Reading In-Between Spectra: Exploiting Laser-Based Mid-Infrared Spectroscopy with Chemometrics As A Tool to Study Continuous Unfolding of Proteins

Abstract

As protein function depends on in its secondary structure, several industries including the pharmaceutical industry rely on analytical techniques such as circular dichroism (CD) spectroscopy for sensitive and dynamic detection of secondary structure changes (denaturation) for quality control. The drawbacks of CD spectroscopy such as high absorption of buffers, low sensitivity to ß-sheet structures and a limited concentration range necessitate alternative methods that are compatible with pharmaceutical formulations that often contain various buffers and high protein concentrations. Fourier transform infrared spectroscopy (FTIR) while being capable of accessing higher protein concentration ranges and being sensitive to different secondary structures, suffers from its own nemesis – water. To avoid total absorption, path lengths used in FTIR spectroscopy must hence be capped, consequently imposing challenges in protein denaturation studies that demand longer path lengths to avoid clogging. Tunable quantum cascade lasers (EC-QCLs) when used as light sources in mid-infrared spectroscopy can provide this extended path length while retaining the benefits of FTIR spectroscopy. This presentation demonstrates the use of EC-QCL spectroscopy as a tool to study the continuous denaturation of proteins effected through two impulses – heat and chemical agents. These impulses themselves have spectral contributions, complicating the retrieval of protein secondary structure information. Chemometrics offers solutions to this in the form of algorithms, such as multivariate curve resolution alternating least squares (MCR-ALS), with the capacity to decompose complex matrices into component-relevant information. Applied to the thermal denaturation of bovine serum albumin (BSA), MCR-ALS indicates the formation of not one, but two intermediate structures, that are overlooked without the application of chemometrics. In contrast, the difficulty in studying the surfactant-induced unfolding and refolding of proteins lies in the highly interfering surfactant contributions. Two approaches were adopted to combat this; the use of a stand-alone MCR-ALS and the use of an automated baseline correction employing partial least squares regression (PLSR) to model the surfactant contributions, followed by MCR-ALS. EC-QCL spectroscopy conjugated to chemometrics in both applications open promising avenues for automation and integrability to industrial processes as a consequence of the ruggedness and stability of such setups.