As biopharmaceutical patents start to expire there is a burgeoning opportunity for the launch of biosimilar pharmaceuticals.
Clearly, biopharmaceuticals are far more complex than the single molecule entities found in traditional pharmaceuticals, and this complexity raises the potential for there to be errors in their production that might prove harmful to patients. Amino acid substitutions, phosphorylation or glycolysation at unintended sites, or deamidation are just some of the potential 'mutations' that need to be looked for and avoided.
The complexity of proteins means that a number of techniques have to be used to characterise them properly. For full characterisation of a protein, the protein's primary, secondary and tertiary structure as well as its physiochemical properties should be assessed.
Routine methods such as SDS-PAGE can be used to gain an idea of purity, the presence of truncated forms and the degree of aggregation. However, they don't give details on the exact amino acid sequence, what post-translational modifications may be present or any details on the secondary or tertiary structure.
Mass spectrometry and protein characterisation
Mass spectrometric techniques are very powerful for the identification and characterisation of proteins. At a basic level, mass spectrometry determines an accurate mass of a protein (often to ≤1 Da), and so provides a quick and easy way to determine if a protein is modified. Mass spectrometry will therefore pick up differences in protein forms that will not be resolved by low-resolution techniques such as SDS-PAGE.
Simple proteomic techniques involving digestion of the protein(s) with a sequencing grade enzyme (e.g. trypsin), followed by LC-MS-MS (coupling of a HPLC to an electrospray mass spectrometer) of the resulting peptide mixture produces vast amounts of additional, useful data.
Fortunately, peptides behave in a very convenient and predictable way in a mass spectrometer, with amino acids falling off sequentially. This means that when the fragmented peptide data is compared against the details held on protein sequence database such as SwissProt, the sequence of the protein may be determined.
The same data will also confirm protein identity and any contaminating proteins in the sample. This approach should also give up to 80% confirmation of the protein sequence.
Mass spectrometry can be used for further investigation of the protein. For example, it will identify the sites of a vast array of different types of modification, such as phosphorylation, oxidation and deamidation. By performing a database search using commonly available search engines, one can tailor the search to look for specific issues, such as oxidised methionine residues. The search software is 'intelligent' enough to consider both the possibility of oxidised methionine as well as unmodified methionine when comparing the sequences to that in a database. Such techniques are essential when looking at the discrete differences that may exist between biobetters and biosimilars.
Protein characterisation by other techniques
Mass spectrometry does not allow full characterisation of a protein. For full characterisation, orthogonal complementary techniques are required, including relatively old techniques such as Edman degradation.
The hydrolytic cleavage of a protein into its constituent amino acids followed by ion exchange chromatography is used to determine the quantity of each amino acid in the sample. Once the amino acid composition is known, UV/visible spectrophotometry can be used to determine the extinction coefficient. This technique may indicate, for example, an amino acid substitution that the standard mass spectrometry techniques described above, will not highlight. Thereafter, mass spectrometric data could mined manually to locate such a substitution.
Edman degradation is still the best way to identify the N-terminal sequence of a protein. Though the N- and C-terminal sequence can be determined by mass spectrometry peptide mapping, the database search engines used for analysis are not primed to detect unexpected modifications or substitutions. Hence, these may be missed without manual data mining.
Protein aggregation is a common problem in the production of biopharmaceutical proteins. Techniques such as size-exclusion chromatography, analytical ultrafiltration and field-flow fractionation may be employed for aggregate analysis.
The growing opportunities for biobetters and biosimilars, means there is an increasing demand for protein characterisation, both to protect patient safety and to meet regulatory requirements. In addition, protein characterisation can be used to help solve expensive production problems, such as excessive aggregation formation, or in a trouble-shooting role, such as the identification of contaminating protein species.