The trastuzumab expression cassette was designed in-house. Indeed, the trastuzumab heavy-chain (Tras-HC) and light-chain (Tras-LC) sequences were obtained from the Protein Data Bank (PDB) (1N8Z) and Drug Bank (DB00072) databases. Cytomegalovirus (CMV) promoter was selected to drive Tras-HC and Tras-LC expression. Then, the dual reporter gene for enhanced green fluorescent protein (eGFP) and puromycin was inserted under the control of the CMV promoter, using HindIII and SpeI restriction sites and terminated by BGH pA. The gene encoding the IL-2 secretory signal sequence was inserted upstream of Tras-HC and Tras-LC and three different poly A elements were inserted upstream of the CMV promoter. Then, the expression cassette was cloned into the donor vector (Piggy-Tras), and while the Piggy-Tras construct was generated by commercial synthesis (Vector Builder, NJ, USA).

Next, Escherichia coli DH5-alpha cells were transformed using the Piggy-Tras and Piggy-Trans plasmids. Sufficient plasmids were prepared for transfection. Following transformation, clones corresponding to each of the vectors were randomly selected for purification. Recombinant plasmids were purified from bacterial cells using the PureLink HiPure Plasmid Filter Maxiprep Kit (Thermo Fisher Scientific, Waltham, MA, USA).

The purified Piggy-Tras were amplified by PCR for Tras-HC and Tras-LC using primers (forward-HC: 5${}^{\prime }$-CAGATCCAAGCTGTGACCGG-3${}^{\prime }$, reverse-HC: 5${}^{\prime }$-TCTTATCATGTCTGGCCAGC-3${}^{\prime }$) and (forward-LC: 5${}^{\prime }$-ACAGATCCAAGCTGTGACC-3${}^{\prime }$, reverse-LC: 5${}^{\prime }$-CTGCATTCTAGTTGTGGTTTG-3${}^{\prime }$), respectively. Afterward, gel electrophoresis was performed on the amplicons using a 1% agarose gel. The sequences of the Piggy-Tras construct were confirmed by means of DNA sequencing.

The suspension Freestyle CHO-S cells (Thermo Fisher Scientific, Waltham, MA, USA) were maintained in FreeStyle™ CHO expression medium (Thermo Fisher Scientific, Waltham, MA, USA). The chemically defined, protein-free medium was supplemented with 8 mM L-glutamine (Invitrogen, Carlsbad, CA, USA), 2 Mm Pen-Strep (Invitrogen, Carlsbad, CA, USA), and an anti-clumping agent (Invitrogen, Carlsbad, CA, USA). The cells were kept in a humidified incubator on a shaker at 37 °C with 8% CO${}_2$. The CHO-S cells were cultivated in shake flasks and sub-cultured twice a week at a density of 2 $\times$ 10${}^6$ cells/mL. Cell viability was evaluated using the Trypan blue exclusion method and cell number was measured using an automated cell counter (Eve Automatic Cell Counter, NanoEnTek, Seoul, Korea).

Lipofectamine 3000 reagent (Life Technologies, Carlsbad, CA, USA) was used to co-transfect the FreeStyle CHO-S cells in triplicate. Twenty-four hours prior to transfection, 30 mL of FreeStyle™ CHO expression medium containing 6 $\times$ 10${}^5$ cells/mL was seeded into 125 mL shaking flasks. Three molar ratios were tested to optimize protein expression: 1:3, 1:5, and 1:10 helper/donor (Piggy-Trans/Piggy-Tras). On the day of transfection, 37.5 µg of helper/donor DNA was diluted in 600 µL of FreeStyle™ CHO expression medium. In a separate tube, 37 µL Lipofectamine 3000 reagent was diluted in 600 µL FreeStyle™ CHO expression medium. Diluted Lipofectamine reagent was added to the diluted DNA, incubated for 10 min at room temperature, and then, added dropwise to the cells. The cells were maintained at 37 °C with 8% CO${}_2$ in a humidified incubator. After 6 hours of incubation, the medium was replaced with a fresh medium containing puromycin. Trastuzumab expression was evaluated on day 4 post-transfection using a human IgG enzyme-linked immunosorbent assay (ELISA) kit (Sigma-Aldrich, St. Louis, MO, USA).

Before transfection, an enhanced stringency selection of the CHO-S cells was performed using a killing curve, which used puromycin concentrations ranging from 20 to 100 µg/mL (Sigma-Aldrich, St. Louis, MO, USA). Following, 48 hours of transfection, the cell media were changed and supplemented with the optimal puromycin concentration. Thereafter, the selective medium was replaced every three days for a month to obtain a stable pool, the stability of which was evaluated by fluorescence microscopy and sodium dodecyl polyacrylamide gel electrophoresis (SDS-PAGE).

The level of mAb expression in the cell culture supernatants (plates and shake flasks) was determined by indirect ELISA. The concentration of the secreted mAb was determined using the Herceptin Trastuzumab/Anti-HER2 Humanized IgG ELISA Kit (Alpha Diagnostics International, San Antonio, TX, USA). Briefly, 8-well strips were used, which were pre-coated with HER2 antigen and post-coated with stabilizers. Standards and culture supernatants were diluted, loaded into each well, and treated with anti-human IgG-horseradish peroxidase (HRP) conjugate. TMB (tetramethylbenzidine) (Sigma-Aldrich, St. Louis, MO, USA) and HRP substrate were added to the plates. After incubating for approximately 10–20 min at room temperature in the dark, sulfuric acid (H${}_2$SO${}_4$) (Merck, New York, NY, USA) at a concentration of 2N was added to stop the reaction. The absorbance was measured at 450 nm using a Multiskan™ FC Microplate Photometer (Thermo Fisher Scientific, Waltham, MA, USA). Washing procedures were performed between all steps using a PBS buffer containing 0.05% (v/v) Tween 20. Human IgG with a defined concentration was applied to draw a standard curve, with the trastuzumab quantitation range from 0.5 ng/mL to 10 ng/mL.

Fluorescence microscopy was used to monitor the stability of the cell pool in the shake flask during a one-month culture without selection pressure via the eGFP expression. The cells were washed twice with 1 mL of PBS, and fixed in 1 mL of ice-cold 100% methanol for 10 min before being washed again with 1 mL of PBS. The coverslips were mounted in ProLong Diamond Antifade Mountant and DAPI was applied (Life Sciences, Cambridge, MA, USA). The images were acquired using an EVOS® FL Cell Imaging microscope.

Since the donor vector was harboring eGFP, the encoding sequence transfection rates in the cells were determined 48 hours post-transfection. The cells were cultivated for two days at 37 °C before being resuspended in PBS. The FACSCalibur™ system (BD Biosciences, San Jose, CA, USA) was used to estimate the green fluorescence intensity of various clones. Green fluorescence at 525 nm was detected through an FL1 set at a PMT voltage of 400, with a logarithmic gain. Ten thousand cells were analyzed for each sample.

The methylcellulose-based semi-solid medium was prepared in-house using a minimum essential medium (MEM) 10X (Gibco, Waltham, MA, USA), 2% methylcellulose 4000 pentose (Sigma-Aldrich, St. Louis, MO, USA), 1% non-essential amino acids (Gibco, Waltham, MA, USA), 14 mM NaHCO${}_3$ (Sigma-Aldrich, St. Louis, MO, USA), 0.5% penicillin/streptomycin. The prepared medium was used to grow FACS-sorted CHO-S cells in isolated colonies. Briefly, the CHO-S population expressing eGFP was prepared by aseptic FACS sorting and plated in 60 mm Petri dishes at a range of cell densities to achieve a final density of 150 colonies per dish. After plating, single-cell separation was assessed under-inverted microscope. Then, the plates were incubated at 37 °C with 8% CO${}_2$. After 12 days of undisturbed incubation, the clones were selected based on the established picking parameters, including colony size, the distance between clones, and colony compactness. Well-dispersed colonies were picked manually and transferred to individual wells in a 96-well plate pre-filled with the Freestyle CHO expression liquid medium. Clones showing high mAb production in supernatants were subsequently expanded in 24-well plates.

The high clone producer T1A7 was cultured in triplicate in a batch culture mode, in a 125 mL Erlenmeyer flask containing 30 mL of FreeStyle™ CHO expression medium (Thermo Fisher, Waltham, MA, USA) supplemented with 8 mM glutamine (Invitrogen, Carlsbad, CA, USA), and 1% anti-clumping agent (Invitrogen, Carlsbad, CA, USA). Flasks were inoculated with 1 $\times$ 10${}^6$ viable cells/mL and kept in a humidified incubator shaker at 37 °C with 8% CO${}_2$. Samples were collected every day and the mAb concentration was quantified using a Herceptin Trastuzumab/Anti-HER2 Humanized IgG ELISA Kit (Alpha Diagnostics International, San Antonio, TX, USA).

After 7 days of batch culture, clarification was carried out on the T1A7 clone by centrifuging the cells at 5000 $\times$g for 15 min, followed by filtering the supernatants. Then, the clarified supernatants were dialyzed with the binding buffer, consisting of 20 mM sodium phosphate, and 0.15 M NaCl, at pH 7.2. Subsequently, protein purification was performed by fast protein liquid chromatography (FPLC) (Cytiva, Uppsala, Sweden). Five column volumes of 20 mM sodium phosphate buffer, at pH 7.2, were used to equilibrate the HiTrap MabSelect SuRe column (GE Healthcare Life Sciences, Danderyd, Sweden) before loading the concentrated supernatants. Then, the column was washed with 10 column volumes of binding buffer to remove any contaminants. Subsequently, elution was conducted using an elution buffer consisting of 0.1 M sodium citrate, at pH 3.3. The collected fractions were analyzed by SDS-PAGE.

Purified mAb produced from the T1A7 clone was analyzed by SDS-PAGE under reducing conditions. Samples were heated at 95 °C for 5 min and loaded into the gel wells. As a negative control, 10 µL of non-transfected CHO-S cells supernatant was loaded alongside 10 µL of the three purified samples. Coomassie blue staining was performed to visualize the resulting protein bands.

The protein aggregates were analyzed using an Agilent HPLC system (1260 Infinity II series). For that purpose, an Agilent AdvanceBio size exclusion chromatography (SEC) column was used with the following characteristics: 7.8 $\times$ 300 mm, 300 Å pore size, and 2.7 µm particle size. The column was previously equilibrated in PBS 1$\times$ (0.8 mL/min) with the oven set to 30 °C. The isocratic program was set to analyze approximately 5–15 µg of each sample via a 10 µL injection at 0.8 mL/min for 20 min. A series of proteins from the Sigma-Aldrich protein standard mix were used to calibrate and test the SEC columns. The proteins used included ribonuclease A (13,700 Da), ovalbumin (44,300 Da), gamma globulins (150,000 Da), and thyroglobulin (670,000 Da). The mass of the mAb secreted by the T1A7 clone was determined with an error of less than 15%. After SEC analysis, 214 and 280 nm chromatograms were extracted from the raw data and analyzed by peak integration.

Statistical analysis was carried out using GraphPad Prism software (version 9.2.0 (283), GraphPad Software, Inc., San Diego, CA, USA), with two-tailed p-values below 0.05 being considered statistically significant, and a 95% confidence interval (*p 0.05, **p 0.01 and ***p 0.001). For cell pool generation, a one-way ANOVA test was performed. Normality tests were also performed in order to evaluate the population distribution.

The generated Piggy-Trans and Piggy-Tras vectors were obtained from Vector Builder, Englewood Cliffs, NJ, USA. The expression cassette designed to express the full-length trastuzumab (Tras-HC sequence encoding 451 amino acids and the Tras-LC encoding 215 amino acids) was inserted between the 5${}^{\prime }$ ITR and the 3${}^{\prime }$ ITR, thereby flanking the ends of the Piggy-Tras (Fig. 1a(1)). Considering that PB transposons can carry large cargo [26, 27], we designed a large expression cassette in which the choice of each element defining the expression cassette was investigated with the goal of achieving high protein expression, including cis-regulatory elements and enhancers [28], CMV promoters, SV40 polyA signal, polyadenylation signal of bovine growth hormone (BGH pA), and Kozak sequences. Other genes enabling plasmid replication and antibiotic resistance in bacteria and mammalian cells were also investigated. Indeed, the Piggy-Trans plasmids included the transposase sequence under the control of the cytomegalovirus immediate early enhancer-chicken $\beta$-actin hybrid (GAG) promoter (Fig. 1a(2)). The insertion of the Tras-HC and Tras-LC was evaluated by PCR. Agarose gel electrophoresis (1%) of the Tras-HC and Tras-LC PCR products confirmed the presence of a 1356 bp fragment that encoded Tras-HC and a 648 bp fragment that encoded Tras-LC (Fig. 1b).

Fig. 1.

Representation of the PiggyBac vector maps for Piggy-Tras and Piggy-Trans and agarose gel illustrating Tras-HC and Tras-LC insertion. (a) Piggy-Tras and Piggy-Trans maps. (1) Piggy-Tras vector containing the expression cassette encoding the full-length trastuzumab (Tras-HC and Tras-LC) under the control of the cytomegalovirus promoter (CMV). It also harbors the pressure selection genes (puromycin and ampicillin) and enhanced green fluorescent protein (eGFP). (2) Piggy-Tras encoding the transposase under the control of the enhancer-chicken $\beta$-actin hybrid (GAG) promoter. (b) Agarose gel electrophoresis of the Tras-HC and Tras-LC PCR products. The gel confirms the presence of a 1356 bp fragment that encodes Tras-HC (band in column 1) and a 648 bp fragment that encodes Tras-LC (band in column 3). Column 2 corresponds to the ladder.

To generate a stable CHO-S cell line, a puromycin kill curve representing a dose-response experiment was performed to determine the minimum concentration of puromycin that can kill all the cells during a specific period. Twenty-four hours after culturing the non-transfected CHO-S cells, the growth medium was replaced with a selection medium that had been supplemented with a range of puromycin concentrations (20 µg/mL–100 µg/mL). A concentration of 50 µg/mL was determined as the minimum concentration that would kill all the cells within 9 days (Fig. 2).

Fig. 2.

Chinese hamster ovary (CHO)-S cell puromycin kill curve. To determine the optimal concentration for selecting transfected cells, a puromycin concentration range of 20 µg/mL to 100 µg/mL was used on CHO-S cells for two weeks and performed in triplicate.

The impact of the Piggy-Tras:Piggy-Trans ratio on the generation of a stable CHO-S pool was also investigated. For that purpose, cells were co-transfected in triplicate with three Piggy-Tras/Piggy-Trans ratios (1:3, 1:5, and 1:10) to determine the optimal conditions for efficient insertion of the gene of interest (GOI) into the host cell genome, and to generate an efficient cell pool with a high mAb expression and production [29, 30]. The transfection supernatants of the three ratios were collected and the secreted antibodies were evaluated using the human IgG ELISA kit. mAb quantification showed that a 1:3 transfection ratio of Piggy-Tras/Piggy-Trans was optimal in CHO cells to obtain higher titers of secreted antibodies (5.1 µg/mL), compared to 0.58 µg/mL for the 1:5 ratio and 0.38 µg/mL for the 1:10 ratio (Fig. 3). Subsequently, the cells from the 1:3 ratio pool were maintained in the selection medium (50 µg/mL of puromycin) for one month. The cell viability was constantly monitored during the 1-month generation of stable cells. During the first days of culturing under selective conditions, the proportion of viable cells was very low (12%) compared to the non-transfected cells (88%). Gradually, the cell viability of the transfected cells improved and reached 98% on day 31 (Fig. 4a). Fluorescence microscope showed that most of the CHO-S cells in the stable pool expressed GFP (Fig. 4b(B)), while non-transfected cells (Fig. 4b(A)) show no expression of GFP. In addition, SDS-PAGE, under reducing conditions, showed that the migration of the cell pool supernatant (in triplicate) was characteristic of an antibody, with the HC and LC bands observed at the expected sizes of 50 kDa and 25 kDa, respectively (Fig. 4b(C)).

Fig. 3.

Impact of Piggy-Tras/Piggy-Trans ratio on antibody expression. Three pools of CHO-S cells were generated (in triplicate) at three different ratios: 1:3, 1:5, and 1:10. Following 4 days of co-transfection, mAb titers were measured using a human IgG enzyme-linked immunosorbent assay (ELISA) kit. Error bars represent the average SD from triplicate measurements. ANOVA (Kruskal–Wallis test) was used to detect statistically none significant (ns) and significant differences between the generated pools.

Fig. 4.

Fluorescence microscopy and sodium dodecyl polyacrylamide gel electrophoresis (SDS-PAGE) evaluation of stable pool generation. (a) Stable cell pool generation: During one month of cell culture under puromycin selection, the CHO-S cell viability was measured. Indeed, the CHO-S cell viability was low (12%) during the first days but increased to 98% by day 23. (b) Fluorescence microscopy and SDS-PAGE evaluation of the stable pool. (A) The supernatant of non-transfected CHO-S cells was visualized by fluorescence microscopy and used as a negative control. (B) The CHO-S stable pool shows the expression of eGFP protein, which is proportional to the expression of the mAb. (C) SDS-PAGE analysis following Coomassie blue staining: migration of the CHO-S stable cell pool supernatants (in triplicate: lane 1, 2, 3) in SDS-PAGE (10%) under reducing conditions showing the expected bands, 50 kDa for Tras-HC and 25 kDa for Tras-LC. Lane 4 refers to the positive control: commercial Herceptin® (Roche, Basel, Switzerland).

The strategy of isolating stable CHO-S clones was based on the combination of fluorescence-activated cell sorting (FACS) followed by methylcellulose semi-solid media. The FACS Calibur system was used to purify the stable cell pool based on the eGFP expression. The eGFP fluorescence was detected in the FL1 channel, and the population of interest was gated to sort the sterile selected CHO-S population (Fig. 5a). The cells were collected for subsequent culturing in 2% MCM. The medium formulation was optimized in-house and included basal 10$\times$ MEM, glutamine, vitamins, amino acids, and sodium bicarbonate. The optimized 2% MCM was used for the expansion of the FACS-purified CHO-S population into 60 mm Petri dishes in a humified 8% CO${}_2$/air mixture at 37 °C. After 12 days of controlled incubation, 1500 clones were selected based on size, compactness, and inter-colony distance (Fig. 5b). Thereafter, 100 clones were picked and expanded. Every single clone was cultured separately in 96-well plates containing Freestyle expression medium for approximately 4 days and observed under an optical microscope. The supernatants were analyzed using the ELISA Herceptin Kit, as shown in Fig. 5c. The three top clones T1A1, T1A7, and T1B7 showed the best antibody expression levels at 26.87 ng/mL, 34.16 ng/mL, and 28.30 ng/mL, respectively.

Fig. 5.

CHO-S cell sorting and clone generation. (a) Cell sorting analysis was performed on the stable pool of CHO-S cells. GFP fluorescence was detected in the FL1 channel and the population of interest with high eGFP expression was gated for sterile sorting. (b) Following the cell sorting, the cells were cultured in optimized 2% MCM, the figure shows the clones generated after 12 days of incubation in a humidified 8% CO${}_2$ air mixture at 37 °C. (c) The selected clones were transferred to 96-well plates and the mAb expression was determined after 4 days using the ELISA Herceptin Kit.

The T1A7 clone was chosen for the triplicate batch culture investigation. Indeed, 30 mL of Freestyle CHO expression medium was inoculated with the clones seeded at a density of 1 $\times$ 10${}^6$ cells/mL and incubated at 37 °C with 8% CO${}_2$ and shaken at 125 rpm. The cell density and viability of each clone were monitored daily during the 7 days of batch culturing (Fig. 6(1)). The trastuzumab in the T1A7 supernatant was quantified every day during the batch-culture period and was observed to reach its maximum on day 5, with 4.24 g/L of the antibody being secreted (Fig. 6(2)). Then, trastuzumab was purified by FLPC using the HiTrap MabSelect SuRe column as described previously.

Fig. 6.

Expression and quantification of mAbs from the supernatant of the T1A7 clone during the 7-day batch-culture mode. (1) Cell density and viability during the batch-culture mode. (2) Determination of the trastuzumab titer by the ELISA Herceptin Kit.

The purity and size of Herceptin® (Roche, Basel, Switzerland) and the mAbs secreted by our T1A7 clones were analyzed by SDS-PAGE and SEC-HPLC. Indeed, the molecular size of Herceptin® and the purified mAbs secreted from the T1A7 clones were also analyzed by SDS-PAGE, under reducing conditions, and showed that the protein bands of both molecules matched, while there was also an absence of any other low or high molecular weight protein bands (Fig. 7b).

Fig. 7.

Size and aggregate analyses. (a) SEC chromatogram obtained from SEC-HPLC showed a main peak with a mass of 165,000 g/mol that was eluted at 8.691 min and a mass of 3000 g/mol that was eluted at 13.982 min. (b) SDS-PAGE analysis of the purified mAbs from the T1A7 clone following Coomassie blue staining: after protein A purification of mAbs from the T1A7 clone supernatant, mAb migration (in triplicate: lane 1, 2, 3) in SDS-PAGE (10%) under reducing conditions showed the expected bands, 50 kDa for Tras-HC and 25 kDa for Tras-LC. Lane 4 refers to the positive control: commercial Herceptin® (Roche).

According to the calibration curve, the purity of the mAbs secreted from the T1A7 clone using SEC-HPLC was 95%, with a retention time of 8.691 min for the main peak, and a mass of 165,000 g/mol. In addition, 3% high molecular weight (HMW) aggregates were present (Fig. 7a), without any low molecular weight (LMW) aggregates.