Therapeutic proteins are produced by means of growing living cell using culture media in a bioreactor. Quantitative and qualitative characterization of these highly complex biogenic broths at all stages of production is of major importance in the BioPharma industry. Compositional and structural analysis of cell culture media and proteins therein often require complicated analytical methods (usually chromatographic based) and extensive sample handling. Multi-dimensional fluorescence (MDF) spectroscopy techniques have been successfully used for routine determination of compositional changes in cell culture media and bioprocess broths as they offer rapid and inexpensive analysis.1 However, conventional MDF cannot always discriminate signals from fluorophores and in particular it can be difficult to observe emission from biomolecules in a sample with small molecule fluorophores (e.g. glycoprotein emission from media component emission).
By incorporating an anisotropy measurement with MDF we produced new 4-D methods (λex × λem × I × r) that were used for fast and non-destructive protein quantification in highly complex mixtures.2 The use of factor-based chemometric data analysis could also be used to resolve the emission of individual or groups of fluorophores in the protein structure based on their anisotropy.3 Here we show how ARMES data (fig.1) can be used to provide domain-specific structural information from two structurally very similar proteins (human and bovine serum albumins, HSA and BSA).
HSA and BSA were subjected to different thermal and chemical stresses and ARMES data was collected using Total Synchronous Fluorescence Spectroscopy (TSFS) as the MDF method. Multivariate Curve Resolution (MCR) was used for spectral deconvolution of the data. The HSA components extracted (fig.1) were: Trp fluorescence (Comp.4), two Tyr groups (Comp. 1&2) and Trp room temperature phosphorescence (Comp.3). The two Tyr components represent different hydrophilic and hydrophobic environments. Although structurally similar, HSA and BSA show different ARMES patterns and different number of components due to the second Trp in the BSA structure.3 ARMES data provides information on the degree of loss in helicity and solvent exposure for different domains, which might correlate with β-sheet formation. The changes in ARMES data for HSA/BSA show interesting differences which are attributable to a combination of photophysical and structural factors. ARMES is a powerful new tool that combines multi-dimensional anisotropy with chemometric data analysis to give a unique, fast, and non-destructive approach to protein structural analysis, without the need of protein labelling or extensive and intricate sample handling methods.
This research was supported by an Irish Research Council postgraduate award to RCG and the loan of a fluorescence spectrometer by Agilent Technologies.