A monoclonal antibody biotherapeutic was supplied in two different formulations. Far UV DMS measurements were made using a Chirascan-plus CD spectrometer, and subsequent analyses of the data were carried out using Global 3, a global analysis software program that is an integral part of the DMS technique. Each DMS data set was obtained in under 100 minutes and used 65 µg of protein.
Robust Tm values were calculated from the CD temperature profiles by global analysis. Significant differences in the mid-points of the first and third transitions were observed; the mid-points of the second transitions were the same within experimental error. The first transition was lower in the lactate than in the acetate; the third transition was the most significant, best defined in both acetate and lactate buffers, and higher in the lactate formulation by 1.8ºC. There was significant variation in the individually calculated van’t Hoff enthalpies, an inevitable consequence of the degree of overlap of transitions in these complex systems.
The samples were measured several times over a period of weeks, and the CD spectra of the DMS data was used to monitor the secondary structures. It became apparent that the antibody behaved differently as it aged in the two formulations. Both formulations maintained the antibody in its mainly b-sheet room temperature conformation as they aged, and initially the conformational changes on heating were similar, with each formulation taking a virtually identical structural pathway to a recognizable unfolded state Figure 1, top.
As the samples aged, the denaturation pathway in acetate changed dramatically Figure 1, bottom left, butit remained virtually unchanged in lactate Figure 1, bottom right, which suggests lactate imparts greater stability.
Examination of the corresponding absorption profiles of the two formulations, recorded simultaneously with the CD data, showed that the aggregation-onset temperature for the antibody in the acetate formulation decreased over time Figure 2, top, whereas the antibody in the lactate formulation showed no tendency to aggregate with time Figure 2, bottom. The onset of aggregation was indicated by an apparent increase in absorption caused by light scattering at wavelengths where there is no chromophore.
The results suggest that the antibody is more stable in the lactate than in the acetate formulation despite its significantly lower Tm1, which if taken in isolation might suggest the contrary. It is speculated that Tm measurements in isolation may not be a particularly good indicator of stability and that resistance to aggregation, measured by DMS, is an important criterion to consider.
Using Chirascan-plus DMS Figure 3 for the testing of two different formulations it was possible to show that a biotherapeutic antibody was more stable in a lactate formulation than in an acetate formulation.
The CD spectra of the DMS data was used to identify changes in the secondary structure of the antibody. Both acetate and lactate formulations maintained the antibody in its mainly b-sheet conformation as they aged at low temperature and, initially, the conformation changes on heating were similar, with each formulation taking a virtually identical structural pathway to a recognizable unfolded state.
As the samples aged, the denaturation pathway of the antibody in acetate buffer changed dramatically, but there was no such change for the antibody in lactate buffer, whose denaturation pathway was identical to that of the initial measurement, suggesting greater stability.