1D2) and TMP. (a) 2D overview of the complex between Flt-1D2 and TMP, (b) 3D overview of the complex

between Flt-1D2 and TMP.Flt-1D2 is displayed in solid ribbon, TMP and the amino acid residues of Flt-1D2 participated in hydrogen bonds are rendered with sticks and the hydrogen bonds are shown in red dash lines

assay for determining the association constant of TMP to VEGFR. Molecular docking

Screening bioactive compounds from Ligusticum chuanxiong

that both TMP and VEGF bind to Flt-1D2 through the same binding site of VEGFR.



Conclusions A universal strategy for the synthesis of immobilized cellsbased stationary phase was proposed using HUVECs as a probe. Validated application of immobilized HUVECs in the analysis of L. chuanxiong extract showed that TMP was the bioactive compound that specifically binds to VEGFR. Compared with previous cell membrane-derived stationary phase, this strategy enables the affinity supports to have better stability. Compared with immobilized receptor based stationary phase, this strategy is able to maintain the native conformation of the receptors on the cell membrane. Through on-line combination with mass spectrometry and classic affinity techniques such as zonal elution and frontal analysis, it is possible to simultaneously screen and identify bioactive compounds in complex media, and probe the exact binding sites of the resulting compounds to their specific protein on the cell membrane as well. Acknowledgments The authors are grateful for the financial support by a grant of the National Natural Science Foundation of China (Nos. 21075097, 21475103), the program for Innovative Research Team of Shaanxi Province (No. 2013KCT-24), and the Ministry of Science and Technology of the People’s Republic of China (No. 2013YQ170525; subproject: 2013YQ17052509).

References Zhao XF, Zheng XH, Fan TP, Li ZJ, Zhang YY, Zheng JB (2015) A novel drug discovery strategy inspired by traditional medicine philosophies. Science 347(6219 Suppl):S38–S40 2. Itokawa H, Morris-Natschke SL, Akiyama T, Lee KH (2008) Plantderived natural product research aimed at new drug discovery. J Nat Med 62:263–280 3. Harvey AL (2008) Natural products in drug discovery. Drug Discov Today 13:894–901 4. Brusotti G, Cesari I, Dentamaro A, Caccialanza G, Massolini G (2014) Isolation and characterization of bioactive compounds from plant resources: the role of analysis in the ethnopharmacological approach. J Pharm Biomed Anal 87:218–228 5. Mishra KP, Ganju L, Sairam M, Banerjee PK, Sawhney RC (2008) A review of high throughput technology for the screening of natural products. Biomed Pharmacother 62:94–98 6. Huang XD, Kong L, Li X, Chen XG, Guo M, Zou HF (2004) Strategy for analysis and screening of bioactive compounds in traditional Chinese medicines. J Chromatogr B 812:71–84 7. Kapetanovic IM (2008) Computer-aided drug discovery and development (CADDD): In silico-chemico-biological approach. Chem Biol Interact 171:165–176 8. Proudfoot JR. (2008) High-Throughput Screening and Drug Discovery. The Practice of Medicinal Chemistry, 3rd ed., Chap 7, Academic Press/Elsevier, Amsterdam, pp. 144–158 9. Murray CW, Verdonk ML, Rees DC (2012) Experiences in fragment-based drug discovery. Trends Pharmacol Sci 33:224–232

















Schiel JE, Mallik R, Soman S, Joseph KS, Hage DS (2006) Applications of silica supports in affinity chromatography. J Sep Sci 29:719–737 Beigi F, Chakir K, Xiao RP, Wainer IW (2004) G-protein-coupled receptor chromatographic stationary phases. 2. Ligand-induced conformational mobility in an immobilized β2-adrenergic receptor. Anal Chem 76:7187–7193 Moaddel R, Jozwiak K, Yamaguchi R, Wainer IW (2005) Direct chromatographic determination of dissociation rate constants of ligand–receptor complexes: assessment of the interaction of noncompetitive inhibitors with an immobilized nicotinic acetylcholine receptor-based liquid chromatography stationary phase. Anal Chem 77:5421–5426 Jozwiak K, Haginaka J, Moaddel R, Wainer IW (2002) Displacement and nonlinear chromatographic techniques in the investigation of interaction of noncompetitive inhibitors with an immobilized α3β4 nicotinic acetylcholine receptor liquid chromatographic stationary phase. Anal Chem 74:4618–4624 Moaddel R, Jozwiak K, Whittington K, Wainer IW (2005) Conformational mobility of immobilized α3β2, α3β4, α4β2, and α4β4 nicotinic acetylcholine receptors. Anal Chem 77:895–901 Beigi F, Wainer IW (2003) Syntheses of immobilized G proteincoupled receptor chromatographic stationary phases: characterization of immobilized μ and κ opioid receptors. Anal Chem 75:4480– 4485 Moaddel R, Oliveira RV, Kimura T, Hyppolite P, Juhaszova M, Xiao Y, Kellar KJ, Bernier M, Wainer IW (2008) Initial synthesis and characterization of an α7 nicotinic receptor cellular membrane affinity chromatography column: effect of receptor subtype and cell type. Anal Chem 80:48–54 Tugulu S, Arnold A, Sielaff I, Johnsson K (2005) Harm-Anton Klok. Protein-functionalized polymer brushes. Biomacromolecules 6:1602–1607 Tugulu S, Silacci P, Stergiopulos N, Klok HA (2007) RGDfunctionalized polymer brushes as substrates for the integrin specific adhesion of human umbilical vein endothelial cells. Biomaterials 28:2536–2546 Yang Z, Yuan S, Liang B, Liu Y, Choong C, Pehkonen S (2014) Chitosan microsphere scaffold tethered with RGD-conjugated poly (methacrylic acid) brushes as effective carriers for the endothelial cells. Macromol Biosci 14:1299–1311 Tiwari R, Mahasenan K, Pavlovicz R, Li C, Tjarks W (2009) Carborane clusters in computational drug design: a comparative docking evaluation using AutoDock, FlexX, Glide, and Surflex. J Chem Inf Model 49:1581–1589 Delano WL (2004) The pyMol molecular graphics system Delano Scientific, San Carlos, CA, USA. Available at: http://www. Ran X, Ma L, Peng C, Zhang H, Qin LP (2011) Ligusticum chuanxiong Hort: a review of chemistry and pharmacology. Pharm Biol 49:1180–1189 Hou XF, Wang S, Zhang T, Ma J, Zhang J, Zhang YM, Lu W, He HZ, He LC (2014) Recent advances in cell membrane chromatography for traditional Chinese medicine analysis. J Pharm Biomed 101:141–150 Li SW, Ling Y, Jin S, Lin YF, Chen ZJ, Hu CX, Wang MH, Yao MZ (2014) Expression of soluble vascular endothelial growth factor receptor-1 and placental growth factor in fetal growth restriction cases and intervention effect of tetramethylpyrazine. Asian Pac J Trop Med 2014:663–667 Xiao Y, Wang YC, Li LL, Jin YC, Sironi LG, Wang Y (2014) Lactones from Ligusticum chuanxiong Hort reduces atherosclerotic lesions in apoE-deficient mice via inhibiting over expression of NFkB-dependent adhesion molecules. Fitoterapia 95:240–246 Roskoski R (2007) Vascular endothelial growth factor (VEGF) signaling in tumor progression. Crit Rev Oncol Hematol 62:179–213

Q. Li et al. 27.









36. 37.

38. 39. 40.


Fan TP, Yeh JC, Leung KW, Yue PY, Wong RN (2006) Angiogenesis: from plants to blood vessels. Trends Pharmacol Sci 27:297–309 Tugues S, Koch S, Gualandi L, Li X, Claesson-Wels L (2011) Vascular endothelial growth factors and receptors: anti-angiogenic therapy in the treatment of cancer. Mol Asp Med 32:88–111 Gumusderelioglu M, Turkoglu H (2002) Biomodification of nonwoven polyester fabrics by insulin and RGD for use in serum-free cultivation of tissue cells. Biomaterials 23:3927–3935 Ebara M, Yamato M, Aoyagi T, Kikuchi A, Sakai K, Okano T (2004) Immobilization of cell adhesive peptides to temperatureresponsive surfaces facilitates both serum-free cell adhesion and noninvasive cell harvest. Tissue Eng A 10(7/8):1125–1135 Elloumi I, Kobayashi R, Funabashi H, Mie M, Kobatake E (2006) Construction of epidermal growth factor fusion protein with cell adhesive activity. Biomaterials 27:3451–3458 Taguchi T, Kishida A, Akashi M, Maruyama I (2000) Immobilization of human vascular endothelial growth factor (VEGF165) onto biomaterials: an evaluation of the biological activity of immobilized VEGF165. J Bioact Compat Polym 15:309– 320 Li L, Wu J, Gao C (2011) Gradient immobilization of a cell adhesion RGD peptide on thermal responsive surface for regulating cell adhesion and detachment. Colloid Surf B 85:12–18 Luk JZ, Cooper-White J, Rintoul L, Taran E, Grøndahl L (2013) Functionalized polycaprolactone films and 3D scaffolds via gamma irradiation-induced grafting. J Mater Chem B 1:171–4181 Cai H, Zhang ZP, Sun PC, He BL, Zhu XX (2005) Synthesis and characterization of thermo- and pH-sensitive hydrogels based on Chitosan-grafted N-isopropylacrylamide via γ-radiation. Radiat Phys Chem 74:26–30 Edmondson S, Osborne VL, Huck WTS (2004) Polymer brushes via surface-initiated polymerizations. Chem Soc Rev 33:14–22 Barbey R, Lavanant L, Paripovic D, Schüwer N, Sugnaux C, Tugulu S, Klok HA (2009) Polymer brushes via surface-initiated controlled radical polymerization: synthesis, characterization, properties, and applications. Chem Rev 109:5437–5527 Yin J, Guo LG (1993) Modern medicine research and clinical applications (1). 1st ed. Beijing, pp. 108–117 Chaiken IM (1987) Analytical affinity chromatography. CRC Press, Boca Raton Mallik R, Wa C, Hage DS (2007) Development of sulfhydrylreactive silica for protein immobilization in high-performance affinity chromatography. Anal Chem 79:1411–1424 Chen J, Ohnmacht C, Hage DS (2004) Studies of phenytoin binding to human serum albumin by high-performance affinity chromatography. J Chromatogr B 809:137–145


Zhao XF, Nan YF, Xiao CN, Zheng JB, Zheng XH, Wei YM, Zhang YY (2010) Screening the bioactive compounds in aqueous extract of Coptidisrhizoma, which specifically bind to rabbit lung tissues β2-adrenoceptor using an affinity chromatographic selection method. J Chromatogr B 878:2029–2034 43. Marszałł MP, Moaddel R, Kole S, Gandhari M, Bernier M, Wainer IW (2008) Ligand and protein fishing with heat shock protein 90 coated magnetic beads. Anal Chem 80:7571–7575 44. Feng J, Xu X, Hu Y, Chen G, Chen W, Yang L (2005) Tetramethylpyrazine inhibition on binding of radiolabeled ligand to VEGFR. Chin Pharmacol Bull 21:939–942 45. Zhang CC, Chen G, Liu JL, Li Y, Xu XY (2013) Cell membrane chromatography and its application progress in studies of traditional Chinese medicines. Chin J Chin Mater Med 38:1881–1886 46. Zhang YX, Xiao YX, Kellar KJ, Wainer IW (1998) Immobilized nicotinic receptor stationary phase for on-line liquid chromatographic determination of drug-receptor affinities. Anal Biochem 264:22–25 47. Wang Y, Yuang BX, Deng XL, He LC, Zhang YY, Han QD (2005) The preparation of HEK293 α1A or HEK293 α1B cell membrane stationary phase and the chromatographic affinity study of ligands ofα1 adrenoceptor. Anal Biochem 339:198–205 48. Zhao XF, Zheng XH, Wei YM, Bian LJ, Wang SX, Zheng JB, Zhang YY, Li ZJ, Zang WJ (2009) Thermodynamic study of the interaction between terbutaline and salbutamol with an immobilized β2-adrenoceptor by high-performance liquid chromatography. J Chromatogr B 877:911–918 49. Vaismanz N, Gospodarowicz D, Neufeld G (1990) Characterization of the receptors for vascular endothelial growth factor. J Biol Chem 265:19461–19466 50. von Tiedemann B, Bilitewski U (2002) Characterization of the vascular endothelial growth factor/receptor interaction and determination of the recombinant protein by an optical receptor sensor. Biosens Bioelectron 17:983–991 51. Vaisman N, Gospodarowicz D, Neufeld G (1990) Characterization of the receptors for vascular endothelial growth factor. J Biol Chem 265:19461–19466 52. Davis-Smyth T, Chen H, Park J, Presta LG, Ferrara N (1996) The second immunoglobulin-like domain of the VEGF tyrosine kinase receptor Flt-1 determines ligand binding and may initiate a signal transduction cascade. EMBO J 15:4919–4927 53. Wiesmann C, Fuh G, Christinger HW, Eigenbrot C, Wells JA, de Vos AM (1997) Crystal structure at 1.7 Å resolution of VEGF in complex with domain 2 of the Flt-1 receptor. Cell 91:695–704

Screening bioactive compounds from Ligusticum chuanxiong by high density immobilized human umbilical vein endothelial cells.

High throughput screening methodologies play a very important role in screening bioactive compounds from complex media. In this work, a new strategy f...
1MB Sizes 2 Downloads 10 Views