In situ construction of an SPR biosensor using cell-free protein synthesis technology and its application in drug screening for CXCR4

Abstract

Surface plasmon resonance (SPR) biosensor has emerged as a transformative tool in high-throughput drug screening and label-free analysis of biomolecular interactions. However, a critical limitation of SPR lies in its stringent requirement for highly purified proteins to ensure reliable quantification of binding affinities and kinetic parameters. In order to address the limitations, the lentiviral particle (LVP) and styrene-maleic acid (SMA) polymer have been previously developed to extract and stabilize transmembrane proteins (TMs) indirectly and thus to detect ligands interaction with TMs by SPR biosensor. The present study proposes a high-throughput SPR-based drug screening system that utilizes cell-free protein synthesis (CFPS) to achieve in situ purification and immobilization of TMs on SPR biosensors. First, C−X−C chemokine receptor 4 (CXCR4) protein with histidine (His)-tag was prepared by CFPS. Then, two types of nickel-nitrilotriacetic acid (Ni-NTA) biosensors, modified with carboxymethylated dextran (CMD) and coated with polycarboxylate hydrogel coating (HC) matrix, were compared with classical carboxymethylated dextran 5 (CM5) biosensor in order to determine the optimal strategy for coupling the CXCR4-CFPS protein. The CMD-NTA/SPR biosensor was next applied to screen for CXCR4 ligands from 96 natural products. Finally, glycyrrhizic acid and ginsenoside Re were proved to be function as CXCR4 inhibitors by affinity test, molecular docking, and cell migration assay. The combination of CFPS with SPR technologies facilitates in-situ purification and immobilization of target proteins in a single step, thereby significantly enhancing the efficiency of SPR assay procedures. The system has broad applicability for targeting various challenging TMs and provides potential candidates for subsequent drug development.

Graphical abstract

Highlights

• •CFPS technology was utilized to in situ construct SPR biosensors targeting transmembrane proteins.
• •The NTA sensor has been proven to be capable of purification and immobilization of CFPS proteins in one step.
• •Glycyrrhizic acid and ginsenoside Re were screened out to be potential inhibitors of CXCR4.

1. Introduction

Surface plasmon resonance (SPR) is a biosensing technology for the real-time detection of interactions between biomolecules, and has the advantages of label free, simple operation, fast response, good selectivity, and high sensitivity [1]. SPR can be used to detect parameters such as biomolecular binding kinetics and binding affinity [2,3], and is also used for high-throughput drug screening [4]. Drozd et al. [5] used SPR technology by immobilizing membrane protein insulin receptor (IR) on the biosensor as a target, thereby developing a platform capable of using kinetic data to assess the interaction characteristics of IR with drugs under study. The platform has the potential to facilitate extensive screening for the discovery of new antidiabetic drugs. Li et al. [6] developed and validated a rapid, high-throughput method based on SPR biosensor for the screening of potential xanthine oxidase (XOD) inhibitors. They identified 14 potential XOD inhibitors from Chrysanthemum morifolium Ramat, including chrysoeriol, acacetin, and luteolin, illustrated the viability of the SPR technique for drug screening, and demonstrated the extensive range of SPR applications.

In general, the need of purified proteins is a prior request in SPR experiments according to the principle of measurement. Nevertheless, transmembrane proteins (TMs) have hydrophobic and lipophilic transmembrane domains, and the production of active TMs in vitro is a recognized technical challenge [7], thus it hinders the using of TMs in SPR analysis. A number of techniques [[8], [9], [10]] have been developed to overcome the problems associated with purifying proteins using TMs in SPR. We have previously developed a novel P-glycoprotein (P-gp), which is a complex TM, ligand screening system based on SPR biosensor with a lentiviral particle (LVP) stabilization strategy to screen potential P-gp inhibitors from natural products [8]. In addition, a high-throughput SPR-based screening system for breast cancer resistance protein inhibitors based on a styrene-maleic acid (SMA) polymer membrane protein stabilization strategy was also established [9]. However, both strategies of immobilizing TMs on SPR biosensors still face many difficulties. In LVP strategy, it is likely that non-target proteins on the surface of the LVP will lead to non-specific response, and control or negative sensors are needed to deduct the non-specific binding. The SMA may alter the behavior of proteins and affect their binding to ligands. Furthermore, the polymer can obscure the active sites of proteins, thereby preventing them from interacting with target molecules. Moreover, the process of using LVP or SMA for SPR-based drug screening is still difficult and complex. To further address these issues, cell-free protein synthesis (CFPS) technology will introduce to SPR in this study to establish a new approach to capture TMs on the SPR biosensor.

CFPS technology is an in vitro approach for protein synthesis utilizing lysed cell extracts or reconstructed biochemical systems rather than expressing proteins in living cells, and is suitable for the expression of various types of proteins, including proteins that are difficult to express, complex proteins, and toxic proteins. It eliminates the dependence on living cells or bacteria, simplifies the purification process [[11], [12], [13]], and is straightforward to be modified with the addition of various tags. DeWinter et al. [14] achieved bedside synthesis of a group of peptide hormones using CFPS technology, thus solving the transport problem of biopharmaceuticals. They synthesized growth hormone and characterized it using liquid chromatography-mass spectrometry (LC-MS), enzyme linked immunosorbent assay (ELISA), amplified luminescent proximity homogeneous assay, and bio-layer interferometry techniques, which proved the stability and feasibility of protein synthesis using CFPS technology. Machida et al. [15] constructed a high-throughput screening system by using the CFPS technique to express the target protein for compounds that inhibit the programmed-1 ribosomal frameshifting (-1PRF) in severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Consequently, they identified a potential inhibitor of the -1PRF from 32,000 compounds. These studies illustrate the viability and promising future of CFPS technology in the protein expression and drug screening applications. Furthermore, the protein tags, such as histidine (His), Flag, glutathione S-transferase (GST), and so on, on the CFPS synthesized protein can facilitate the protein purification and immobilization on the SPR sensor surface in one step. The combination of CFPS and SPR technologies avoids the necessity for conventional protein purification, enhances protein stability and functionality, and improves the efficiency of SPR procedures. Therefore, for rapid and high throughput drug screening targeting TMs by SPR technology, using CFPS to produce interest proteins may be a desirable choice.

C−X−C chemokine receptor 4 (CXCR4) is a TM that has been proved to be a pharmacological target for many drugs [16]. But due to its seven times transmembrane structure, the purification of CXCR4 is still challenging [17]. In this study, CXCR4 was used as an example target protein to in situ establish a CFPS/SPR-based drug screening system for the TM-targeted drugs. The characteristic of this screening system lies in the direct synthesis of His-tagged membrane proteins by CFPS, and the protein purification and immobilization of target proteins could be fulfilled in one step by the specific binding of His and nickel-nitrilotriacetic acid (Ni-NTA) on the SPR sensor. The applicability and efficacy of this screening system was demonstrated by the discovery of several potential CXCR4 inhibitors from a natural product library, thereby substantiating the feasibility of integration SPR and CFPS technology. By in situ construction of the SPR biosensor, this study also provides a valuable reference and guidance for the rapid screening of inhibitors against various TMs.

2. Materials and methods

2.1. Reagents

The TnT® Quick Coupled Transcription/Translation System (L1170) was procured from Promega Corporation (Madison, WI, USA). Series S carboxymethylated dextran 5 (CM5) sensor chip, series S NTA sensor chip, amino coupling kit, 10× phosphate-buffered saline (PBS), 10× 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)-buffered saline with polysorbate 20 (HBS-P) buffer, NTA Reagent Kit, Dulbecco's modified Eagle's medium (DMEM), and fetal bovine serum (FBS) were obtained from Cytiva (Marlborough, MA, USA). The XanTec-NTA chip was purchased from XanTec Bioanalytics (Duesseldorf, Germany). The plasmids were synthesized by GenScript (Nanjing, China). Natural Product Library for High-Throughput Screening (L6000) was obtained from TargerMol (Boston, MA, USA). Precast gel (Tris-Gly, 4%–15%, 15 wells), precast gel electrophoresis solution (Tris-Gly), Western rapid membrane transfer solution (10× ), nitrocellulose membrane (0.45 μm), His-Tag antibody (mouse monoclonal antibody), horseradish peroxidase labeled goat anti-rabbit IgG (H+L), and Cell Counting Kit-8 (CCK-8) kit were purchased from Beyotime Biotechnology (Shanghai, China). 3,3′-Diaminobenzidine (DAB) horseradish peroxidase chromogenic kit was purchased from Epizyme (Shanghai, China). CXCR4 antibody (UMB2) was purchased from Abcam (Cambridge, UK). AMD3100, ginsenoside Re, glycyrrhizic acid, gracillin, and alkannin were procured from Efebio (Shanghai, China). The 4% paraformaldehyde was obtained from Univ-bio (Shanghai, China). Transwell cell chambers were purchased from Corning Inc. (Corning, NY, USA). Virus-like particle (VLP) isotype control protein and human CXCR4-VLP were purchased from Sino Biological (Beijing, China). Super FastPure Cell RNA Isolation Kit and ChamQ SYBR Color quantitative polymerase chain reaction (qPCR) Master Mix were purchased from Vazyme (Nanjing, China). PrimeScript™ RT reagent Kit with genomic DNA (gDNA) Eraser was purchased from Takara (Tokyo, Japan).

2.2. Protein synthesis and validation

2.2.1. Plasmid design

The full-length of CXCR4 gene (NM_001008540.2) was cloned into the pET-17b expression vector to construct a plasmid with His-tagged CXCR4. The start codon (AUG) was designed in a 'Kozak consensus' context (A/GCCAUGG), and a poly(A) sequence was constructed downstream of the CXCR4 gene. A hexa-His tag was incorporated into the N-terminus of the CXCR4 gene.

2.2.2. CFPS

Expression of CXCR4 was conducted using the Transcription/Translation System Kit according to the instruction of manufacture. For the reaction, 40 μL of TNT® T7 Quick Master Mix, 1 μL of 1 mM methionine, and 1 μg of plasmid DNA (CXCR4 or vehicle) were sequentially added, and ddH₂O was added to a final volume of 50 μL. The reaction was gently mixed with a pipette and incubated for 90 min at 30 °C. Then the obtained CXCR4-CFPS mixture was stored at 4 °C for later use.

2.2.3. Western blot assay

Western blots were performed with precast 4%–15% Tris-glycine polyacrylamide gels and transferred onto nitrocellulose membranes. Blots were blocked with 5% bovine serum albumin (BSA) for 1 h, and anti-His or anti-CXCR4 antibodies were incubated overnight at 4 °C with 5% BSA in tris-buffered saline with 0.1% Tween 20 (TBST). Following three times of washing with TBST, the peroxidase conjugated secondary antibody was incubated in TBST with 5% BSA for 1 h at room temperature. Membranes were incubated with ECL light chemiluminescence kit and then detected by the ChemiDoc Imaging System (Bio-Rad, Hercules, CA, USA).

2.3. SPR assay

All SPR experiments were performed on the Biacore T200 system (Cytiva) equipped with NTA or CM5 sensor chip at 25 °C.

2.3.1. Immobilization of CXCR4-VLP on SPR sensor CM5

The CXCR4-VLP was coupled to the flow cell (FC) 2 of the CM5 sensor chip using the amino-coupling method, while a control VLP was coupled to the FC1. The CXCR4-VLP and control VLP were dissolved into stock solutions with a final concentration of 1 mg/mL respectively, and the samples were diluted to a final concentration of 100 μg/mL with 10 mM sodium acetate buffer (pH 4.5). The chip surface was treated with a 1:1 (v/v) mixture of 0.2 M 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC) and 50 mM N-hydroxy succinimide (NHS) at a flow rate of 10 μL/min for 7 min. This was followed by injection of the sample solution for 10 min, and finally sealing of the activated chip surface with a 1 M ethanolamine injection at pH 8.5 for 7 min.

2.3.2. Immobilization of CXCR4-CFPS on SPR sensor NTA

FC1 was used as the reference cell while FC2 was used as detection cell. The running buffer was HBS-P. The NTA chip (either the NTA sensor chip from Cytiva or XanTec Bioanalytics) was cleaned by injecting 350 mm ethylenediaminetetraacetic acid (EDTA) solution for 60 s at a flow rate of 10 μL/min. Then the chip was loaded with Ni²⁺ solution (0.5 mM) by injection for 60 s. The CXCR4-CFPS was injected for 180 s at a flow rate of 5 μL/min on FC2, while the control protein was on FC1.

2.3.3. Activity and specificity of the SPR biosensor

AMD3100 was used as the positive drug for CXCR4, and the samples (0, 2, 4, 8, 16, 32, 64, and 128 μM) were injected at a flow rate of 30 μL/min. The binding time was 60 s, and the dissociation time was 90 s. The affinity constant (K_D) was measured and calculated by Biacore T200 Evaluation Software (Cytiva) to assess the activity of the sensor chip. Ciprofloxacin, a negative compound for CXCR4, was assayed in the same way to verify whether the sensor chip was specific or not.

2.3.4. Screening of natural product compounds

Ninety-six selected compounds from Natural Product Library were injected into the NTA chip at a concentration of 50 μM, with a binding time of 60 s and a dissociation time of 90 s. The flow rate was 30 μL/min. Compounds with higher resonance unit (RU) were screened out, and kinetic assays were further carried on to examine the specificity of these binders to the CXCR4 protein.

2.4. Cell culture

The SH-SY5Y cells were cultured in DMEM containing 10% FBS, with 100 units per milliliter of penicillin and 100 μg per milliliter of streptomycin. The cells were cultured in an atmosphere of 5% CO₂ at a constant temperature of 37 °C. The cells were observed under a microscope, and when they had reached approximately 80% of the full field of view, passaging was performed in order to continue the culture.

2.4.1. Cytotoxicity assay

A total of 100 μL of 5000 cells were added to each well of a 96-well plate, and the cells were treated with culture media containing different concentrations of compounds. The incubation was continued for 24 h. 10 μL of CCK-8 solution was added to each well. The wells without cells but with the corresponding amounts of cell culture medium, drug, and CCK-8 solution were served as the blank controls. The incubation was continued in a cell culture incubator, and the absorbance was measured with BioTek Synergy 4 microplate reader (Winooski, VT, USA) at 450 nm at 1 h.

2.4.2. Cell migration assay

A transwell assay was used to investigate the effect of the compounds on cell migration. 600 μL of culture medium containing 10% FBS was added to a 24-well plate and different concentrations of compounds were added, the final concentrations were 0, 10, 20, and 50 μM, respectively. A transwell chamber was placed in each well and SH-SY5Y cells were resuspended in DMEM medium without FBS and then added to the transwell chamber at a density of 1 × 10⁵ cells/well. The cells were further incubated for 24 h to allow the cells to migrate. SH-SY5Y cells were fixed with formaldehyde solution, and stained with 0.1% crystal violet and rinsed with water. Cells were observed and counted under 10× magnification using an Olympus IX71 microscope (Tokyo, Japan).

2.5. The validation of CXCR4 pathway inhibition by two compounds

Real-time fluorescent qPCR was performed to validate the effect of the two compounds on the CXCR4 downstream pathway. 8 × 10⁵ SH-SY5Y cells were added to each well of a six-well plate. After the cells have adhered, they were incubated with 50 μM glycyrrhizic acid and ginsenoside Re for 12 h. RNA was extracted using the Super FastPure Cell Isolation kit according to the operating instructions. gDNA was removed from the RNA using the PrimeScript™ RT Reagent Kit with gDNA Eraser, followed by reverse transcription into complementary DNA (cDNA). qPCR reactions were performed using the SYBR Green I hybrid fluorescent method. Primers for the target genes (interleukin-4 (IL-4), IL-6, matrix metalloproteinase-9 (MMP-9), tumor necrosis factor-α (TNF-α), and vascular endothelial growth factor (VEGF)) and the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were designed using Primer 3 Online, then synthesized by Sangon Biotech (Shanghai, China). Fluorescence signals were monitored in real time using the SYBR Green channel of the Roche LightCycler 96 (Basel, Switzerland) and the instrument automatically generated amplification and melting curves.

2.6. Molecular docking

The crystal structure of the CXCR4 protein was obtained from the Protein Data Bank (PDB) with the PDB ID 4RWS. Prior to docking, the viral chemokine antagonist vMIP-II was removed along with the water molecules. The chemical structures of all compounds were optimized using Discovery Studio 3.0 (DS 3.0) software and exported in mol2 format. The pairs were subjected to molecular docking using the CDOCKER module. All other options were maintained at the default settings, and the optimal conformations were evaluated based on CDOCKER binding energy. The results of the docking process were visualized using the PYMOL software.

2.7. SPR verification

Glycyrrhizic acid and ginsenoside Re were prepared in dimethyl sulfoxide (DMSO) to a final concentration of 10 mM as stock solutions respectively. The samples were then diluted with PBS buffer containing 5% DMSO to a series of concentration (64, 32, 16, 8, 4, 2, 1, 0.5, and 0 μM). The running buffer was PBS containing 5% DMSO. The samples were injected for 60 s at a flow rate of 30 μL/min, and the dissociation time was 90 s. The data obtained were fitted and analyzed using the steady-state binding model using the Biacore T200 evaluation software.

2.8. Statistical analysis

All cell assays were performed in at least three independent replicates. Data are presented as the mean ± standard deviation (SD). Differences between two groups were analyzed using Student’s t-test.

3. Results and discussion

3.1. Synthesis of CXCR4 protein using CFPS

In order to enhance the efficiency of translation process, a Kozak consensus context (A/GCCAUGG) was incorporated for the AUG start codon. DNA constructs with poly(A) sequences were designed to be located downstream of the genes, which are important for messenger RNA (mRNA) stability and may play a role in translation initiation in rabbit reticulocyte lysate. His can form ligand bonds with transition metal ions such as Ni²⁺ and Co²⁺, thereby a His-tag was incorporated into the N-terminus of the CXCR4. This allowed the CXCR4 protein to bind directly to the SPR NTA chip without the need for prior purification in subsequent experiments. The structure of the CXCR4 plasmid is shown in Fig. 1A. In accordance with the design, the Kozak sequence is situated after the T7 promoter sequence and before the target gene CXCR4, while the poly A tail is positioned subsequent to the target gene CXCR4.The TnT® T7 Quick Transcription/Translation System was used for in vitro synthesis of CXCR4 protein following Promega's instructions. Then Western blot analysis was performed on CFPS reaction mixtures to confirm the synthetic product. The experimental group contained the CXCR4-pET-17b plasmid, while the control group contained a vehicle plasmid. When employing a mouse monoclonal anti-His antibody as the primary antibody, a clear CXCR4 band was visible in the CXCR4-pET-17b plasmid group (Fig. 1B). Western blot assay result with the anti-CXCR4 antibody as the primary antibody also shows a clear band in the experimental group (Fig. 1C). Therefore, the CXCR4 protein was successfully synthesized by CFPS technology.

Fig. 1.

Schematic diagram of C−X−C chemokine receptor 4 (CXCR4) plasmid and validation of cell-free protein synthesis (CFPS) product. (A) Schematic of CXCR4-pET-17b plasmid. (B) The experimental and control groups were incubated with anti-histidine (His) antibody. (C) The experimental and control groups were incubated with anti-CXCR4 antibody. ORF: open reading frame.

3.2. Immobilization strategies of SPR biosensor

3.2.1. Selection of biosensor type

Three different coupling strategies to immobilize the CXCR4 protein on the biosensor were designed and the activity of biosensor was investigated. Specifically, the classic CM5 amino coupling approach using commercially available CXCR4-VLP were performed as a control group. VLP is the vehicle to indirectly express and stabilize the CXCR4 protein, and a control VLP vehicle without CXCR4 is needed to deduce the non-specific binding in SPR assay. By using the CXCR4-VLP protein, the amino coupling method that fixes proteins by covalently binding to carboxymethyl dextran on the sensor surface was adopted (Fig. 2A).

Fig. 2.

Three different strategies to construct the surface plasmon resonance (SPR) biosensor. (A) In C−X−C chemokine receptor 4 (CXCR4)-virus-like particle (VLP)/carboxymethylated dextran 5 (CM5) strategy, commercially available CXCR4-VLP was immobilized on the CM5 sensor chip by amino coupling method. (B) In CXCR4-cell-free protein synthesis (CFPS)/carboxymethylated dextran (CMD)-nitrilotriacetic acid (NTA) strategy, CXCR4 synthesized by CFPS was in situ purified and immobilized on the CMD-NTA sensor by histidine (His)-tag. (C) In CXCR4-CFPS/polycarboxylate hydrogel coating (HC)-NTA strategy, CXCR4 synthesized by CFPS was in situ purified and immobilized on the HC-NTA sensor by His-tag. Ni: nickel.

By using the CXCR4-CFPS protein, two types of NTA sensor chips with different modified surfaces were compared. One is consisted of carboxymethylated dextran (CMD) modified by NTA (Fig. 2B), and the other is coated with a biologically inert polycarboxylate hydrogel coating (HC) matrix modified NTA (Fig. 2C). NTA molecules can chelate metal ions, such as Ni²⁺, to form ligand sites that bind to His tags on the CFPS protein. Thus, the CXCR4-CFPS/CMD-NTA strategy and the CXCR4-CFPS/HC-NTA strategy utilized His-tag based method to purify CXCR4 protein from the CFPS mixture and to immobilize the protein on the NTA sensor in one step.

In order to verify the feasibility of the NTA sensor strategies, the activity of sensor was measured by CXCR4 inhibitor AMD3100. In CXCR4-VLP/CM5 strategy, the coupling amount of CXCR4-VLP was 1223 RU (Fig. 3A), and the K_D value of AMD3100 was 73.9 μM (Figs. 3B and C). In CXCR4-CFPS/CMD-NTA strategy, the coupling amount of CXCR4-CFPS was 1143 RU (Fig. 3D), and the K_D value of AMD3100 was 4.64 μM (Figs. 3E and F). In CXCR4-CFPS/HC-NTA strategy, the coupling amount was 24676 RU (Fig. 3G), which is higher than in the other two strategies. The K_D value was 23.5 μM (Figs. 3H and I). These results showed that all the three strategies demonstrated good response values and binding activities. Thus, CXCR4 synthesized by CFPS system can be immobilized by His-tag both on the CMD-NTA sensor and the HC-NTA sensor, and the activity of CXCR4-CFPS is in accordance with CXCR4-VLP. Interestingly, the NTA chip can purify and immobilize the target protein in one step, which is significantly better than the CM5 strategy in terms of speed, simplicity, and cost. Ultimately, the CXCR4-CFPS/CMD-NTA strategy was selected to perform the drug screening.

Fig. 3.

Immobilization of C−X−C chemokine receptor 4 (CXCR4) and affinity test of AMD3100 in three different strategies. (A) The coupling of CXCR4-virus-like particle (VLP) on the carboxymethylated dextran 5 (CM5) sensor. (B) The sensorgrams of AMD3100 on the CM5 sensor. (C) The fitting curves of AMD3100 on the CM5 sensor. (D) The coupling of the CXCR-cell-free protein synthesis (CFPS) on the carboxymethylated dextran (CMD)-nitrilotriacetic acid (NTA) sensor. (E) The sensorgrams of AMD3100 on the CMD- NTA sensor. (F) The fitting curves of AMD3100 on the CMD-NTA sensor. (G) The coupling of the CXCR-CFPS on the polycarboxylate hydrogel coating (HC)-NTA sensor. (H) The sensorgrams of AMD3100 on the HC-NTA sensor. (I) The fitting curves of AMD3100 on the HC-NTA sensor. The dashed lines in Figs. 3C, E, F, and I indicate the concentration corresponding to the affinity constant (K_D) value. RU: resonance unit.

3.2.2. Selection of running buffer

To ensure the activity of the CXCR4 protein on the SPR chip, we investigated the affinity of the positive drug AMD3100 under different running buffer conditions (including PBS and HBS-P). At the same time, to eliminate the interference of the SPR sensor surface, we compared the effects of different reference FC processing methods on the affinity determination, including NTA surface (NTA group), Ni covered surface (Ni group), and coupling with the control CFPS product (control protein group). The results are shown in Fig. S1. Figs. S1A−C are the NTA group, Ni group, and control protein group under the PBS running buffer condition, while Figs. S1D−F are the NTA group, Ni group, and control protein group under the HBS-P running buffer condition. The data indicate that the NTA group using the HBSP running buffer has the best activity and response, so this condition is adopted for the subsequent SPR detection.

3.2.3. The influence of non-specific binding and consistency verification

Since the CFPS product is a complex mixture containing many proteins, in order to eliminate the influence of other proteins, we coupled the control CFPS product onto the reference FC of SPR, so as to subtract the responses of these proteins from CXCR4 FC. From the results, it can be seen that the immobilization level of the control CFPS product on the SPR chip is close to zero (Fig. S2), indicating that proteins without His tags can hardly be coupled on the Ni-NTA chip. These results ensure that the other proteins in the CFPS product do not interfere with the specific detection of CXCR4. Furthermore, the negative compound ciprofloxacin was also injected and it produced essentially zero response values, thus further confirming the specificity of the constructed biosensors (Fig. S3). We compared the results of SPR purification and immobilization of 3 batches of CFPS synthesized proteins. The results are shown in Fig. S4. This indicates that the immobilization level of different batches of CFPS synthesized proteins are similar.

3.3. Screening of CXCR4 ligands from natural products

The SPR biosensor was used to screen 96 natural products, which may have anticancer or antitumor effects. The compounds (50 μM each) were analyzed by CXCR4-CFPS/CMD-NTA biosensor to compare the relative response (RR) adjusted by molecular weight (expressed as 100× RU/Da). As is shown in Fig. S5, the RR value of positive control was 1.8, and the RR values of S55, S60, and S96 were higher than that of positive control. Another five compounds with RR values near this threshold were also selected in order to avoid missing out the positive results, so totally 8 compounds were further subjected to kinetic assay to confirm whether their binding to CXCR4 were specific. The result shows that glycyrrhizic acid (Figs. 4A and B), gracillin (Figs. 4C and D), ginsenoside Re (Figs. 4E and F), and alkannin (Figs. 4G and H) could bind to CXCR4 specifically (Table S1). This determination was made based on the analysis of affinity curves and the subsequent fitting results. The results of the fitting analysis of the experimental data demonstrated that the interaction of artesunate (Fig. 4I), glucosyringic acid (Fig. 4J), germacrene (Fig. 4K), and β-elemene (Fig. 4L) with the CXCR4 protein exhibited robust non-specific binding characteristics, indicating an absence of the compounds' capacity to bind specifically to the CXCR4 protein. Consequently, these compounds were excluded from further studies.

Fig. 4.

Affinity test of the eight compounds screened out by surface plasmon resonance (SPR). (A) The sensorgrams of glycyrrhizic acid. (B) The fitting curves of glycyrrhizic acid. (C) The sensorgrams of gracillin. (D) The fitting curves of gracillin. (E) The sensorgrams of ginsenoside Re. (F) The fitting curves of ginsenoside Re. (G) The sensorgrams of alkannin. (H) The fitting curves of alkannin. (I−L) The response of artesunate (I), glucosyringic acid (J), germacrone (K), and β-elemene (L). The dashed line in Figs. 4A−H indicates the concentrations corresponding to the affinity constant (K_D) values. RU: resonance unit.

3.4. Inhibition of two compounds on cell migration

In order to confirm the four compounds having the potential to function as ligands of CXCR4, the pharmacological activity of these compounds against SH-SY5Y cells was evaluated by molecular biology experiments. First, the compounds' toxicity on cells was determined using the CCK-8 assay, and the results show that gracillin and alkannin exhibited high toxicity (Fig. 5). Gracillin demonstrated less than 50% cell survival at 6.25 to 50 μM (Fig. 5A), and the cell survival of alkannin was lower than 46% at 50 μM (Fig. 5B). Consequently, glycyrrhizic acid (Fig. 5C) and ginsenoside Re (Fig. 5D) were selected for subsequent experimentation. Then, the inhibitory effect of the two compounds on cell migration was tested using the transwell assay. The results demonstrate that glycyrrhizic acid and ginsenoside Re effectively hindered the migration of SH-SY5Y cells. Moreover, the number of transmembrane cells inhibited by both glycyrrhizic acid (Figs. 5E and F) and ginsenoside Re (Figs. 5G and H) at 25 and 50 μM exhibited a significant difference compared with the control group.

Fig. 5.

Results of biological experiments. (A−D) Cytotoxicity of gracillin (A), alkannin (B), glycyrrhizic acid (C), and ginsenoside Re (D) on SH-SY5Y cells. (E) Images of cell migration assay for glycyrrhizic acid. (F) Percentage of cell number passing through the membrane with different concentrations of glycyrrhizic acid. (G) Images of cell migration assay for ginsenoside Re. (H) Percentage of cell number passing through the membrane with different concentrations of ginsenoside Re. (I−M) Quantitative polymerase chain reaction (qPCR) results regarding the effects of glycyrrhizic acid and ginsenoside Re on the interleukin-4 (IL-4) (I), IL-6 (J), matrix metalloproteinase-9 (MMP-9) (K), vascular endothelial growth factor (VEGF) (L), and tumor necrosis factor-α (TNF-α) (M) pathways. ^∗P < 0.1, ^∗∗∗P < 0.001, and ^∗∗∗∗P < 0.0001, treatment group vs. control group.

3.5. Validation of CXCR4 pathway inhibition by two compounds

The impact of the compounds on the expression of downstream genes of CXCR4 was evaluated using qPCR method. Glycyrrhizin acid significantly reduced the relative expression of the IL-4 gene (Fig. 5I), but was not found to affect the expression of the IL-6 (Fig. 5J), MMP-9 (Fig. 5K), VEGF (Fig. 5L), and TNF-α (Fig. 5M). This reveals that glycyrrhizin acid may inhibit IL-4 gene transcription by blocking G protein-mediated signal transduction subsequent to CXCR4 activation. In contrast, Ginsenoside Re could reduce the relative expression of the MMP-9 gene (Fig. 5K), while the expression of IL-4 (Fig. 5I), IL-6 (Fig. 5J), VEGF (Fig. 5L), and TNF-α (Fig. 5M) was not significantly affected. Ginsenoside Re may inhibit nuclear factor kappa-B phosphorylation and nuclear translocation by blocking CXCR4-activated phosphatidylinositol-3-kinase/protein kinase B (PI3K/Akt) signaling, thereby reducing its activation of the MMP-9 promoter.

3.6. Molecular docking of the 2 compounds with CXCR4

In order to further validate and elucidate the potential interactions of glycyrrhizic acid and ginsenoside Re with CXCR4, a comprehensive molecular docking analysis was conducted (Fig. 6). The docking results for AMD3100 are displayed in Fig. 6A. It can be observed that AMD3100 formed hydrophobic interactions with amino acid residue ILE-284, and hydrogen bonds with ASP-97 and ASP-262, with a CDOCKER binding energy of −57.66 kJ. The docking results of glycyrrhizic acid and CXCR4 are illustrated in Fig. 6B. It is observed that glycyrrhizic acid forms hydrophobic interactions with the protein amino acids PHE-189 and VAL-196, and hydrogen bonds with TYR-190 and GLN-200, with a CDOCKER binding energy of −31.59 kJ. The docking result of ginsenoside Re is shown in Fig. 6C. It forms hydrophobic interactions with protein amino acids ILE-204, LEU-208, ILE-209, and ILE-259, and is capable of hydrogen bonding with amino acid residues of GLN-200 and TYR-256 with a CDOCKER binding energy of −9.92 kJ. The docking result indicates that glycyrrhizic acid and ginsenoside Re have negative CDOCKER binding energies, suggesting that these two compounds may be able to bind specifically to the CXCR4. The molecular docking result along with the affinity test and the cell migration assay indicate that glycyrrhizic acid and ginsenoside Re are potential CXCR4 small molecule ligands.

Fig. 6.

Schematic representation of the mode of action of C−X−C chemokine receptor 4 (CXCR4) with compounds. (A) AMD3100 forms hydrophobic interactions with protein amino acids ILE-284 and is capable of hydrogen bonding with amino acid residues of ASP-97 and ASP-262. (B) Glycyrrhizic acid forms hydrophobic interactions with the protein amino acids PHE-189 and VAL-196, and hydrogen bonds with TYR-190 and GLN-200. (C) Ginsenoside Re forms hydrophobic interactions with protein amino acids ILE-204, LEU-208, ILE-209, and ILE-259, and is capable of hydrogen bonding with amino acid residues of GLN-200 and TYR-256.

3.7. SPR verification

To further validate the reliability of the CXCR4-CFPS/CMD-NTA biosensor and its screening results, binding affinity of both compounds using SPR biosensors with other two strategies, CXCR4-VLP/CM5 and CXCR4-CFPS/HC-NTA, was determined respectively. The results of the three biosensors were consistent (Fig. 7). Specifically, in the CM5 biosensor, the K_D value of glycyrrhizic acid was 16.57 μM (Figs. 7A and B), while the K_D value of ginsenoside Re was 63.94 μM (Figs. 7C and D). In comparison, the HC-NTA group exhibited K_D values of 95.92 μM for glycyrrhizic acid (Figs. 7E and F) and 44.43 μM for ginsenoside Re (Figs. 7G and H). Although substantial variations in response intensity were observed among different biosensors, the K_D values obtained by each strategy exhibited a certain degree of similarity. This finding serves to substantiate the validity and reliability of using CFPS protein and Ni-NTA strategy in SPR biosensor, with the advantages of simplicity, rapidity, and easy operation. Consequently, this study furnishes a substantial foundation for combining CFPS technology and SPR analysis in drug screening and molecular interaction studies.

Fig. 7.

Affinity test of the two compounds by the other surface plasmon resonance (SPR) biosensor. (A) The sensorgrams of glycyrrhizic acid by C−X−C chemokine receptor 4 (CXCR4)-virus-like particle (VLP)/carboxymethylated dextran 5 (CM5) biosensor. (B) The affinity constant (K_D) value of glycyrrhizic acid by CXCR4-VLP/CM5 biosensor. (C) The sensorgrams of ginsenoside Re by CXCR4-VLP/CM5 biosensor. (D) K_D value of ginsenoside Re by CXCR4-VLP/CM5 biosensor. (E) The sensorgrams of glycyrrhizic acid by CXCR4-cell-free protein synthesis (CFPS)/polycarboxylate hydrogel coating (HC)-nitrilotriacetic acid (NTA) biosensor. (F) K_D value of glycyrrhizic acid by CXCR4-CFPS/HC-NTA biosensor. (G) The sensorgrams of ginsenoside Re by CXCR4-CFPS/HC-NTA biosensor. (H) K_D value of ginsenoside Re by CXCR4-CFPS/HC-NTA biosensor. The dashed lines mean the concentration corresponding to the K_D value. RU: resonance unit.

4. Conclusion

In this study, we proposed a high-throughput drug screening system that integrates CFPS and SPR technologies, and demonstrated the in-situ purification and immobilization of TMs on SPR biosensors, addressing a significant challenge in drug screening. The utilization of CFPS technology facilitates the expeditious synthesis of small amounts of target proteins in the nano-to-microgram range, accompanied by the seamless incorporation of labels, thereby ensuring optimal convenience for experimental procedures. This enhances the efficiency and conciseness of the experimental process, while concomitantly simplifying the expression and purification of protein ligands, particularly those that are challenging to purify using conventional methodologies. The combination of CFPS technology with SPR technology facilitates the purification and immobilization of target protein in situ in a single step, thus streamlining the experimental process.

For the construction of SPR biosensor, proteins with a His-tag were synthesized using CFPS, which can be purified and immobilized in Ni-NTA sensor chips in one step in SPR technology. Three different coupling strategies were constructed using proteins from different sources and different principles of biosensors, demonstrating the feasibility of coupling CFPS and SPR techniques. The verification results demonstrated that in the three immobilization strategies the CXCR4 antagonist AMD3100 elicited a specific response signal, whereas the control ciprofloxacin did not. This indicates the CFPS protein immobilized on the Ni-NTA biosensor is identical to the purified or VLP protein immobilized on the CM5 biosensor. In the drug screening stage, 96 natural products were screened at 50 μM. Consequently, eight compounds demonstrated relative higher response and were subjected to the verification experiments. The results of these assays indicate that four compounds, namely glycyrrhizic acid, gracillin, ginsenoside Re, and alkannin, exhibited strong binding affinitied to the CXCR4. Furthermore, glycyrrhizic acid and ginsenoside Re were found could inhibit the migration SH-SY5Y cell line and has the potential to act as a small molecule inhibitor of CXCR4 protein. Also, the molecular docking results for both compounds demonstrated their capacity to bind to the target protein. In order to further validate the CXCR4-CFPS/CMD-NTA biosensor and the screening results, the other two biosensors, CXCR4-VLP/CM5 and CXCR4-CFPS/HC-NTA, were utilized to assess the affinity of glycyrrhizic acid and ginsenoside Re. The results of the three biosensors were consistent. These validation steps not only proved the efficiency and accuracy of the CFPS-based SPR screening, but also provided important foundations for bioactivity testing and drug development by SPR analysis.

To sum up, a drug screening system was successfully constructed using CFPS technology and Ni-NTA biosensor for in situ purification and immobilization. The system was validated by screening compounds that act on CXCR4, and it is reasonable to conclude that it has broad applicability for screening other proteins. The system's functionality can be further enhanced through the exploration of novel SPR immobilization strategies, protein tags, and CFPS reaction systems, thereby facilitating drug studies targeting various proteins. Furthermore, CFPS and SPR technologies can be combined with other analytical techniques, such as MS and fluorescence imaging, to create a multifunctional high-throughput screening platform, and thus to improve the screening efficiency and accuracy.

CRediT authorship contribution statement

Xiaofei Wang: Investigation. Yanting Wang: Methodology. Hao Chen: Methodology. Ningqi Xia: Methodology. Jiayu Gu: Methodology. Linfeng Zhang: Methodology. Dongyao Wang: Methodology. Bin Lu: Resources, Data curation. Diya Lv: Writing – review & editing. Xiaofei Chen: Resources. Yan Cao: Writing – review & editing. Yifeng Chai: Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. As an editorial board member, Yifeng Chai recused himself from all review processes related to this article to ensure the fairness and objectivity of the review.

Appendix A. Supplementary data

The following is the Supplementary data to this article.