Structure Reveals Primate-Specific Regulation of the Glycolipid Gatekeeper Enzyme UGCG

The first structural-level explanation has been provided for how UGCG, known as the “gatekeeper enzyme” of glycosphingolipid synthesis, brings two different substrates together at one site and for the distinctive regulatory mechanism found only in primates.

According to the international journal Nature on the 28th, a research team led by Professors Xu Huachang, Zhang Yi and Wu Chanlong reported these findings in a paper titled “Regulation of the Primate-Specific Human Glycosphingolipid Gatekeeper Enzyme UGCG.”

This is the third Nature research paper published by Professor Xu Huachang’s team this year.

Glycosphingolipids (GSLs) are important and highly diverse lipid molecules that make up eukaryotic cell membranes. Using ceramide as a hydrophobic backbone, cells sequentially attach, extend and modify various sugars to produce hundreds of structurally distinct glycosphingolipids.

These molecules are not only major components of cell membranes but also play broad roles in membrane microdomain organization, cell recognition and adhesion, receptor signaling, and immune and nervous system function. Abnormal glycosphingolipid metabolism is closely associated with multiple human diseases, including lysosomal storage diseases, neurologic disorders, metabolic diseases and cancer.

Although glycosphingolipids are structurally and functionally diverse, their biosynthesis shares a single key starting point. UDP-glucose ceramide glucosyltransferase (UGCG), located in the Golgi membrane, transfers glucose from UDP-glucose to ceramide to produce glucosylceramide (GlcCer).

GlcCer is subsequently glycosylated and modified to produce complex glycosphingolipids such as lactosylceramide and gangliosides. UGCG is therefore considered the central “gatekeeper enzyme” of glycosphingolipid biosynthesis, controlling whether ceramide enters the complex glycosphingolipid synthesis network.

UGCG is also an important drug target clinically. Inhibiting UGCG to reduce the production of GlcCer and downstream glycosphingolipids has become an important treatment strategy for some glycosphingolipid storage disorders.

Despite this substantial physiologic and pharmacologic importance, the most basic molecular mechanisms have long lacked a structural explanation. How UGCG simultaneously recognizes soluble UDP-glucose in the cytosol and hydrophobic ceramide embedded in the lipid bilayer, how sugar transfer occurs at the membrane-water interface, whether the membrane lipid environment regulates UGCG activity, and how existing UGCG inhibitors block this process through distinct mechanisms have remained unresolved questions.

The research team used cryo-electron microscopy, quantitative enzymology, structure-guided mutagenesis, pharmacologic analyses and molecular dynamics simulations to obtain high-resolution structures of full-length human UGCG in multiple functional states. The work systematically revealed the molecular basis by which UGCG recognizes the two substrates, catalyzes sugar transfer, senses membrane lipids and binds different types of inhibitors.

The reaction catalyzed by UGCG presents a unique spatial challenge. UDP-glucose, the sugar donor, is a soluble molecule that comes from the cytosolic side, whereas the other substrate, ceramide, contains long hydrophobic fatty acid chains and is located primarily within the lipid bilayer. UGCG must therefore obtain substrates simultaneously from the aqueous and membrane phases and position them precisely at the same catalytic center.

According to the full-length UGCG structure interpreted by the research team, UGCG is anchored in the membrane by multiple transmembrane helices, while its catalytic domain is positioned at the boundary between the cytosol and the membrane. This distinctive structural arrangement enables UGCG to form two substrate pathways that differ in character but converge. UDP-glucose enters from the cytosolic side through a hydrophilic sugar-nucleotide-binding region, while ceramide enters laterally from the lipid bilayer through a hydrophobic channel connected to the membrane.

The two substrate pathways ultimately meet at the UGCG catalytic center. This structural model shows that UGCG does not simply capture both substrates from solution. Instead, it uses its transmembrane architecture to spatially connect the cytosolic sugar donor with the lipid acceptor in the membrane, completing sugar transfer at the membrane-water interface.

After comparing ceramides with different chain lengths, the research team also confirmed that UGCG displays different reaction characteristics depending on the lipid substrate. This supports the model that ceramide approaches the catalytic center through the membrane-side channel.

After identifying the substrate-entry pathways, the research team also elucidated the chemical mechanism by which UGCG carries out sugar transfer. Structurally, UDP-glucose is precisely fixed in the cytosolic catalytic pocket of UGCG. Key positively charged residues, including R272, R275 and R376, form an interaction network that stabilizes the diphosphate group of the sugar nucleotide.

Unlike many glycosyltransferases, UGCG carries out this process without the typical assistance of a metal ion. D236 in the catalytic center occupies a key position linking UDP-glucose with the reaction center of ceramide. Taken together, the structural, mutational and functional experiments support an explanation in which D236 acts as a key catalytic residue, activating the hydroxyl group of ceramide, which then performs a nucleophilic attack on the sugar of UDP-glucose to ultimately form GlcCer.

In addition, the beta-hairpin formed by residues 370-378 is located near the sugar-donor-binding region. Structural and functional analyses showed that this region is not a static structural element but is closely involved in the recognition and stabilization of UDP-glucose. Structurally disrupting this region markedly weakened UDP-glucose binding and UGCG activity.

The research team thereby established an interconnected catalytic framework encompassing sugar-donor binding, lipid-substrate entry, activation of the catalytic residue and formation of the sugar linkage.

Because UGCG is embedded directly in the Golgi membrane, whether the surrounding lipid environment affects enzyme activity was also of interest. The researchers observed lipid densities consistent with phospholipids in the UGCG structure. Functional screening of multiple phospholipids showed that phosphatidylserine (PS) clearly regulated UGCG activity, whereas lipids including PC, PE, PG and DAG did not show the same effect.

Structural analysis indicated that PS can enter a region adjacent to the ceramide-binding pathway and form a nonproductive lipid-occupancy state. Molecular dynamics simulations and biochemical experiments also supported this model.

This suggests that the lipid bilayer may provide more than a passive reaction environment for membrane metabolic enzymes such as UGCG. The local membrane lipid composition itself may affect substrate entry and the catalytic state of the enzyme, thereby helping regulate the flow of glycosphingolipid synthesis.

In addition to the highly conserved core catalytic mechanism, the research team discovered an unexpected species-specific regulatory element. Structural comparisons showed that Y196 occupies a key position in the two substrate-binding regions and undergoes marked conformational changes depending on the substrate state.

In the UDP-glucose-bound state, Y196 adopts a relatively closed, upward-occluded conformation that partially restricts the space in the substrate pocket. In contrast, in the UDP/C6-ceramide-bound state, Y196 shifts to a downward-open conformation, expanding the lipid-substrate-binding region. The corresponding S196 in mouse UGCG predicted by AlphaFold3 showed a more open local conformation.

Phylogenetic analysis showed that position 196 underwent a characteristic substitution from serine (S) to tyrosine (Y) during mammalian evolution, producing a conserved tyrosine in primate UGCG. In functional experiments, replacing Y196 in human UGCG with serine lowered the apparent Km for the ceramide substrate and increased the maximum reaction rate. This supports the conclusion that Y196 acts as a local catalytic “gatekeeper,” restricting substrate entry and precisely regulating the enzyme’s catalytic kinetics.

These findings show that UGCG can maintain a highly conserved core catalytic system during evolution while precisely regulating substrate recognition and catalytic efficiency through a change in a single key residue near the catalytic pocket.

Accordingly, the “primate-specific regulation” emphasized in the paper does not mean that primates use a different glycosphingolipid synthesis pathway. Rather, it refers to a more refined evolutionary regulatory mechanism in which primate-specific Y196 changes the local pocket conformation and enzyme kinetic properties while preserving UGCG’s basic catalytic chemistry.

UGCG is not only an important subject of basic lipid biology research but also a clinically validated drug target. The research team obtained structures of UGCG bound to several representative inhibitors and confirmed that although the different drugs all ultimately inhibit GlcCer production, they do not act through the same mechanism.

Eliglustat and ibiglustat bind primarily to the hydrophobic channel associated with ceramide recognition. They occupy the binding space for the lipid substrate and prevent ceramide from entering the catalytic center in the correct manner.

In contrast, miglustat binds closer to the UDP-glucose recognition region and interferes with completion of the catalytic cycle from the opposite side.

Thus, different classes of UGCG inhibitors target different steps in the enzyme’s catalytic process. One strategy blocks lipid-substrate entry, whereas another disrupts the catalytic environment on the sugar-donor side.

This distinction has important implications for UGCG drug design. Previous UGCG inhibitor development relied more heavily on compound screening and efficacy optimization, but the multiple high-resolution complex structures obtained in this study now allow researchers to directly compare the binding modes, key interactions and determinants of selectivity for different drugs.

In the future, structure-guided design targeting the sugar-donor pocket, ceramide-entry channel and membrane-lipid regulatory region may enable development of new UGCG modulators with different pharmacokinetic properties, tissue distributions and disease indications.

By integrating structural and functional findings from different substrate and inhibitor-bound states, the research team also proposed a working model of the UGCG catalytic cycle. UDP-glucose enters the sugar-donor-binding pocket from the cytosolic side, while ceramide can enter the catalytic center laterally through the membrane. Upon substrate binding, Y196 and surrounding regions undergo conformational changes, providing the structural basis for the two substrates to assemble at the catalytic site.

After the reaction is complete, UDP and GlcCer are released toward the cytosolic side and into the membrane, respectively. Miglustat primarily interferes with effective binding in the sugar-donor pocket, whereas eliglustat and ibiglustat occupy the lateral channel through which ceramide enters. However, the current structures alone cannot rigorously establish the order in which the two substrates bind or the two products are released.

Tracing back hundreds of structurally complex glycosphingolipids, their biosynthesis ultimately converges on this key initiating step catalyzed by UGCG. Through structures representing multiple functional states, this study broke down UGCG’s operation into interconnected molecular events: how UDP-glucose enters from the cytosolic side, how ceramide enters laterally from the membrane, how the two substrates meet at the membrane-water interface, how sugar transfer occurs, how surrounding membrane lipids affect substrate occupancy, and how different drugs block the catalytic process at different sites.

These findings not only establish a structural framework for UGCG catalysis and regulation but also connect glycosphingolipid biosynthesis, the membrane lipid environment, evolutionary regulation and drug mechanisms within a single molecular system. They are expected to provide a new foundation for understanding glycosphingolipid homeostasis and developing intervention strategies for related diseases.

The paper’s co-first authors are Wu Chanlong and Jin Shanshan, while Xu Huachang, Zhang Yi and Wu Chanlong are co-corresponding authors. This is the third Nature research paper published by the research team this year.