Abstract
D-tagatose is a high-value, low-calorie sweetener that can be produced from dairy lactose via a two-step enzymatic route: lactose hydrolysis to galactose followed by galactose isomerization to tagatose. Here, we combined genomics, in silico structural analysis, and biochemical assays to evaluate the lactose-to-tagatose conversion potential of an Antarctic isolate, L47, identified as Ewingella americana (NCBI accession SAMN54554459). Genome mining revealed one L-arabinose isomerase gene (araA) and three β-galactosidase genes (bgaA, bglY, lacZ), an uncommon combination in a single bacterium. Recombinant AraA was produced in Escherichia coli and biochemically characterized, showing Mn(2+) dependence and measurable D-galactose isomerization, reaching ~18% tagatose from 100 mM galactose after 48 h under the tested conditions. In contrast, the β-galactosidases were predominantly recovered as insoluble aggregates in E. coli; therefore, β-galactosidase activity was assessed using washed inclusion-body preparations. Under these conditions, BgaA displayed the most consistent o-NPG hydrolyzing activity, whereas BglY and LacZ did not yield reproducible activity. Overall, our results identify BgaA as the most tractable lactose-hydrolyzing candidate from L47 in the current workflow and indicate that AraA performance is the principal bottleneck toward an efficient lactose-to-tagatose process, motivating future optimization at the enzyme and process levels.