By Flynn Watson
Proline is typically the most abundant amino acid present in wine however, it has historically been overlooked as it’s not consumed by yeast. Proline itself is known to be sweet though it’s impact in wine has been scarcely studied despite it being present at above-threshold levels in some wines. In this study, we developed an automated processing workflow to identify and subsequently quantify the proline multiplet seen at 2.8 – 2.4 ppm in the wine 1H spectrum. The script, built in Python 3, utilizes the water and ethanol suppressed spectra generated using the noesygpps1d pulse sequence and does not require any expert interpretation. The NMR method was compared against the traditional HPLC quantitation method (N = 28) in addition to standard analytical checks of recovery and reproducibility. The validated workflow was then used to screen 25+ wines and rapidly select three blending components: low, moderate or high levels of proline. These components were then blended in duplicate in a ternary experimental design, subjected to sensory and chemical analysis as well as consumer acceptance testing.