Introduction

In recent years, biopharmaceuticals have been widely used in the fields of diagnosis and treatment, which mainly include various vaccines, hormones, and protein. At present, they are widely used in the treatment of blood diseases, cancer, and cardiovascular diseases1. Since 2002, more than 300 biopharmaceuticals have been approved by the United States Food and Drug Administration. As an increasing number of biological preparations are applied to the fields of diagnosis and treatment, the application scope is also increasing2. There is a need to further improve production of biological agents, strive to improve the product quality of biological agents, and reduce the production cost in a shorter time to meet the increasing demand for these products3.

For more than 20 years, therapeutic recombinant proteins have been produced primarily by Chinese hamster ovary (CHO) cells. CHO cells have many advantages in this field, such as being able to survive and grow in suspension culture, possessing high safety, and being able to grow in serum-free medium and medium with defined chemical composition4. CHO cells can perform post-translational modification on the recombinant protein, and the glycosylation of glycoprotein produced by CHO cells is similar to that in humans5,6. Multiple gene amplification systems can exploit the genomic instability of CHO cells to achieve gene amplification and ultimately increased recombinant protein yields. At present, the titer of recombinant protein produced by CHO cells has reached the g/L level, and the titer level has been remarkable improved. Therefore, CHO cells are currently a commonly used expression system for recombinant proteins7,8,9.

However, in the field of biopharmaceuticals, high-efficiency expression of protein could not be separated from stable and high-yielding cell lines, and screening out more monoclonal cell lines could obviously improve the industrial production efficiency10,11. At present, the limited dilution method and flow cytometry are the most used cloning methods. Although both of them can obtain satisfactory cloning results, it takes a long time and slow efficiency. Under normal conditions, it takes 2 months to complete the screening of 1000 cell lines. In the cloning process, cells may also be subjected to mechanical damage, which dramatically reduces the number of cells and seriously affects the efficiency of the experiment12. Semi-solid medium is more suitable for isolating monoclonal cells because of its soft physical form, some enterprises often use semi-solid medium for monoclonal screening. In semi-solid medium, the cells are less affected during screening, which greatly increase the number of surviving cells and significantly improve the experimental efficiency13,14. At present, there are few choices of semi-solid medium in the international market, and its price is high. Some commercial semi-solid mediums are also unfavorable for the growth of certain cells. For example, the commercial ClonaCell-HY Medium D is a methylcellulose-based semi-solid medium that can be used for the screening and cloning of hybridomas15. However, when using this medium for cell cloning, cells often float on the surface of the colloid and move around, making it very difficult to select monoclonal clones. The Clone Pix system screening method is proposed by Molecular Devices. This method can complete the cell screening in shorter time than the artificial method. However, due to the long production period and high transportation cost, cell culture requires much money. Technically, this method requires the use of antibodies containing fluorescent markers. The secondary antibody is added into the semi-solid medium, which can detect the Fc of the antibody produced by the monoclonal cell mass, and the AntiFc-secondary antibodies can produce a fluorescent effect after being irradiated with light of a specific wavelength. When the cells are screened, high-yield monoclonal cells can be selected in accordance with different fluorescence intensities16,17. However, the method still needs at least 5 weeks during the whole screening process.

According to the above research status, an efficient screening system for recombinant proteins in CHO cells based on semi-solid medium and bicistronic vector was established in this study. The screening system can shorten the screening cycle and reduce the research cost while preserving the advantage of the semi-solid medium to screen monoclonal cell strains, thereby effectively promoting the large-scale production of recombinant proteins.

Results

Optimization of semi-solid medium

Six formulations of semi-solid medium were prepared, and their viscosity and the degree of methylcellulose dissolution were observed. When agarose and methylcellulose were added simultaneously and regardless of how the dissolution conditions were changed, insoluble impurities were always present in the medium. Thus, agarose was removed during formulation optimization. When the final mass concentration of methylcellulose was 1.5%, the viscosity of semi-solid medium was too low and the fluidity was large, which was not conducive to the screening of monoclonal cells. When the final mass concentration of methylcellulose was 3.5%, the viscosity of semi-solid medium was too high, which is not conducive to mixing with 2 × H-DMEM. The comparison among six different concentrations of methylcellulose showed that the semi-solid medium had the best performance when the final mass concentration of methylcellulose was 2.5% (Table S1).

In order to observe whether cells could survive in the semi-solid medium 48 h after transfection, cells transfected with pCMV-eGFP-F2A-RFP were inoculated into the prepared semi-solid medium (the final mass concentration of methylcellulose was 2.5%), and the fluorescence intensity of RFP was observed after culturing for 48 h. The results showed that the cells could survive in the semi-solid medium, and the transient expression of RFP could be observed in the semi-solid medium (Fig. 1A). Then, the cells continued to be cultured for 1 week. Their state was observed at any time during the culture, and an appropriate amount of fluid was administered. After 1 week of culture, cell clusters were visible to the naked eye (Fig. 1B).

Fig. 1
figure 1

Optimization of semi-solid medium. (A) Fluorescence of CHO-S cells in semi-solid medium for 48 h (magnification × 100); (B) results of CHO-S cells grown in semi-solid medium for 7 days, and RFP expression could be observed; (C) comparison of the activity in different monoclonal cells.

Full size image

Three cell clusters were selected from the semi-solid medium and cultured in a 24-well plate by using the complete medium. After the cells grew to a sufficient number, they were inoculated into a serum-free medium at 800,000 cells/mL and cultured in suspension. During the suspension culture, cell counts were performed every day, and the cell status was checked. The cells were compared with the conventional cultured CHO-S cells in the complete medium all the time. The results showed that cell viability of the control group decreased more rapidly from day 6 to day 9 of suspension culture. The monoclonal cells selected from the semi-solid medium with complete medium grew normally, which could also be suspension culture in serum-free medium, and no significant difference in cell viability was found from that of conventional cultured cells (Fig. 1C).

Vector optimization mediated by 2A peptide

In order to establish an efficient semi-solid medium screening system, we adopted a method that combines both vector construction and semi-solid medium. Among the strategies used to construct recombinant protein vectors, the most promising and widely used currently are the use of internal ribosomal entry sites (IRES) and self-cleavage 2A peptide. Of all 2A peptides, equine rhinitis virus (E2A), thosea asigna virus (T2A), foot-and-mouth disease virus (F2A) are the most commonly used. Experiments were performed by the 2A peptide-mediated biscistronic vectors. In this study, three expression vectors pCMV-eGFP-IRES-RFP-Bla, pCMV-eGFP-F2A-RFP-Bla, and pCMV-eGFP-E2A-RFP-Bla were successfully constructed (Fig. 2). After CHO-S cells were transient transfected with the three vectors, then three vectors were successfully screened using semi-solid medium, which were observed under the fluorescence microscope to detect the fluorescence expression of eGFP and RFP. The fluorescence results showed that upon transfection with F2A, E2A, and IRES vectors, when the downstream RFP expression was high, the upstream eGFP expression increased correspondingly. The F2A-mediated vector had the highest RFP expression (Fig. 3). Therefore, pCMV-eGFP-F2A-RFP-Bla was selected as the experimental vector. Secreted alkaline phosphatase (SEAP) is a secreted protein that can be used to obtain experimental samples from culture medium without lysing cells. Human serum albumin (HSA) is a highly water-soluble globular monomer plasma protein, which is a common recombinant protein drugs. During the subsequent experiments, the upstream gene (eGFP) of F2A was replaced with SEAP and HSA, two expression vectors pCMV-SEAP-F2A-RFP-Bla and pCMV-HSA-F2A-RFP-Bla were successfully constructed.

Fig. 2
figure 2

The schematic diagram of vector construction. CMV cytomegalovirus promoter, RFP red fluorescent protein, F2A foot-and-mouth disease virus, IRES internal ribosome entry site, eGFP enhanced green fluorescent protein, E2A equine rhinitis virus, Bla blasticidin, SEAP secreted alkaline phos-phatase, HAS human serum albumin.

Full size image
Fig. 3
figure 3

Comparison of eGFP and RFP expression levels in F2A, E2A and IRES vectors. (A) eGFP and RFP fluorescence diagrams of F2A, E2A and IRES vectors; (B) the fluorescence quantitative analysis of eGFP and RFP.

Full size image

Comparison of gene expression levels screened by semi-solid medium and limited dilution method

To compare whether the target gene expression was affected by different screening methods, in the present study, monoclonal cell lines transfected with SEAP and HSA vectors were cloned by limited dilution method and semi-solid medium screening system. Monoclonal cell lines transfected with the SEAP vector were screened, and the fluorescence of RFP was observed under the microscope. In accordance with the fluorescence intensity of each monoclonal cell line, 13 monoclonal cell lines with different RFP expressions were selected from the limited dilution method (Fig. 4A) and semi-solid medium screening system (Fig. 4B). The same experiments were performed using HSA transfected cells and the results can be found in Figs. S1–S5. Findings from HSA transfected pools validated the results obtained in SEAP transfected cells.

Fig. 4
figure 4

The expression levels of RFP in monoclonal cell strains transfected with SEAP vectors (magnification × 100). (A) Fluorescence diagrams of RFPs from 13 monoclonal cell strains selected by the limited dilution method, which were cultured for 7 days; (B) fluorescence diagrams of RFPs from 13 monoclonal cell strains selected by the semi-solid medium screening system.

Full size image

Thirteen monoclonal cells selected by limited dilution method were suspended in serum-free medium, and the supernatant medium and cell pellets were collected after 7 days. For the collected cell pellets, intracellular proteins were collected after cell lysis, and the RFP expression of each monoclonal cell line was detected by Western blot (Fig. 5A,B). The collected supernatant medium was detected for SEAP activity by using an alkaline phosphatase kit. The results showed that when the expression of the downstream gene RFP was high in the 13 monoclonal cell lines cloned by limited dilution method, the expression of the upstream gene SEAP increased correspondingly (Fig. 5C,D).

Fig. 5
figure 5

Fluorescence quantitative analysis of RFP and SEAP expression in monoclonal cell strains screened by limited dilution method. (A) Western blot was used to detect the expression of RFP; (B) the gray-scale analysis results of RFP; (C) stable expression levels of SEAP in 13 monoclonal cells; (D) Comparison of RFP and SEAP expression levels in 13 monoclonal cells.

Full size image

The 13 monoclonal cells selected by the semi-solid medium screening system were suspended in serum-free medium, and the supernatant medium (SEAP) and cell precipitates (intracellular protein RFP) were collected after 7 days. The expression levels of RFP in each monoclonal cell line were detected by Western blot (Fig. 6A,B). The supernatant medium was collected, and the screened cells were detected using an alkaline phosphatase kit for SEAP activity. The results showed that when the expression of downstream gene RFP was high in the 13 monoclonal cell lines screened by the semi-solid medium screening system, the expression of upstream gene SEAP increased (Fig. 6C,D).

Fig. 6
figure 6

Fluorescence quantitative analysis of RFP and SEAP expression in monoclonal cell strains screened by semi-solid medium screening system. (A) Western blot was used to detect the expression of RFP; (B) the gray-scale analysis results of RFP; (C) Stable expression levels of SEAP in 13 monoclonal cells; (D) comparison of RFP and SEAP expression levels in 13 monoclonal cells.

Full size image

The SEAP expression levels in monoclonal cells screened by the limited dilution method and semi-solid medium screening system were compared (Fig. 7). The results showed that the SEAP expression of eight monoclonal cells screened by semi-solid medium was higher than that of monoclonal cells screened by limited dilution method. Meanwhile, the SEAP expression of four monoclonal cells (Clone 6, 7, 9, 12) screened by the semi-solid medium screening system was lower than that of the monoclonal cells screened by limited dilution method. Therefore, the expression of SEAP from the monoclonal cells screened by limited dilution method and the semi-solid medium screening system had no significant difference, indicating that semi-solid medium screening did not reduce the target gene expression of monoclonal cells.

Fig. 7
figure 7

The SEAP expression levels in monoclonal cell strains obtained by limited dilution method and semi-solid medium screening system.

Full size image

Expression of target gene is independent of gene copy number

To investigate the molecular mechanism of transgene expression of target gene in monoclonal cell lines, the gene copy numbers were analyzed. The total DNA in CHO cells transfected with SEAP vector was extracted, further to explore the molecular mechanism of transgene expression of SEAP in the monoclonal cells screened by the semi-solid medium screening system, and the relationship between the relative gene copy number and SEAP expression was analyzed by qPCR. Three groups of monoclonal cell lines with different SEAP expression levels were selected. In particular, 1, 2, and 3 are monoclonal cell lines with high expression of SEAP; 4, 5, and 6 are monoclonal cell lines with moderate expression of SEAP; and 7, 8, and 9 are monoclonal cell lines with low expression of SEAP (Fig. 8). The results showed that a group of monoclonal cell lines with high expression of SEAP also showed the higher gene copy numbers. However, in a group of monoclonal cell lines with moderate expression of SEAP, the gene copy number of clone 6 was lower than those of monoclonal cell lines with low expression of SEAP. The experiments showed that the expression of SEAP was not dependent on the gene copy number in CHO cells.

Fig. 8
figure 8

The gene copy number of SEAP in monoclonal cell strains obtained by semi-solid medium screening system.

Full size image

Discussion

CHO cells have become an important host for the industrial production of recombinant protein drugs because they can adapt to high-density suspension growth, have strong ability to recombinant expression proteins, and have a post-translational processing mechanism close to that of humans18,19,20. In recent years, the related technologies and products about the CHO cell expression system have developed rapidly, including the transformation of host cell line21,22,23, the optimization of serum-free medium24,25,26, the design of target protein expression vector27,28,29, cell line screening strategy30,31,32, and the upgrade of downstream production process and target protein purification33,34. For improvement of the expression of recombinant proteins, the production cost of recombinant protein drugs must be reduced, which could more quickly occupy the market and maximize the use of patent protection period. At present, with the frequent occurrence of public emergencies, such as the spread of COVID-19, people have shifted their focus to obtaining high-yielding clonal cell lines more quickly than before35,36,37.

Due to the heterogeneity and transgene silencing in CHO cells38,39, the proportion of high-yielding clonal cells lines is low. When high- and low-yield clonal cells coexist, high-yielding clonal cells could further decrease or even disappear with cell expansion due to slower growth. Therefore, the screening of high-yielding clonal cells is particularly important. The construction steps of high-yielding clonal cell lines include obtaining cell pool under pressure, monoclonal screening, and expression level identification40,41. The traditional classical methods usually use MTX, MSX, blasticidin, puromycin, and neomycin to obtain a cell pool with a certain expression through screening pressure and further obtain monoclonal cell lines through multiple rounds of limited dilution method. Finally, a large number of monoclonal cell lines could be screened by ELISA to obtain high-expression cell strains. This process usually takes at least 3 months, which is time-consuming and laborious, as well as affects the rapid development of modern biomedicine. Image-based methods are currently the focus of research. By pressurizing cell pools and using various imaging principles to simultaneously monitor the formation of monoclonal cells and protein expression levels, the construction time of cell lines could be significantly shortened. With the help of staining technology, high-throughput monoclonal screening by flow cytometry can quickly sort the recombinant CHO cell pool42. However, this kind of methods require antibodies to remain on the surface of the cell membrane after being secreted into the extracellular layer, and the conditions are relatively harsh. Molecular Devices Co., Ltd. developed a monoclonal screening method that is based on semi-solid medium and ClonePix to overcome this shortcoming17,43. The single cells obtained by this method had limited growth space in the semi-solid medium, and the secreted proteins were retained in the peripheral cells. After immunostaining is performed, high-yielding clones could be effectively observed and directly selected using ClonePix imaging, thereby improving the screening efficiency of high-yielding clones. However, this method is highly dependent on instruments and equipment, and relies on the patented formula of semi-solid medium.

In this study, on the basis of the independent development of CHO culture medium and through the optimization of basic components and methyl cellulose concentration, a semi-solid medium with good properties, which were conducive to the growth of cell clones, was obtained, and CHO cells could be drawn by pipette. The monoclonal cell strain obtained using this medium can quickly adapt to the serum-free suspension medium without a domestication process and avoid the process of adherent screening to suspension production. First, we obtained the best properties of the semi-solid medium from six formulations of semi-solid medium. The CHO cells were inoculated into the semi-solid medium, and three monoclonal cells were randomly selected into the complete medium for expansion culture a week later. After the cells grew to a certain number, they were subjected to serum free medium for suspension culture, counted, and observed daily. The results showed that the monoclonal cells selected from the semi-solid medium to the complete medium could grow normally and be suspended in serum-free medium, and no significant difference was found between cell viability and conventional culture. Moreover, the screening of semi-solid medium did not affect the subsequent culture of cells.

In the present study, in order to establish an efficient semi-solid medium screening system, we developed a method that combines the vector construction and cell culture in semi-solid medium. 2A peptide has a short structure, usually consisting of 18 amino acids, and the expression of two linked genes is relatively balanced, which has gradually become an effective tool for constructing multi-gene expression. By designing different biscistronic vectors, the results showed that the upstream eGFP and downstream RFP expression levels of F2A vectors were the highest, and the expression of downstream reporter gene could be used to indicate the expression of upstream target protein without affecting the target protein expression. Therefore, the F2A vector was selected for subsequent experiments. On the basis of the F2A vector, the eGFP upstream of F2A was replaced with target genes, such as SEAP and HSA. First, the SEAP vector was transfected into the CHO cells, and 13 monoclonal strains were screened using limited dilution method and the semi-solid medium screening system separately to verify the expression of SEAP and RFP in these monoclonal cells. The results showed that when the expression of the downstream gene RFP in the monoclonal cells obtained by the two screening methods was high, the expression of the upstream gene SEAP increased. Therefore, following the above interesting results, monoclonal cell lines with high expression of target genes can be directly selected on the basis of reporter gene expression when using semi-solid medium screening method. Meanwhile, no significant difference was observed in the expression of SEAP in the monoclonal cells obtained by the two screening methods. Therefore, semi-solid medium screening does not reduce the target gene expression of monoclonal cells. Experiments done with the HSA vector further validated our observations with the SEAP vector.

When using the semi-solid medium screening system, monoclonal cells with high expression of target genes can be directly selected in accordance with the expression of reporter genes. Moreover, the time can be controlled within 3 weeks, which greatly shortened the time for screening high-yielding clones and reduced the cost and labor (Table 1). Therefore, the basic research, development, and production of recombinant protein drugs could be accelerated. In addition, to investigate the molecular mechanism of transgene expression of SEAP and HSA in monoclonal cell lines, the gene copy numbers of SEAP and HSA were analyzed. The results showed that the gene copy number of high expression clones was lower than those of low expression clones, and the expression levels of SEAP and HSA were independent of gene copy number. This novel system established in the present study is more suitable for the screening of recombinant proteins. In future, verifying the long-term effects of downstream reporter genes on cell growth at the large-scale production level is necessary, and further determining whether the reporter genes occupy protein expression resources and affect the expression of upstream target protein remains to be studied.

Table 1 Comparison of semi-solid medium screening and limited dilution method for screening high-yielding clones.
Full size table

Conclusions

In this study, a semi-solid medium that could effectively support the growth and monoclonal formation of CHO cells was designed, and it is suitable for the screening of CHO cells. A CHO cell screening system based on semi-solid medium was established by designing a bicistronic vector and using the downstream reporter gene to effectively indicate the expression of the upstream target protein. The screening strategy of CHO monoclonal cell line provides a novel method for the production of recombinant protein drugs and the rapid development of biomedical industry.

Methods

Formulation optimization of semi-solid medium

A certain amount of H-DMEM (Proteinasay, Xinxiang, China) recommended by manufacturer was dissolved in ultrapure water and filtered using a membrane with a diameter of 0.22 μM to prepare 2 × H-DMEM medium. Methyl cellulose was weighed and added into deionized water for high-pressure sterilization after dissolution. A 2 × H-DMEM medium containing 20% fetal bovine serum (Gibco, Grand Island, USA) and 20 µg/mL of blasticidin (Giemmyl, Shanghai, China) were prepared and mixed with a 2 × methyl cellulose (Sigma, St. Louis, MO, USA) solution (in a 1:1 ratio) at 600 rpm overnight under stirring at 4 °C. A semi-solid medium with final methylcellulose concentrations of 4%, 3.5%, 2.5%, and 1.5% was prepared by the above method. The specific formula is shown in Table S1.

Vector construction

In this study, pCMV-eGFP-IRES-Luc (Proteinasay, Xinxiang, China) was used as the skeletal vector, and both genes (eGEP and Luc) were under the control of the same CMV promoter. The Luc (GenBank no: M15077.1) in this vector was replaced by RFP (GenBank no: TID02864.1), and IRES (GenBank no: JQ692169.1) was replaced by F2A (GenBank no: EF576923) and E2A (GenBank no: M73260), which were synthesized by Sangon Biotech (Shanghai, China). Meanwhile, the resistance gene blasticidin (GenBank no: NW_019170238.1) was added into the vector, and RFP-blasticidin was treated as a fusion protein gene using In-Fusion Cloning technology, then three expression vectors of pCMV-eGFP-IRES-RFP-Bla, pCMV-eGFP-F2A-RFP-Bla, and pCMV-eGFP-E2A-RFP-Bla were constructed (Fig. 2A). In accordance with the pre-experimental results, pCMV-eGFP-F2A-RFP-Bla was determined to be the target vector, and the vectors of pCMV-SEAP-F2A-RFP-Bla and pCMV-HSA-F2A-RFP-Bla were constructed by biosynthesis of SEAP (GenBank no: KM403567.1) and HSA (GenBank no: CAA01491.1) sequences, which were replaced with eGFP (GenBank no: U55763.1) upstream of F2A (Fig. 2B). The above vectors were successfully constructed through enzyme digestion and sequencing.

Cell culture and transfection

CHO-S cells (National Biomedical Cell-Line Resource Center, Beijing, China) were cultured in an incubator maintained at 37 °C in a humidified atmosphere with 5% CO2 by using DMEM/F12 medium containing 10% fetal bovine serum (Gibco, Grand Island, USA). Twenty-four well plates were plated with a cell density of 1.8 × 105 cells per well. After approximately 24 h, the cells were transfected with Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) at a density of 70–80% in accordance with the manufacturer’s instructions. Forty-eight hours after transfection, transient transfection assay was performed, and monoclonal cells were cloned by limited dilution method using DMEM/F12 complete medium (DMEM/F12 medium containing 10% fetal bovine serum) containing 10 µg/mL of blasticidin (Genomeditech, Shanghai, China). The monoclonal cells were expanded and suspended in chemically defined serum-free medium (Proteinasay, Xinxiang, China), which can be used for suspension culture. After 7 days, the medium was collected to the concentration of target protein. All experimental protocols were approved by the Ethics Committee of Xinxiang Medical University. All methods were carried out in accordance with relevant guidelines and regulations.

Transient transfection analysis

CHO-S cells were subjected to fluorescence analysis using a fluorescence microscope (Nikon, Tokyo, Japan) 48 h post-transfection, and the fluorescence intensity of RFP was observed using green excitation light. The efficiency of transient transfection was evaluated in accordance with fluorescence intensity.

Establishment of the semi-solid medium screening system

CHO-S cells were plated into 12-well plate and transfected with Lipofectamine 2000 in DMEM. After approximately 48 h, 500–2000 cells/2 mL were diluted into wells containing semi-solid medium, and then screened in semi-solid medium for 7 days. During the culture period, an appropriate amount of fluid was administered to avoid drying of the semi-solid medium. The monoclonal cells screened by the semi-solid medium were transferred to the culture plate, further selected using 1 mL pipet tips (free sterile) for expansion culture, and then serum-free medium (Proteinasay, Xinxiang, China) was used for suspension culture. After 7 days, the medium was collected and concentration of target protein was assayed.

Secreted alkaline phosphatase (SEAP) detection

SEAP was measured in strict accordance with the kit’s instructions (Cat no. P0321S; Beyotime, Shanghai, China). Samples were sequentially added following the manufacturer’s instruction, and finally, the reaction termination solution was added. The solution absorbance was detected at 405 nm, and the SEAP activity of the samples was calculated in accordance with the OD value.

ELISA assay

Human serum albumin (HSA) was measured using an ELISA kit (Mlbio, Shanghai, China). A 96-well immunoassay plate was used to coat the captured antibody, and the captured HSA was detected using the conjugate of goat anti-mouse antibody and horseradish peroxidase. 3,3′,5,5′-Tetramethylbenzidine (TMB) substrate solution was used to measure the OD value of each well at 450 nm, and the concentration of HSA was calculated in accordance with the OD value. The detailed operating procedures were performed following the manufacturer’s instruction.

Western blot

The selected CHO-S cell lines were lysed, and the intracellular protein RFP was collected. After denaturation at 100 °C for 15 min, RFP was detected by Western blot. Then, 20 µL of the supernatant was extracted from the denatured samples and subjected to electrophoresis on 10% SDS-PAGE gel (Beyotime, Shanghai, China). The protein was transferred to PVDF membrane and incubated overnight with RFP (dilution of 1:1000, Santa Cruz, CA, USA) at 4 °C and then with horseradish-peroxidase-labeled goat anti-mouse secondary antibody (dilution of 1:5000, Sigma, St. Louis, MO, USA) for 2 h at 37 °C. After being washed with TBST, the protein bands were visualised using an enhanced chemiluminescence substrate kit (Amersham, GE Healthcare, Chicago, IL, USA). The relative expression levels of protein were determined using Image J software (version 1.41, National Institutes of Health, USA).

Fluorescent quantitative PCR

The relative gene copy numbers of SEAP and HSA were detected by fluorescent quantitative PCR (qPCR). Genomic DNA was extracted from the stable monoclonal cells in accordance with the manufacturer’s instruction (TaKaRa, Dalian, China). The primers for qPCR are shown in Table S2. For quantification, qPCR was performed using the ABI 7500 fluorescent quantitative detection system (Applied Biosystems, Foster City, CA, USA). The PCR reaction conditions were as follows: 95 °C for 4 min, followed by 42 cycles of 94 °C for 10 s, 60 °C for 30 s, and 72 °C 10 s. Moreover, qPCR was performed using the ABI 7500 SYBR PCR instrument (Applied Biosystems, Foster City, CA, USA) with cycle parameters of 4 min at 95 °C, 42 cycles; 94 °C for 10 s; 60 °C for 15 s; and 72 °C for 10 s. The relative copy numbers of SEAP and HSA were calculated using 2−△△CT method.

Statistical analysis

All data were analysed using SPSS 18.0 software (SPSS Inc, Chicago, IL, USA). One-way analysis of variance was used for inter-group comparison, and pairwise comparison was used for t-test. All experiments were repeated three times as independent analyses. P < 0.05 indicated that the difference was statistically significant.

Data availability

Data is provided within the manuscript or supplementary information files.