Simulation of absorption spectra of native and unfolded proteins

Protein concentrations are routinely determined using the absorbance measured at 280 nm and Beer-Lambert's law. However, traditional single-wavelength approaches may be inferior to a multi-wavelength analysis of complete spectra given the larger amount of data that can be processed. Hence, the current study was aimed at simulating protein UV spectra (from 250 to 350 nm) with a view to more accurately estimate protein concentrations. We demonstrate that the spectra of unfolded proteins are well-simulated using primary sequence data and the wavelength-dependent molar absorption coefficients of l-cystine, l-phenylalanine, N-acetyl-l-tyrosinamide and N-acetyl-l-tryptophanamide (the latter two serving as L-Tyr and L-Trp model compounds). Alternatively, simulations can be performed with the coefficients of the Tyr and Trp mimics replaced by a pseudo-Voigt (pV) function, which mathematically fully describes the spectra of these model compounds. Furthermore, a pV function, generated from the analysis of the spectra of 14 proteins, can be utilized to simulate the spectral contributions of Tyr and Trp in native proteins with a reasonable degree of accuracy. A Microsoft Excel-based multi-wavelength fitting routine can be employed to simulate the spectra of proteins and compare them with those experimentally recorded, thereby facilitating the determination of protein concentrations.