Summary

A bioRxiv preprint reports the first three-dimensional structure of a class II GalT enzyme from Bifidobacterium longum at 2.25 Å resolution. The study links its unusual architecture to a preference for N-acetylated sugars involved in human milk oligosaccharide metabolism.

A bioRxiv preprint reports the first three-dimensional structure of a class II UDP-glucose–hexose-1-phosphate uridylyltransferase, or GalT, from Bifidobacterium longum. The enzyme, named BlGalT2, was resolved at 2.25 Å and examined through structural, kinetic and mutational analyses.

BlGalT2 is part of the GNB/LNB metabolic pathway used by infant-associated bifidobacteria to process lacto-N-biose I (LNB) and galacto-N-biose (GNB). These sugars are major components of human milk oligosaccharides and intestinal mucin. The findings provide a molecular view of a rare enzyme class that differs substantially from the better-known class I GalTs.

Contents

What the study examined

The researchers studied BlGalT2 from B. longum JCM 1217, a bacterium associated with the infant gut. The enzyme performs a key step in the pathway by which bifidobacteria use host-derived glycans as nutrients.

The work was posted on bioRxiv on September 16, 2026, as a preprint. Its central result is a crystal structure at 2.25 Å resolution. At this scale, structural biology can reveal the arrangement of protein domains and the shape of the pocket where substrates bind and react.

The study also measured enzyme activity with different sugar-phosphate substrates, analysed the role of metal ions and used AlphaFold3-based structural prediction alongside targeted mutations. Together, these experiments connected the observed protein structure with substrate recognition and a proposed catalytic mechanism.

A different enzyme architecture

Class II GalTs belong to the histidine triad, or HIT, superfamily. They are distinct from canonical class I GalTs, which carry out related uridylyltransferase chemistry using a different structural framework.

BlGalT2 is a monomeric protein with two characteristic catalytic core subdomains, called HIT1 and HIT2. It also contains an extended auxiliary domain that helps form a self-contained active site. The arrangement gives class II GalTs a structural identity of their own rather than representing a simple variation of the class I fold.

The structure therefore helps explain how bifidobacteria can carry out a step in milk-sugar metabolism using an enzyme family that is uncommon compared with the canonical GalTs. The authors describe this structure as closing the remaining structural gap in the known pathway of bifidobacterial human milk oligosaccharide metabolism.

How the enzyme recognises sugars

The kinetic experiments showed a strong preference for GalNAc-1P, or N-acetylgalactosamine-1-phosphate, over Gal-1P, or galactose-1-phosphate. The measured Michaelis constant for GalNAc-1P was 50 times lower than for Gal-1P. This difference is consistent with a hydrophobic pocket that can accommodate the N-acetyl group on GalNAc-1P.

Metal analysis and activity measurements indicated that Zn2+ is required for full enzyme activity. Structural prediction suggested that the metal is coordinated by several residues, including the first histidine in the HIT motif. Mutations and AlphaFold3 modelling also identified residues involved in recognising the nucleotide-sugar substrates.

The results support a Ping-Pong reaction mechanism. In this type of mechanism, the enzyme reacts with one substrate first, temporarily forming a covalent intermediate, before reacting with the next substrate. For BlGalT2, the proposed intermediate is a covalent enzyme–uridine monophosphate complex.

Taken together, the findings show how a class II GalT combines a distinctive HIT-superfamily architecture, zinc-dependent activity and a substrate pocket suited to N-acetylated sugars. They also offer a molecular explanation for how bifidobacteria have adapted enzymes to process glycans supplied by their host environment. The evidence concerns purified enzyme structure and activity, rather than a human or animal intervention.

Sources