Traditionally, chemists have designed artificial enzymes based on proteins or RNA. Meanwhile, glycans—the third most widespread biomolecules in nature—have mostly been regarded as biological markers or structural components rather than active players in chemical reactions. But what if glycans could also catalyze reactions and serve as catalytic scaffolds? This question has driven our research, building on the discovery of the “glycan hairpin,”¹ which demonstrated that glycans can be folded into stable three-dimensional (3-D) structures soluble in water.

Glycans’ potential as catalysts was largely untapped yet intriguing. Compared with proteins, they offer higher structural diversity, with over a hundred different monosaccharides available as building blocks. They also have the ability to form branches, and exhibit different directionality waiting for function along with intrinsic chirality.  Last but not least, glycans can engage in multiple interactions with a substrate due to synergies of hydrogen bonding and hydrophobic effects. These intrinsic properties make them ideal for fine-tuned folding behavior, creating tailored 3-D environment to accommodate substrates via the endowed interactions, further for performing functions, i.e. catalysis. 

Figure 1. The design of this work

We first turned to nature for inspiration. Many biological systems use glycans to interact with substrates, particularly biomolecules-containing aromatic groups, through a type of weak bonding known as CH/π interactions (Fig. 1a). These interactions are foundational to carbohydrate-binding proteins such as lectins, glycosyltransferases, and enzymes, which selectively recognize and bind sugar structures. The challenge was to identify and then engineer the right glycan scaffold—one that could support CH/π interactions and recruit an aromatic substrate in proximity of the reactive site while maintaining a stable 3-D conformation.

Sialyl Lewis X (sLeX), a naturally occurring glycan known for its rigid glycan turn, is stabilized by an unconventional hydrogen bonding, along with an extended sialic acid (Neu5Ac) arm (Fig. 1b). Based on this structure, we introduced two key modifications. First, we replaced the Neu5Ac unit with β-Galactose (β-Gal), a sugar particularly suited for interacting with aromatic molecules like L-tryptophan (Trp) thanks to its strategically placed C-H bonds. The second tweak involved swapping fucose (Fuc) for rhamnose (Rha), allowing precise positioning of a C-4 hydroxyl (-OH) group near the CH/π interaction site (and thus aimed to the installation of a catalytic group). These modifications transformed sLeX from a natural biologically important glycan into a promising synthetic scaffold, capable of binding an aromatic substrate and positioning it next to a reactive site for catalysis (Fig. 1c).

Figure 2.  CH/π interactions between 4mer-I and 4mer-IV with Trp.

To test whether it could interact with Trp, we synthesized a neutral glycan structure, 4mer-I, and analyzed its folding behavior. NMR spectroscopy and molecular dynamics simulations confirmed that 4mer-I retained its rigid conformation in water—a crucial feature for supporting catalytic function (Fig. 2a). More importantly, when we incubated 4mer-I with Trp, we observed chemical shift changes concentrated around the β-Gal unit, confirming that our glycan successfully engaged in CH/π interactions (Fig. 2b). This was the validation we had been hoping for: our planned interaction site could selectively bind an aromatic substrate in the presence of other sugar units.

The next step was to introduce functional groups onto the scaffold and assess their impact on structural stability and catalytic performance. We synthesized three variations: 4mer-II with a phosphoric acid group, 4mer-III with a sulfuric acid group, and 4mer-IV with a carboxylic acid group. Each maintained its structural integrity, but their ability to interact with Trp varied. Both 4mer-II and 4mer-III exhibited weaker CH/π interactions due to steric hindrance, where bulky functional groups interfered with Trp binding. In contrast, 4mer-IV maintained strong CH/π interactions, similar to the original 4mer-I (Fig. 2c). The presence of an extra methylene group in 4mer-IV provided additional flexibility, allowing it to better accommodate Trp in the correct position.

Figure 3. Application of a glycan foldamer as catalyst in a Pictet-Spengler reaction.

It was then time to test their catalytic activity using the Pictet-Spengler reaction, in which Trp reacts with propionaldehyde in water (Fig. 3). The results were definitive. Without a catalyst, the reaction produced almost no yield. When we tested 4mer-II and 4mer-III, they also failed to accelerate the reaction, implicating that the decreased CH/π interactions prevented effective catalysis. This defied our expectations and prompted a change of strategy: we added flexibility to 4mer-IV to achieve stronger CH/ π interactions to bind aromatic substrates. And this is how 4mer-IV stood out, remarkably improving reaction efficiency. To our excitement, it became clear that CH/π interactions played a crucial role in this catalytic process.

To validate our findings, we synthesized two additional control molecules. The first was a shorter glycan, 3mer, which lacked the CH/π binding site. As expected, it showed no interaction with Trp and resulted in a low reaction yield. The second control was 4mer-V, a glycan with glucose instead of galactose at the binding site. This variant exhibited moderate CH/π interactions and an intermediate reaction yield (Fig. 3). The conclusive proof came when 4mer-IV demonstrated both the strongest CH/π interactions and the highest reaction yield, confirming that these interactions were directly linked to catalytic efficiency.

Our findings suggested that glycans can be designed as rigid biomolecules with modular adaptability to catalyze an organic transformation, paving the way toward the design of glycan-based catalysts. This work was envisioned to push the boundaries of glycan synthesis and conformational analysis, shedding new light on glycans’ interactions with other molecules. Beyond their well-established roles in biological recognition and signaling, glycans might play other overlooked function in biological systems.

Reference

1. Fittolani, G.; Tyrikos-Ergas, T.; Poveda, A.; Yu, Y.; Yadav, N.; Seeberger, P. H.; Jiménez-Barbero, J.; Delbianco, M. Synthesis of a Glycan Hairpin. Nat. Chem. 2023, 15 (10), 1461-1469.