Alditols, which have a sweet taste but produce much lower
calories
than natural sugars, are widely used as artificial sweeteners. Alditols
are the reduced forms of monosaccharide aldoses, and different alditols
are diastereomers or epimers of each other and direct and rapid identification
by conventional methods is difficult. Nanopores, which are emerging
single-molecule sensors with exceptional resolution when engineered
appropriately, are useful for the recognition of diastereomers and
epimers. In this work, direct distinguishing of alditols corresponding
to all 15 monosaccharide aldoses was achieved by a boronic acid-appended
hetero-octameric Mycobacterium smegmatis porin A (MspA) nanopore (MspA-PBA). Thirteen alditols including
glycerol, erythritol, threitol, adonitol, arabitol, xylitol, mannitol,
sorbitol, allitol, dulcitol, iditol, talitol, and gulitol (l-sorbitol) could be fully distinguished, and their sensing features
constitute a complete nanopore alditol database. To automate event
classification, a custom machine-learning algorithm was developed
and delivered a 99.9% validation accuracy. This strategy was also
used to identify alditol components in commercially available “zero-sugar”
drinks and healthcare products, suggesting their use in rapid and
sensitive quality control for the food and medical industry.