Sequence variants (SVs) in proteins can affect the quality, safety, and efficacy of biologic drugs, as well as challenge standard purification processes. It is crucial to identify SVs within proteins, as they can lead to misfolding, aggregation, and fragmentation. But detecting and quantifying these SVs are complex due to limitations in existing analytical techniques. To address this, Samsung Biologics has developed a high-resolution method that can detect both genetic and non-genetic variants at low levels within a significantly reduced timeline. This innovative approach has been successfully applied during cell line and upstream process development, ensuring greater efficiency with higher quality in biologic drug production.
Risks Posed by Sequence Variants
The functionality and efficacy of antibodies and other therapeutic proteins are significantly influenced by their structure, which is mainly determined by the amino acid sequence. Errors in protein biosynthesis that lead to sequence variants (SVs) can result in protein misfolding, aggregation, and fragmentation. These variants typically arise from amino acid substitutions due to genetic mutations during cell line development (CLD) or amino acid misincorporations during upstream processing (USP).
DNA transcription into messenger RNA and subsequent protein translation are critical phases where mutations can impact both gene and protein levels. During cell line development, genetic mutations may arise during transient transfection and DNA replication. These mutations can persist through clone selection, affecting subsequent development.
Furthermore, during USP, multiple mechanisms contribute to SVs at the protein level. The dense cell culture environment can exacerbate cellular stress, increasing the likelihood of transfer RNA mispairing or ribosomal inaccuracies and leading to the incorporation of incorrect amino acids during protein synthesis. Similarly, a deficiency of certain key nutrients in the media can prompt the translation machinery to substitute an available amino acid for one that is limited or absent, thereby altering the amino acid sequence of the protein.
Although sequence variants generally constitute less than 0.1% of biotherapeutic molecules, the push for increased titers and productivity can intensify stress on cell lines, potentially leading to production errors. Complex proteins, such as bispecific antibodies, fusion proteins, and other advanced modalities, are particularly susceptible to a higher incidence of SVs. This increased risk of SVs may largely be due to the inclusion of repetitive sequences within flexible linker regions, which are susceptible to duplications or deletions via homologous recombination at the repeated sequences.
Proteins with incorrect sequences pose significant challenges in removal through standard purification processes, impacting the quality, safety, and efficacy of biologic drugs. While next-generation sequencing can detect and quantify SVs in cell lines, it is restricted to identifying mutations at the DNA level and does not extend to critical non-genetic variants.
LC-MS/MS for Sequence Variance Analysis (SVA)
Liquid chromatography–tandem mass spectrometry (LC-MS/MS) is a powerful tool for detecting and quantifying genetic as well as non-genetic protein sequence variants during cell line and upstream process development. This method identifies SVs by comparing observed amino acid sequences against their theoretical expectations.
However, conventional LC-MS/MS techniques often yield low-resolution results, generating a vast array of potential sequence variants, as many as 1,000 in one analysis. This necessitates extensive and meticulous data analysis, typically performed by skilled analysts, to distinguish genuine variants from false positives caused by misidentification.1
Accelerating SVA with an Advanced LC-MS/MS Method
To address the limitations of traditional analysis in detecting and quantifying SVs, Samsung Biologics developed a cutting-edge LC-MS/MS methodology. This advanced method incorporates ultra-high-performance liquid chromatography (UHPLC) for rapid separations — coupled with a high-resolution Orbitrap mass spectrometer for detection — and uses a comprehensive protein characterization software for detailed analysis. The synergy between these high-end MS acquisition and data processing technologies allows for more precise and efficient detection of SVs, even at very low levels.
Use of NIST mAb, a commercially available protein with documented SVs, has shown results that align with established literature values. For example, the detection of a glycine (G)-to--aspartic acid (D) variant in NIST mAb at a level as low as 0.01% demonstrates the method's capability to identify minute levels of SVs with significant accuracy, as illustrated in the charts within Figure 1.
▲ Comparison of sequence variants (%) between Samsung Biologics’ method and a reference method2 for NIST mAb
▲ MS/MS spectra of VYACEVTHQGLSSPVTK peptide containing G199D variant in NIST mAb
Figure 1. Selected SVA results obtained for NIST mAb using Samsung Biologics’ new LC-MS/MS technique
The workflow for SVA is akin to that used in peptide mapping, beginning with the standard steps of reduction, alkylation, and protease digestion. The acquired LC-MS/MS data are meticulously processed to delineate between wild-type and variant peptides. This is done by comparing retention times and the MS/MS spectra and measuring parts per million (ppm) errors.
A distinctive aspect of this method compared with traditional peptide mapping is the application of multiple proteases to facilitate orthogonal digestions, which enhances the reliability of detecting SVs. While trypsin is typically used to cleave proteins at lysine (K) and arginine (R) residues, yielding positively charged peptides conducive to mass spectrometry detection, its use can also create long peptides with ionization challenges. Introducing additional proteases mitigates this issue by producing multiple data sets, which helps confirm the presence of both real and false positive variants.
The protein characterization software includes a variety of validator algorithms to filter out false positive results for modifications, such as SVs. Samsung Biologics adapted this validator algorithm to suit the equipment and analysis conditions for SVA. False positives can be intuitively identified, allowing experienced analysts to quickly and accurately inspect SVA results.
By optimizing the workflow, the SVA timeline has been reduced from the typical 6–8 weeks to 1–2 weeks for analyzing two to six samples, significantly accelerating the process and reducing the risk of misidentification. This enhanced technique has been applied effectively during both CLD and USP development phases, including the analysis of clonal pools to identify superior clones as well as the assessment of media and nutrient feed strategies on the formation of SVs.
Case Study One: Detecting Mutations During Cell-Line Development
Samsung Biologics employs SVA during cell line development to ensure the selection of more stable cell lines that produce minimal undesired SVs. The first case study illustrates the detection of a significant mutation (sample A) and amino acid misincorporations (sample B).
In sample A, a critical mutation involved the substitution of asparagine (N) with histidine (H) at a specific site. The mutation was precisely identified using the advanced LC-MS/MS technique, which differentiated the two peptide sequences not only by their distinct masses but also through the optimized MS/MS conditions that allowed for full b and y ion coverage. This led to clear and unambiguous fragmentation of both peptides, as shown in Figure 2 (top panel). The SV was quantified at a concerning level of 6.0% by comparing the extracted ion chromatogram (XIC) areas of the wild-type and variant peptides. Notably, despite its high levels, this variant could have remained undetected in typical in-process testing owing to its physicochemical characteristics, which showcases the robustness of the new method and its ability to optimize quality.
Figure 2. Results for SV analysis performed during cell line development
For sample B, the analysis revealed multiple low-level misincorporations from threonine (T) to serine (S) at different sites. Interestingly, no variants, including other kinds of substitutions, were detected in those samples, suggesting that these misincorporations were likely due to nutrient depletion during cell culture. This finding was instrumental in advancing this cell line and provided critical data to optimize media composition during subsequent upstream process development.
Case Study Two: Detecting Variants During Upstream Processing
Substantial benefits of SVA have also been demonstrated throughout USP. The second case study highlights how the proportion of SVs increased as the project transitioned from cell line development to process development stages, as illustrated in Figure 3.
Notably, the detected amino acid misincorporations shifted exclusively from asparagine (N) to serine (S), indicating a depletion of asparagine in the media or feed. This specificity in misincorporation underscores the utility of SVA in pinpointing exact stages within the manufacturing process where errors might occur, enabling targeted interventions. The rapid application of SVA across various developmental stages enhances Samsung Biologics’ ability to detect and rectify these errors promptly, ensuring higher quality and consistency in biologic drug production.
Figure 3. SVA results at different manufacturing stages
Advantages of Streamlined SVA
Analytical development for biopharmaceutical products presents numerous challenges. Samsung Biologics adopts a tailored approach, understanding the specific needs of each client’s project and leveraging extensive experience across various compounds. Committed to innovation, Samsung Biologics develops advanced solutions that streamline development processes and provide crucial data to support informed decision-making.
The LC-MS/MS technique developed by Samsung Biologics for SVA offers significant advantages in mitigating risks associated with product development. This streamlined method effectively identifies both single-base mutations during CLD and amino acid misincorporations during both CLD and USP development. By integrating SVA at strategic points within the CLD and process development workflows, Samsung Biologics not only reduces development time and costs but also enhances the quality of the final product.
Optimized Analytical Services at Samsung Biologics
A team of highly skilled scientists with proven track record in delivering comprehensive, high-quality analytics for a wide variety of molecular types is critical in ensuring the success of biotherapeutic programs. Beyond utilizing platform solutions for standard monoclonal antibodies (mAbs), the team excels in developing, qualifying, and characterizing analytical methods for more complex modalities. Their extensive in-house portfolio of analytical techniques allows for a thorough understanding and assessment of target product profiles and critical quality attributes, all while reducing timelines and costs. This capability demonstrates product quality across cell line, USP, and DSP development, as well as clinical and commercial manufacturing stages.
References
1. Cadang, Lance et al. “A Highly Efficient Workflow for Detecting and Identifying Sequence Variants in Therapeutic Proteins with a High Resolution LC-MS/MS Method.” Molecules. 28: 3392 (2023).
2. Schiel, John E., Darryl L. Davis, and Oleg V. Borisov. State-of-the-Art and Emerging Technologies for Therapeutic Monoclonal Antibody Characterization Volume 2. Biopharmaceutical Characterization: The NISTmAb Case Study. 63–117 (2015).