High‐throughput assessment of antigen conformational stability by ultraviolet absorption spectroscopy and its application to excipient screening

Abstract

In high-throughput screening (HTS) assays, the use of ultraviolet absorption spectroscopy (UA) is commonly limited to concentration and turbidity measurements. Our aim was to evaluate microplate-based UA and its second-derivative [(2d)UA] for measuring the conformational stability of two recombinant antigenic proteins in the presence of 44 excipients. Protein conformational stability was assessed by (2d)UA upon titration with guanidine hydrochloride. (2d)UA was compared with tryptophan fluorescence spectroscopy (TF) and differential scanning fluorimetry (DSF), both commonly used techniques for measuring protein conformational stability. The HTS data were corrected for plate, row and column effects by applying a median polish procedure. Irrespective of the unfolding method applied, similar stabilizing excipients were identified by all analytical methods for a given antigen. The native forms of both antigens were destabilized by arginine, hydroxypropyl-β-cyclodextrin, and sodium docusate, and were protected by polyols. The median polish correction improved the quality of the prediction models and the screening resolution. The higher sensitivities of TF and DSF compared with (2d)UA allowed the identification of a larger number of stabilizing excipients. However, similar screening resolutions (z'-factor > 0.8) were observed for 2dUA, TF, and DSF in a HTS of excipients applied to one of the antigens. Therefore, (2d)UA deserves more attention in HTS studies focused on protein conformational stability.

Development of a High-Throughput Screening Platform to Study the Adsorption of Antigens onto Aluminum-Containing Adjuvants

“…The degree of antigen recovery in aluminum‐free formulations was at least 88% for both concentrations, and the median values were 95% for β‐casein and 93% for BSA (Supplementary Table 1). The UV–Visible absorbance at 320 nm was determined to evaluate potential particle formation . For both BSA and β‐casein, low signals were recorded (Supplementary Table 1).…”

“…The pH values of the adsorbed formulations were determined using a Cavro Mini Sample Processor (Männedorf, Switzerland) with Gemini software that was customized for pH measurements. The optical density at 320 nm of the supernatant of adjuvant‐free (protein‐only) solutions was measured in the nonabsorbing regions of the absorption spectra to monitor the presence of aggregates, using UV‐transparent microplates and a Varioskan Flash microplate reader . The blank signal was subtracted for each protein formulation.…”

“…The optical density at 320 nm of the supernatant of adjuvant-free (protein-only) solutions was measured in the nonabsorbing regions of the absorption spectra to monitor the presence of aggregates, using UV-transparent microplates and a Varioskan Flash microplate reader. 23 The blank signal was subtracted for each protein formulation.…”

“…The UV-Visible absorbance at 320 nm was determined to evaluate potential particle formation. 23 For both BSA and $-casein, low signals were recorded (Supplementary Table 1). High-protein recoveries and negligible particle formation suggested that no major aggregation/precipitation occurred at the protein concentration and volume ranges examined.…”

Lessons learned in building high‐throughput process development capabilities

“…The use of HTPD in formulation is less advanced when compared to purification at Merck & Co. and is still being developed. There are limited descriptions of formulation HTPD in the scientific literature and the main focus is in evaluating HT assays that are relevant to formulation development [22][23][24][25][26][27][28][29][30]…”