Evaluation of the Structural, Physicochemical, and Biological Characteristics of SB11, as Lucentis^® (Ranibizumab) Biosimilar

EU/US-ranibizumab, which expired from December 2016 to August 2021, was purchased from local distributors and then stored according to the manufacturer’s instructions.

Peptide mapping was performed to analyze structural integrity, including post-translational modifications and disulfide bond by liquid chromatography-electrospray ionization-tandem mass spectrometry (LC–ESI–MS/MS). To achieve denaturation and reduction, ranibizumab was mixed with 8 M urea and 1 M dithiothreitol; 1 M iodoacetamide was added. The sample was digested with trypsin (11047841001, Roche, Basel, Switzerland), Lys-C (90051, Thermo scientific, Waltham, MA, USA), and Asp-N (11058541103, Roche). For the disulfide linkage analysis, the reducing step was not performed, and the non-reduced samples were digested with trypsin. The digestion samples underwent reverse-phase ultra performance liquid chromatography-mass spectrometry (RP-UPLC-MS) using a BEH300 C18 column (186003687, Waters, Milford, MA, USA) at 60 °C. Peptides were eluted by a linear gradient at a flow rate of 0.3 ml/min for 100 min. Data were collected and processed by MassLynx v4.1 (Waters) and/or BiopharmaLynx v1.2 (Waters).

For far-UV studies, ranibizumab was dialyzed with 10 mM sodium citrate buffer and diluted with 10 mM sodium citrate. For near-UV studies, the samples were not dialyzed but instead diluted with SB11 formulation buffer. Circular dichroism spectroscopy (CD) measurements were performed using a Chirascan Q100 (Applied Photophysics, Leatherhead, UK) with 0.1-mm pathlength cells for far-UV and 10-mm pathlength cuvette for near-UV. The observed CD data in ellipticity for each sample were blank-subtracted and the average of the triplicate scans was used to make the CD plot. The CDNN algorithm was used to fit the far-UV CD data for prediction of secondary structure.

The Fourier transform infrared spectroscopy (FT-IR) spectra were recorded from a Nicolet iS50 FT-IR spectrophotometer (Thermo Scientific), equipped with a Smart Orbit diamond ATR accessory and the OMNIC software package for spectrometer control and data analysis. All spectra were acquired at 4 cm⁻¹ resolution between 4000 and 525 cm⁻¹, and a background scan with the matching SB11 formulation buffer was acquired prior to a scan of each sample. Ranibizumab spectra were partitioned into peak areas according to structural contribution and the results averaged over three replicates per sample.

Nano-differential scanning calorimetry (DSC) (TA Instruments, New Castle, DE, USA) was used to analyze the melting temperature of samples. The ranibizumab and SB11 formulation buffer were scanned from 15 to 100 °C using a scan rate of 1.5 °C/min. Ranibizumab was diluted with the SB11 formulation buffer prior to the run. To determine the baseline value, the SB11 formulation buffer was loaded into a sample channel and scanned to verify a reproducible baseline for subtraction from sample data. Data were analyzed by TA Instruments NanoAnalyze software.

Hydrogen/deuterium exchange-mass spectrometry (H/DX-MS) was adapted to compare higher order structures between ranibizumab samples. H/DX-MS was initiated by a 1:8 dilution of sample in D₂O buffer at intervals of 10 s, 1 min, 10 min, 1 h, and 4 h before quenching and injecting into the mass spectrometer. Peptides were digested on an immobilized pepsin column, and the trapped peptide fragments were eluted by a gradient of 8% to 92% acetonitrile in 15 min. Mass spectra were collected in MS^E mode, and data were analyzed by ProteinLynX Global Server™ (Waters) to identify peptides and by DynamX software (Waters) to calculate deuterium uptake and generate butterfly and difference plots.

Ranibizumab was injected onto a TSK-Gel Super SW2000 SW_XL analytical column (0018674, Tosoh, Tokyo, Japan), which was connected to a Waters HPLC system; monitoring was done by UV detection at 280 nm. The flow rate was 0.2 ml/min. Data were acquired and processed by Empower^®3 software (Waters).

Reduced and non-reduced capillary electrophoresis-sodium dodecyl sulfate (CE-SDS) analyses were conducted with a high-performance capillary electrophoresis system (PA800 plus Pharmaceutical Analysis System; SCIEX, Framingham, MA, USA). For the reduced condition, ranibizumab was mixed with a 10-kDa internal standard, SDS-MW sample buffer (A10663, Beckman Coulter, Brea, CA, USA), and 2-mercaptoethanol and then boiled at 70 °C for 5 min. For the non-reduced condition, 2-mercaptoethanol was replaced with iodoacetamide. The sample was electrokinetically introduced onto a capillary (base fused-silica capillary; Beckman Coulter) by applying voltage at -5 kV for 20 s and separated in the capillary cartridge. Electrophoresis was performed at a constant voltage with an applied field strength of -497 V. Data were acquired and processed by 32 Karat software (SCIEX) with integration capabilities.

Ranibizumab was mixed with pharmalyte 3–10, pharmalyte 8–10.5, 1% methyl cellulose, distilled water, pI 6.61 marker, and pI 9.5 marker. The mixture was loaded onto an iCE3 icIEF instrument (ProteinSimple, San Jose, CA, USA) using a capillary cartridge at 4 °C. Data were acquired by CFR software (ProteinSimple) and processed by Empower^®3 software (Waters).

Relative binding activities of ranibizumab to VEGF-A 165/189/121/110 were measured using an enzyme-linked immunosorbent assay (ELISA). A recombinant human vascular endothelial growth factor (rhVEGF), VEGF-A 165 (293-VE/CF, R&D Systems, Minneapolis, MN, USA), VEGF-A 189 (8147-VE/CF, R&D Systems), VEGF-A 121 (4644-VS/CF, R&D Systems), and VEGF-A 110 (5336-VE-010/CF, R&D Systems), was absorbed onto a 96-well plate. Serial dilutions of ranibizumab bound to rhVEGF followed by the sequential addition of anti-human IgG (Fab-specific)-peroxidase antibody, tetramethylbenzidine (TMB, T0440, Sigma-Aldrich, St. Louis, MO, USA) substrate, and 1 N sulfuric acid. The measurement of the absorbance at a wavelength of 450 nm quantitated the relative binding activity of ranibizumab to rhVEGF. The VEGF binding activity of ranibizumab sample was calculated relative to a reference standard using a four-parameter logistic model [PLA (Parallel Line Analysis); Stegmann Systems GmbH, Rodgau, Germany].

HUVEC (C2519A, Lonza, Basel, Switzerland) in endothelial cell growth medium (EBM™-2, CC-3156, Lonza) was treated on microplates where VEGF A-165 (293-VE/CF, R&D systems) and serially diluted ranibizumab were pre-incubated and proliferated for 4 days at 37 °C, 5% CO₂ incubator. The relative cell proliferation was measured by CellTiter-Blue^® cell viability assay kit (G8082, Promega, Madison, WI, USA) using a SpectraMax (M3; Molecular Devices, San Jose, CA, USA). Anti-proliferation activity was calculated relative to a reference standard four-parameter logistic model (PLA; Stegmann Systems GmbH).

Engineered cell lines for VEGF signaling (NFAT-RE-Luc2P/KDR 293 cells, E8510, Promega) in assay media (DMEM, 11995, Gibco, Grand Island, NY, USA) were treated on microplates where VEGF A-165 (293-VE/CF, R&D systems) and serially diluted ranibizumab had been pre-incubated. After cell treatment, the microplate was incubated at 37 °C, 5% CO₂ incubator for 3.5–6 h to induce luciferase gene expression by VEGF signaling. The relative VEGF neutralizing activity was measured by a luminescent detection system (Steady-glo^®, E2520, Promega) via microplate reader (EnVision, 2104, Perkin Elmer, Waltham, MA, USA). VEGF neutralization activity was calculated relative to a reference standard four-parameter logistic model (PLA; Stegmann Systems GmbH).

The peptide chromatograms of trypsin or Lys-C digested SB11 and EU/US-ranibizumab showed a highly similar peak profile without missing or additional new peaks with comparative retention time (Fig. 1A, B). In addition, the data confirmed that the amino acid sequences of SB11 and EU/US-ranibizumab were identical and that the sequence coverage by MS/MS was 100% (data not shown). Based on the results of intact mass and peptide mapping analysis, the molecular weights of SB11 and EU/US-ranibizumab were identical (considering assay variability; 0.01% of molecular weight and 30 ppm for intact protein and peptide, data not shown). Sequence variants generated by post-translational modifications such as oxidation and deamidation were identified. The CDR region of Fab contains methionine (Met) residues that are susceptible to oxidation and asparagine (Asn) residues that are susceptible to deamidation [15]. The relative levels of Met oxidation and Asn deamidation were similarly very low (< 0.4%) in all samples (data not shown). Disulfide bond analysis of SB11 and EU/US-ranibizumab under non-reducing condition after trypsin digestion showed that all ten cysteine residues are properly involved in the formation of the five disulfide bonds (Fig. 1C–E, Table 2).

The higher order structures of SB11 and EU/US-ranibizumab were compared using CD, FT-IR spectroscopy, and DSC. The far UV CD spectrum of SB11 and EU/US-ranibizumab showed comparable profiles throughout much of the far-UV region, especially at the local minimum at around 218 nm, which strongly corresponds to the secondary structure of protein folding (Fig. 2A). However, as the curves reach the far-UV region lower than about 210 nm, there was some divergence within samples. In this region, the buffer absorbance and concentration differences can play a role; this non-overlap does not necessarily correlate with a difference in protein folding. In addition, the samples showed comparable near-UV CD profile maxima at 258 nm, 265 nm, 285 nm, and 291 nm, respectively (Fig. 2B). No significant differences between the samples were observed, suggesting the samples share common tertiary structure characteristics. Secondary structure was further identified by FT-IR spectroscopy. The second derivative spectra of SB11 and EU/US-ranibizumab in the amide I band (1600–1700 cm⁻¹) and their distribution into the amide regions support high similarity in secondary structures (Fig. 2C). The DSC thermograms of samples showed highly similar thermal profiles and thermal transition midpoint temperatures, indicating that the thermal stability and conformation of SB11 are highly similar to those of EU/US-ranibizumab (Fig. 2D). Tertiary structure was also determined using H/DX-MS. The dynamics of deuterium uptake over time (10 s to 4 h) revealed high symmetry between SB11 and EU/US-ranibizumab (Fig. 2E, F).

The purity of the products was analyzed via product monomer content and amount of intact Fab, heavy chain, and light chain determined by SE-HPLC and CE-SDS, respectively. In SE-HPLC analysis, all samples showed prominent monomer peaks (Fig. 3A, B). The level of high molecular weight (%HMW) species of SB11 and EU/US-ranibizumab was comparably low (Table 3). The results were confirmed by orthogonal analyses using size exclusion chromatography-coupled multi-angle light scattering (SEC-MALS) and sedimentation velocity analytical ultracentrifugation (SV-AUC) (data not shown). In CE-SDS analysis, the results showed that the electrophoretic profiles of SB11 and EU/US-ranibizumab were similar (Fig. 3C, D). The level of intact Fab species (%Main for non-reduced CE-SDS), heavy chain (%Main 2 for reduced CE-SDS), and light chain (%Main 1 for reduced CE-SDS) in SB11 were similar to those of EU/US-ranibizumab (Table 3).

To assess charge variants of SB11 and EU/US-ranibizumab, icIEF analysis was performed. The results of icIEF showed that the isoelectric point (pI) values of all samples were identical and their electropherograms were similar (Fig. 3E, F). The results for %Main, %Acidic, and %Basic of all samples were comparable (Table 3). These results have been confirmed by orthogonal analytical method which is cation exchange-high performance liquid chromatography (CEX-HPLC) (data not shown).

As MoA-related biological activities of ranibizumab reflect physiological and pathological conditions, the VEGF-A 165 binding, VEGF-A 165 neutralization assay and anti-proliferation assay were performed. Additionally, binding activities for VEGF-A isoforms, excluding the VEGF-A 165, were evaluated to demonstrate similarity between SB11 and EU/US-ranibizumab.

Reference product lots with different expiry dates within the approved shelf life were used for establishing similarity ranges (mean ± 3 SD). SB11 drug products manufactured at commercial scale were compared with EU-ranibizumab and US-ranibizumab for the biological activities. Evaluation of the VEGF-A 165 binding activity by ELISA showed the mean %relative binding activity of SB11, EU-ranibizumab, and US-ranibizumab were 98%, 100%, and 100%, respectively. No meaningful differences between SB11, and EU/US-ranibizumab were observed and VEGF-A 165 binding activities of SB11 were within the similarity range (Fig. 4A). The relative VEGF-A 165 neutralization potency of SB11 was similar to that of EU/US-ranibizumab. Mean %relative potencies of SB11, EU-ranibizumab, and US-ranibizumab were 99%, 100%, and 101%, respectively. In addition, VEGF-A 165 neutralization potencies of SB11 were within predetermined mean ± 3 SD of similarity range (Fig. 4A). For the HUVEC anti-proliferation assay, the mean values of %relative anti-proliferation potencies of SB11, EU-ranibizumab, and US-ranibizumab were 100%, 101%, and 102%, respectively. All SB11 batches also showed relative potencies within the similarity range, demonstrating promising similarity in HUVEC anti-proliferation potency (Fig. 4A). These results support that SB11 is similar to EU/US-ranibizumab in terms of MoA-related biological activities.

Additional biological activities and VEGF-A 110, VEGF-A 121, and VEGF-A 189 binding activities were assessed of SB11 in a side-by-side comparison. Mean values of %relative VEGF-A 110 binding of SB11, EU-ranibizumab, and US-ranibizumab were 102%, 98%, and 101%, respectively, and those of %relative VEGF-A 121 binding activities were 101%, 103%, and 98%, respectively. In addition, mean values of %relative VEGF-A 189 binding activities of SB11, EU-ranibizumab, and US-ranibizumab were 99%, 97%, and 96%, respectively. No differences were observed between SB11 and EU/US-ranibizumab in the side-by-side comparison; relative VEGF-A 110, VEGF-A 121, and VEGF-A 189 binding activities of SB11 were similar to those of EU/US-ranibizumab (Fig. 4B).