Quality control in SARS-CoV-2 RBD-Fc vaccine production using LCMS to confirm strain selection and detect … – Nature.com

In vaccine production, QC is an important step to confirm identities of raw materials, relevant ingredients and final products. They are also screened for unwanted contaminants within the QC processes25,26. In case of SARS-CoV-2 vaccine production, the presence of other viral strains in the drug substances and/or end products would lead to failure of product tests and may cause a drop in efficacy and/or increase the risk of adverse events27. Therefore, establishing a suitable method for checking authenticity of the SARS-CoV-2 strain selected for vaccine production and detecting contaminants from other strains is essential for vaccine quality assessment. LCMS technique is widely known for its sensitivity, robustness, accuracy and high throughput28. It has a benefit over an enzyme-linked immunosorbent assay (ELISA) because it can detect several peptides from one or multiple viral strains in a single run29. LCMS method has been broadly used in clinical research to identify SARS-CoV-2 variants, but surprisingly its application in manufacturing process of biological products has less been found30. This study demonstrated that LCMS can be a useful tool for evaluating product quality in biopharmaceutical production.

Wuhan and early developed SAR-CoV-2 strains, for example Alpha, Beta, Gamma, Delta and Epsilon show little variation among their RBD amino acid sequences, demonstrating their close relationships. However, more recently emerging Omicron strains show 17 amino acids different from Wuhan wild type, indicating substantial evolution of Omicron lineages (Fig.1). All distinct amino acids detected among 14 SARS-CoV-2 strains contribute to nine characteristic peptides upon tryptic digestion. Most of the characteristic peptides can be readily detected in LCMS analysis but unmodified peptide 6 is hardly detected. The sequence between 150D and 192R positions (43 amino acids in length) has neither arginine (R) nor lysine (K), digestive sites of trypsin enzyme, causing a lengthy peptide after digestion. The large peptide could become low hydrophilicity, poor ionization, limited fragmentation and could comprise multiple charge states, resulting in low or no signal in LCMS analysis31,32. In the RBD-Fc sequences of different SARS-CoV-2 strains, including Delta, Iota, Theta and Eta strains, T161 or E167 position is substituted with K, yielding digestible peptides, which are useful for differentiating these strains from the others and from each other (Fig.3). Peptides 1, 2, 3, 4 and 7 are applicable for detecting Omicron BA1 and BA2. Peptide 3 is helpful for identifying Beta and Gamma strains. Peptides 8 and 9 are required for identifying Zeta and Alpha strains, respectively. Peptide 5 is unique for Lambda strain and applicable for distinguishing Kappa and Epsilon strains from the others. When all characteristic peptides were used together (Fig.3) and the identification processes were followed step by step from the top to the bottom, identification of SAR-CoV-2 strains can be achieved in a single attempt, even from a mix of recombinant RBD-Fc proteins derived from different viral strains (Fig.4).

However, Kappa and Epsilon strains cannot be differentiated from each other with this LCMS method because they have only one amino-acid difference, which does not contribute to a characteristic peptide after tryptic digestion. To overcome this issue, GluC protease could be an alternative enzyme for using alone or together with trypsin enzyme to additionally cleave the proteins at C-terminus of glutamic acid (E). GluC enzyme could facilitate the differentiation of SARS-CoV-2 RBD-Fc proteins derived from Epsilon and Kappa strain because the sequence of Epsilon strain contains glutamic acid at position 167 (E167), which is cleavable by GluC enzyme. However, this amino acid is mutated to glutamine (Q167) for Kappa strain. After GluC digestion, the resulting peptides would be different between the samples derived from Epsilon and Kappa strains and therefore distinguishable by LCMS analysis. Apart from GluC enzyme, other proteases, such as LysC, ArgC and AspN enzymes, could be additionally applied to increase completeness of protein digestion at lysine (K) and arginine (R) positions or additionally cleave the peptide at aspartic acid (D) position to increase number of resulting peptides. This could lead to better detection and differentiation of vaccine products derived from closely related SARS-CoV-2 strains. However, time and cost would be factors of concern as increased number of enzymes will double digestion time and chemical use.

Post-translational modifications of N-glycosylation could be another factor, affecting LCMS results and analytical performance to confirm protein identity and detect contaminations. In this study, characteristic peptide 1 has one glycosylation site at N26 amino acid. Sometimes, MS/MS signal of the non-glycosylated form was not observed. Peptide 1 is important for differentiating RBD-Fc proteins derived from Omicron BA1 and BA2 from the others. To improve the MS/MS signal of peptide 1 and abate the effects of N-glycosylation, protease enzymes, such as Endo-S and PNGase A, could be additionally used to release N-glycans from the core protein structure. Nonetheless, other characteristic peptides, including peptides 2, 3, 4 and 7, are appliable for distinguishing Omicron BA1 and BA2 strains from each other and from the others. Hence, Endo-S and PNGase A enzymes might not be necessary in this study, but they would benefit the other studies, discovering that only N-glycosylated peptide is a characteristic peptide.

In typical LCMS analysis, the resulting peptides are matched to the reference sequence of individual proteins. For example, the result from Wuhan SARS-CoV-2 RBD-Fc sample is usually compared against the reference amino acid sequence of Wuhan strain only (Fig.2). Identity of Wuhan strain might be confirmed but potential contaminants from other strains would not be identified. By having the peptide key alongside the reference of target protein, purity of intermediate substances and vaccine products can be more ascertained. Generally, the limit of LCMS detection is in a range of 110ngml1 solution or ng mg1 protein33,34. Within this study, a protein spike-in assay was performed to examine the limits of peptide detection. Among nine characteristic peptides, the lowest points of peptide detection at MS/MS level were around 5ngml1, observed from peptide 5 and 8 (Table 1). These two peptides are unique for detecting RBD-Fc derived from Lambda, Zeta, Epsilon or Kappa strains. Peptide 3 for detecting RBD-Fc derived Gamma strain and peptide 6.1 for detecting RBD-Fc derived from Iota strain were detectable at concentration around 20ngml1, demonstrating relatively sensitive detection by this developed LCMS method. Detection limits of other peptides were in a range of 50200ngml1, which would be sufficient for detecting even small contaminations in our vaccine production. Contaminants from other strains likely occur at microgram level since the tested materials are typically prepared at 1mgml1.

According to the WHO guideline, cross-contamination of biological materials at any stage during biopharmaceutical manufacturing is a risk and must be assessed and controlled19. In our production system, biological materials can be derived from different SARS-CoV-2 strains, therefore appropriate logistic plans and activities, such as a clear separation of storage areas between different bacteria cell banks and avoiding infiltrating different strains at the same time, must be exercised to reduce the risk. These processes have already been implemented in the manufacturing plant of Baiya Phytopharm. An implementation of LCMS analysis in the QC process will ensure the purity of starting materials and final vaccine products. Furthermore, molecular techniques, such as real-time PCR, could be further developed to check contaminations of different SARS-CoV-2 strains in bacterial clones and DNA plasmids during cloning and transformation steps. It could also be used to track down microbial contaminations and host cell DNA residues in the final vaccine products35. By using bioanalytical and molecular techniques in combination, QC analysis in biologic manufacture would be more forceful to ensure production integrity and product quality.

Continued here:

Quality control in SARS-CoV-2 RBD-Fc vaccine production using LCMS to confirm strain selection and detect ... - Nature.com