Tailoring the natural rare sugars D-tagatose and L-sorbose to produce novel functional carbohydrates

Rare sugars are scarce monosaccharides and derivatives with promising applications in food and health. Traditional Izumoring strategies enable interconversion of monosaccharides, while newer enzymatic approaches create novel disaccharides using rare sugars as acceptors. D-Tagatose, a GRAS, low-calorie sweetener with potential prebiotic effects, is substantially absorbed in the small intestine (≈80% in humans), limiting its colonic fermentation. The study hypothesizes that enzymatic fructosylation (via levansucrase SacB) of D-tagatose (and testing L-sorbose and D-psicose) can generate non-digestible oligosaccharides that resist small-intestinal digestion and better modulate the gut microbiota. The work aims to biosynthesize, structurally characterize (by NMR), and functionally assess (digestibility and in vitro fermentation) these novel rare sugar-based oligosaccharides, and to rationalize SacB acceptor specificity by molecular docking.

• Enzymatic synthesis of rare sugar derivatives has expanded beyond Izumoring, including aldose epimerization, enzymatic condensation, and phosphorylation–dephosphorylation cascades. Few reports exist on synthesizing disaccharides using rare sugars as acceptors (e.g., xylosylation of D-psicose; glucosylation of D-galactose). Fujita et al. reported L-sorbose transfructosylation by a β-fructofuranosidase, yielding a β-(2→2) disaccharide.
• D-Tagatose shares organoleptic properties with sucrose but has a lower glycemic index and caloric value; it is GRAS (FDA) and approved in multiple jurisdictions. However, human studies show ≈80% small-intestinal absorption, restricting its prebiotic reach.
• Levansucrases (GH68) like Bacillus subtilis SacB catalyze sucrose hydrolysis and β-(2,6) levan formation, and can transfer fructosyl units to alternative acceptors. Prior acceptor studies did not cover D-tagatose, L-sorbose, or D-psicose.
• Inulin-type FOS with β-(2→1) linkages are resistant to mammalian digestion and are established prebiotics, providing a comparative benchmark for fructosylated tagatose.

• Enzyme production and activity: SacB from Bacillus subtilis CECT 39 overproduced in E. coli and purified. Activities at 37 °C, pH 6.0 (50 mM potassium phosphate): total 20.77 U mg−1 (glucose release), fructosidase 10.79 U mg−1 (fructose release), transfructosylation 9.98 U mg−1 (fructose transfer).
• Enzymatic synthesis: Transfructosylation reactions using sucrose as donor and D-tagatose (primary), L-sorbose or D-psicose (tests) as acceptors at 37 °C, pH 6.0. SacB concentrations tested: 0.6, 3.1, 6.2 U mL−1; time-course up to 72 h with aliquots at 0.5–72 h. Substrate ratios for tagatose optimization: 200:200, 200:300, 200:450, 300:300 g L−1 (sucrose:tagatose). Enzyme inactivation at 100 °C for 5 min.
• Product analysis and purification: GC-FID of carbohydrates as TMS-oximes (DB-5HT column; oven 150→380 °C at 3 °C/min; N2 carrier; internal standard phenyl-β-glucoside). Preparative LC-RID (Zorbax NH2 PrepHT, 65:35 acetonitrile:water, 21 mL/min) to isolate fractions by degree of polymerization: tagatose T1–T3; sorbose S1–S4; fractions pooled, evaporated (<25 °C), and freeze-dried.
• NMR structural elucidation: 1D 1H/13C and 2D gCOSY, TOCSY, gHSQC (multiplicity-edited), gHMBC, and gHSQC-TOCSY at 298 K in D2O on 500/125 MHz system with HCN cryoprobe. Chemical shift referencing: HDO δH 4.79; external 1,4-dioxane δC 67.40.
• Molecular docking: AutoDock Vina using SacB structure (PDB 1PT2). Receptor modeled as covalent fructosyl–Asp86 intermediate; grid 25×25×25 Å, exhaustiveness 32, center at (41.5, 36.6, 13.0 Å). Ligands (D-tagatopyranose, L-sorbopyranose, D-psicopyranose) from PubChem, stereochemistry validated with Phenix eLBOW; validation by redocking sucrose. Analysis in PyMOL focusing on catalytic residues Asp86, Asp247, Glu342.
• In vitro digestion: Purified tagatose-based di- (T1) and trisaccharides (T3) incubated in simulated intestinal fluid with pig small intestinal BBMV (60 mg) at 37 °C; aliquots at 1, 2, 3 h; analysis by GC-FID. BBMV disaccharidase activities profiled with standard substrates.
• In vitro fermentation: Batch anaerobic fecal fermentations (two pools, n=3 donors/pool) at 37 °C with substrates at 0.5% w/v: β-D-Fru-(2→1)-D-Tag (from T1), unmodified tagatose, commercial FOS, and no-carbon control. Sampling at 0, 8, 24 h for 16S rRNA sequencing (V3–V4, Illumina MiSeq; QIIME2; SILVA 138) and SCFA/lactate quantification by GC-FID after ether extraction and MTBSTFA derivatization. Diversity metrics (alpha/beta), differential abundance (ANCOM, LEfSe, metagenomeSeq), correlation network (ccrepe, qgraph), and Pearson correlations between taxa and SCFAs.

• Acceptor specificity: SacB transferred fructosyl units from sucrose to D-tagatose and L-sorbose, producing oligosaccharides with DP2–DP5; no detectable transfer to D-psicose.
• Structures by NMR: • Tagatose series: Main disaccharide β-D-fructofuranosyl-(2→1)-D-tagatopyranose (two anomers, α:β ≈ 6:1). Trisaccharides included β-D-Fru-(2→6)-β-D-Fru-(2→1)-D-Tag (major) and minor species; a disaccharide β-D-Fru-(2→6)-D-Glc was also observed (from glucose acceptor). • Sorbose series: Main disaccharide β-D-Fru-(2→5)-α-L-sorbopyranose and elongated β-(2→6)-linked fructan-type tri-, tetra-, and pentasaccharides terminating at L-sorbose.
• Molecular docking: Single productive poses supported β-D-Fru transfer to C1 of D-tagatopyranose and to C5 of L-sorbopyranose, consistent with SacB’s catalytic mechanism (involvement of Glu342, Asp86, Arg246/Arg360). D-psicopyranose’s C5 stereochemistry misorients OH5, preventing productive catalysis, rationalizing the lack of transfer.
• Process optimization (tagatose): SacB 3.1 U mL−1 yielded the highest β-D-Fru-(2→1)-D-Tag levels (5× vs 6.2 U mL−1; 13× vs 0.6 U mL−1). Substrate ratio effects: maximum product concentrations ranged 27 g L−1 (200:200) to 64.04 g L−1 (300:300). Highest conversions: 21.5% (300:300) and 26.0% (200:450). Selected optimal: 300:300 g L−1 for highest concentration (64.04 g L−1) and yield (21.3%). Product formation peaked at 24 h and remained stable to 72 h across conditions.
• Digestibility (pig BBMV): Sucrose fully digested by 1 h. β-D-Fru-(2→1)-D-Tag remained essentially intact (99.8% of initial after 3 h). The trisaccharide β-D-Fru-(2→6)-β-D-Fru-(2→1)-D-Tag retained 78.0% after 3 h, indicating high resistance.
• In vitro fermentation (human fecal pools): Beta-diversity showed time and donor effects; β-D-Fru-(2→1)-D-Tag clustered with FOS, distinct from unmodified tagatose. Both FOS and β-D-Fru-(2→1)-D-Tag increased Bifidobacterium and Mitsuokella; unmodified tagatose favored Collinsella. Differential abundance across 25 genera showed β-D-Fru-(2→1)-D-Tag selectively stimulated Eggerthella, Ruminococcus gnavus group, Sutterella, Romboutsia, Succinivibrio, and SCFA-producing Mitsuokella. Correlations: Bifidobacterium, Lactobacillus, Enterococcus positively associated with acetic acid; Mitsuokella, Lactobacillus, Turicibacter with lactate; several genera (e.g., Alistipes, Parabacteroides, Eggerthella) with butyrate/valerate and propionate.
• Overall, fructosylated tagatose exhibited higher bifidogenic effect than tagatose and similar to commercial FOS.

The study demonstrates that tailoring rare sugars via SacB-mediated transfructosylation yields structurally defined, digestion-resistant oligosaccharides with prebiotic-like functionality. NMR elucidation confirmed β-(2→1) linkage to D-tagatose and β-(2→5) to L-sorbose, with further β-(2→6) fructan extensions. Molecular docking provided a mechanistic explanation for SacB’s acceptor specificity and the observed lack of transfer to D-psicose, linking active-site interactions and stereochemistry to product profiles. Process optimization achieved substantial β-D-Fru-(2→1)-D-Tag yields and stability. In vitro BBMV assays indicated strong resistance to small-intestinal disaccharidases, addressing the limitation of high small-intestinal absorption of native tagatose and supporting enhanced delivery to the colon. In fecal fermentations, β-D-Fru-(2→1)-D-Tag modulated microbiota similarly to FOS, increasing beneficial bifidobacteria and SCFA-associated taxa compared with unmodified tagatose, aligning structure (β-2,1 fructosylation) with function (prebiotic effect). These findings support the strategy of increasing polymerization degree to enhance bioavailability and targeted microbiota interactions of rare sugars.

SacB from Bacillus subtilis CECT 39 efficiently transfers fructosyl units from sucrose to D-tagatose (C1) and L-sorbose (C5), yielding the principal disaccharides β-D-Fru-(2→1)-D-Tag and β-D-Fru-(2→5)-L-Sor, and longer β-(2→6)-extended oligosaccharides. Structural NMR and docking analyses clarified linkage patterns and acceptor specificity, including the inability to fructosylate D-psicose. Optimized synthesis produced up to 64.04 g L−1 β-D-Fru-(2→1)-D-Tag with 21.3% yield at a 300:300 g L−1 sucrose:tagatose ratio. The main tagatose-based products were highly resistant to pig BBMV digestion, and β-D-Fru-(2→1)-D-Tag showed microbiota modulatory properties akin to FOS with a higher bifidogenic effect than tagatose. Bioconverting rare sugars into β-fructosylated species with higher DP is an effective approach to improve their bioavailability and functional impact as emerging prebiotics and low-calorie sweeteners. Future work should include in vivo validation, dose–response studies, broader donor cohorts, exploration of L-sorbose-based oligosaccharides’ bioactivity, and process scale-up/technoeconomic assessment.

• Functional assays were in vitro: small-intestinal digestibility used pig BBMV, and microbiota studies used batch fecal fermentations from two donor pools; in vivo human validation is needed.
• Molecular docking of the covalent intermediate was qualitative (no further charge adjustments), serving as structural rationale rather than quantitative prediction.
• Microbiota responses were influenced by donor and time effects; larger, more diverse cohorts would refine generalizability.
• Only tagatose-based products were functionally tested; L-sorbose-based oligosaccharides’ digestibility and fermentability were not evaluated.
• Some product fractions contained side products (e.g., β-D-Fru-(2→6)-D-Glc), and lack of commercial standards required extensive purification and NMR, which may limit throughput.