The formulation of biopharmaceutical products is aimed, among other things, at eliminating or minimizing chemical instabilities, increasing the conformational and colloidal stability of proteins, and optimizing interfacial instability through the precise selection of excipients. Proteins, in our case monoclonal antibodies, can undergo complex degradation pathways and changes not only in primary but also in higher order structures. The field of excipients includes buffer substances, stabilizers, salts, antioxidants, and surfactants. In the past, the selection of these excipients in biopharmaceutical development relied on systematic testing and compatibility determination. However, in recent times, there is an increasing need for more rational approaches, which are highly desirable in a competitive environment due to their efficiency.
In the scope of my doctoral work, we identified alternative buffer systems that can present equally good, if not better, options in development compared to conventionally used buffers. We have demonstrated their effectiveness under recommended, accelerated, and stress storage conditions, other forms of stress (including light stress and freeze/thaw stress), as well as at low and high protein concentrations. We investigated the interactions between excipients and monoclonal antibodies as one of the proposed stabilization mechanisms. We employed 1D and 2D nuclear magnetic resonance (NMR) methods and molecular dynamics (MD) simulations. To ensure appropriate resolution of NMR measurements, we cleaved the model monoclonal antibody into Fab and Fc domains using the cysteine protease IgdE and purified the fragments using preparative affinity (ALC) and size exclusion (SEC) chromatography. We addressed the instability of the Fc fragment using conventional buffers and stabilizers, supported by MD simulations. The amide and methyl fingerprints obtained with 2D NMR methods were further supported by chemometric methods of principal component analysis (PCA) and combined chemical shift deviation (CCSD), showing that these methods are sensitive enough to identify interactions not only with buffers but also with polysorbates. The latter posed a greater challenge, presumably due to the weakness of the interaction, leading us to synthesize spin-labelled polysorbate (SLPS), which is structurally very similar to PS80 and PS20, commonly used surfactants in biopharmaceuticals. The interaction between PS80 and the model mAb was hypothesized to be likely due to the protective effect against the chemical hydrolysis of the surfactant observed in their coformulation. SLPS also allowed us to use the electron paramagnetic resonance (EPR) method and the effect of paramagnetic relaxation enhancement (PRE) in NMR measurements. Through both, we demonstrated that SLPS interacts with the model mAb and that the interaction is indeed weak, necessitating increased solution viscosity for the interaction to be detectable in EPR. Our findings were supported by MD simulations, indicating that the interaction is most likely hydrophobic in the hinge region of the mAb, as suggested by our stability data of PS in the presence of fragments. Lastly, we characterized the chemical hydrolysis of PS80 in histidine and acetate buffer, proposed a degradation mechanism, and showed that the plastic primary packaging significantly affects the breakdown. The results were supported by PCA.
Our findings add a new piece to the mosaic of understanding stabilization and degradation processes in biological formulations, representing a new step towards rational development of biological formulations.
|
