Application of Laser-Based Mid-Infrared Spectroscopy for Inline Monitoring of Proteins from Preparative Liquid Chromatography

Abstract

Laser-based mid-infrared (IR) spectroscopy is an emerging technique for analyzing proteins in aqueous solutions. Novel spectrometers based on external cavity-quantum cascade lasers (EC-QCLs) offer significant advantages compared to gold-standard Fourier-transform IR (FTIR) instrumentation such as higher sensitivity, larger applicable optical path-lengths and increased ruggedness. These advantages open a wide range of possible applications, including measurements of proteins from complex purification operations. In this work, a commercial EC-QCL based mid-IR spectrometer was applied for inline monitoring of proteins from preparative liquid chromatography (LC). The large optical path length (25 µm) of the equipped transmission cell and the broad tuning range of the laser (1350-1750 cm-1) enabled robust spectra acquisition in the most important wavenumber region for protein secondary structure analysis. To demonstrate the advantages of QCL-IR spectroscopy over conventional LC detectors, two different purification operations based on ion-exchange chromatography (IEX) and size exclusion chromatography (SEC) were monitored. In IEX, a major challenge was caused by the applied sodium chloride gradient, inducing mid-IR absorbance bands that overlap with protein bands and dominate to recorded spectra. Here, a novel background compensation approach was implemented to eliminate salt bands, leading to high quality protein spectra. In case of SEC, proteins with similar molecular weights, leading to overlapping chromatographic peaks that cannot be distinguished with conventional UV detectors were monitored. Here, the combination of laser-based mid-IR spectroscopy and chemometrics allowed estimation of individual protein concentrations across the chromatographic run based on their secondary structure. For both systems, the obtained results were compared to reference high-performance LC (HPLC) offline measurements, showing excellent agreement in terms of protein identification as well as quantification. Consequently, QCL-IR spectroscopy can be successfully applied for inline detection of proteins from LC effluents, providing information that is typically only accessible by time and cost-intensive offline methods.