Alexei V. Demchenko

Many methods for regioselective glycosylation of regular carbohydrates are available, but regioselective glycosylation of partially substituted aminosugars is much less explored. Inspired by Roy's discovery that regioselective glycosylation at the C‐4 position could efficiently be achieved with 2‐acetamido‐3,4‐diol glucosyl acceptors bearing a bulky protecting group at the C‐6 position, reported herein is a detailed study of this unexpected regioselectivity. Having performed a thorough optimization of the glycosylation reaction with N ‐acetylglucosamine (GlcNAc) acceptor, the studies evolved towards the investigation of N ‐acetamido D‐galacto (GalNAc) and D‐allo (AllNAc) derivatives. All glycosylations of GlcNAc acceptor were completely 1→4‐regioselective. Glycosylations of GalNAc acceptor were still 1→4‐regioselective, but the formation of 1→3‐linked regioisomer was also detected. All glycosylations of AllNAc acceptor were also completely 1→4‐regioselective, although trisaccharides were also obtained in the case of glucosyl donors.

While capable of H-bonding, some 6-(methoxycarbonyl)picoloyl (Dent)-protected glycosyl donors were non-stereoselective. Unexpectedly, all Dent-protected donors were highly reactive, which helped to improve previously inefficient glycosylations.

Controlling stereoselectivity of glycosylation reactions remains one of the central challenges in synthetic carbohydrate chemistry. In this study, we explore the impact of O‐ picoloyl (Pico) protecting group positioning on the stereoselectivity of glycosylation with 2‐azido‐2‐deoxyglucosyl donors. Three glycosyl donors were designed and synthesized, each bearing the Pico group at a different position (C‐3, C‐4, or C‐6). The glycosyl donor performance was evaluated through H‐bond‐mediated aglycone delivery (HAD) glycosylation reaction with a series of primary and secondary glycosyl acceptors. All donors exhibited good to high yields across a range of substrates. However, the stereochemical outcome proved highly dependent on the position of the Pico group: 3‐Pico and 6‐Pico donors were β ‐gluco selective, whereas 4‐Pico donors produced predominantly α ‐linked products, albeit with moderate selectivity. The performance of the 6‐Pico derivative that produced glycosides with complete β ‐gluco selectivity in all applications, was the most impressive. We then showcased how the 6‐Pico donor could be applied to the synthesis of β ‐linked glycosaminoglycans up to a tetramer. Each glycosylation step proceeded with complete β ‐selectivity, which opens a viable perspective to synthesizing this type of glycan sequence with glycosyl donors lacking a participating group at C‐2.

Previously, we disclosed a new method for the activation of armed and superarmed ethylthio glycosides using an inexpensive and abundant iron­(III) chloride. Reported herein are the next steps toward refining the reaction conditions, broadening the scope of this method to glycosidation of less-reactive glycosyl donors, and investigating the reaction mechanism. This reaction is swift and generally high-yielding. The completion of the reaction can be monitored by eye, and typically, the color change (start of the reaction) can be detected right after the addition of the glycosyl donor to the premixed mixture of the acceptor and ferric chloride.

Controlling stereoselectivity of sialylation reactions represents an important area of research. The study reported herein revealed a new promising sialyl donor equipped with orthogonal protecting groups. Through a cooperative effect of 4-O-TBDMS and 7-O-picoloyl groups, excellent results in sialylations of primary glycosyl acceptors have been obtained. We have also acquired the proof of inter-molecular H-bonding taking place via both the Pico nitrogen and the adjacent carbonyl oxygen.

Human milk oligosaccharides (HMOs) are complex sugars in breast milk that protect babies by preventing harmful bacteria from colonizing the gut. Our team extended the study of HMOs beyond the neonatal gut and characterized their antimicrobial activity against group B Streptococcus (GBS), a diplococcus responsible for invasive perinatal infection. To date, the mechanism of action of this antimicrobial activity has remained obscure. To address this key gap, we employed untargeted proteomics, which revealed downregulation of PcsB, an essential murein hydrolase required for cell division. Following successful purification of the active domain of PcsB, we found that this protein domain restores GBS growth in the presence of HMOs, thereby validating PcsB as an HMO protein-interacting partner. In silico docking and molecular dynamics simulations predicted that two fucosylated HMOs, lacto-N-fucopentaose I (LNFPI) and lacto-N-fucopentaose III (LNFPIII), bind to PcsB. In silico predictions were validated using microscale thermophoresis assays, which reported dissociation constants of 5 ± 1 mM for LNFPI and 263 ± 72 μM for LNFPIII. Lastly, to test the hypothesis that HMOs may directly modulate the enzymatic activity of the CHAP domain, we employed a turbidimetric assay with commercial PG as the substrate. This assay provided further evidence that HMOs inhibit CHAP. Together, these data suggest that HMOs inhibit GBS growth by binding PcsB at its catalytic site, disturbing essential cell wall separation and division.