From Protein-Rich to Protein-Free: CryoSeek Uncovers a Structural Continuum of Glycofibrils

Carbohydrates, or glycans, are the most abundant organic molecules on Earth and participate in a wide range of biological processes, from protein folding and immune regulation to cell wall architecture. However, Glycan Profiling and determining high-resolution structures of glycans have long been exceptionally challenging due to the remarkable diversity of monosaccharides, complex branching patterns, and the absence of direct genetic templates encoding glycan biosynthesis. While high-resolution structures have been solved for mono- or oligosaccharides bound to proteins, the higher-order structural organization of extended glycan assemblies has remained largely unexplored.

A breakthrough came with the introduction of CryoSeek, a research strategy that employs cryo-electron microscopy (cryo-EM) to image minimally processed samples directly from natural sources, supplemented with AI-facilitated data processing and bioinformatic analyses. In a study posted on Cell Chemical Biology, Li, Z., et al. (2025) from Tsinghua University and the Shenzhen Medical Academy of Research and Translation report the CryoSeek-driven discovery and structural determination of five additional glycofibrils isolated from the Tsinghua Lotus Pond (TLP). The study, titled CryoSeek identification of glycofibrils with diverse compositions and structural assemblies, expands the repertoire of known fibrous assemblies and challenges the conventional view that proteins alone dictate high-order architecture.

A Structural Continuum of Glycofibrils

Together with previously reported TLP-1a/1b and TLP-4a/4b, the researchers have now resolved nine distinct fibrils from the Tsinghua Lotus Pond at near-atomic resolutions ranging from 3.0 to 3.5 A. These nine fibrils fall into three compositional classes:

• Protein-dominant fibrils: TLP-1a and TLP-1b, which are composed primarily of proteins with typical pili-like structures.
• Glycofibrils with protein cores: TLP-IPT, TLP-12, TLP-3, TLP-2, and TLP-4a/4b, in which a protein scaffold is densely decorated with glycans that constitute the majority of the molecular mass.
• Protein-free glycofibril: TLP-0, a fibril composed entirely of glycans with no protein component whatsoever.

This spectrum reveals a remarkable structural continuum in which the protein-to-glycan ratio progressively decreases, culminating in TLP-0, an assembly built solely from carbohydrate interactions.

Fig. 1 Discovery of various fibrils from the Tsinghua Lotus Pond via the CryoSeek strategy. (Li, et al. 2025)

TLP-IPT: A Glycofibril with Folded IPT Domains

TLP-IPT represents the most protein-rich glycofibril in the newly reported set. Its central axis consists of tandem immunoglobulin-like domains, specifically the IPT (immunoglobulin-like, plexins, transcription factors) domain. Automatic model building using CryoNet and other bioinformatics software readily generated an atomic model for the protein domains, while the surrounding glycan densities required manual interpretation. Each asymmetric unit comprises a single IPT domain modified with 13 O-linked glycan chains (provisionally assigned to Ser) that form a dense glycoshell around the protein core.

A striking feature of TLP-IPT is the inter-repeat assembly mechanism. In cross-section, glycans from repeat n+1 form a C-shaped contour, and the cleft is filled by glycans from repeat n to complete a closed O ring. This organization is reminiscent of the donor-strand exchange packing mechanism observed in many bacterial pili, except that the inter-repeat interface in TLP-IPT is mediated entirely by glycans rather than protein segments. Bioinformatic analysis suggests TLP-IPT may originate from Baffinella frigidus, a species belonging to the class Cryptophyceae.

TLP-12: A Trimer of Dodecapeptide Repeats

TLP-12 features a unique protein core formed by three polypeptide chains that intertwine and spiral upward in parallel, creating a central cylinder with a triangular cross-section. Each chain comprises repeating dodecapeptide motifs that fold into short beta-strands, weaving an elongated beta-sheet ribbon around the central axis. Two glycosylation sites per repeat — at the 2nd and 5th positions — give rise to smaller and larger glycans that form parallel helical ridges wrapping around the fibril. Due to the resolution limit of the cryo-EM map, the precise identities of these glycans were tentatively modeled. Within a 14-nm TLP-12 segment, 648 residues fold into 18 repeats per chain and are modified with 1,134 sugar moieties. The S2 and S5 glycans not only interact within the same repeat but also mediate inter-chain contacts, with striking di-glycosyl stacking between adjacent monomers that may contribute to glycocalyx stabilization. Sequence analysis links TLP-12 to parasitic dinoflagellates of the genus Amoebophrya within the Alveolate clade.

TLP-3 and TLP-2: Glycofibrils with Linear Peptide Scaffolds

TLP-3 and TLP-2, together with the previously reported TLP-4a/4b, represent a class of glycofibrils whose protein cores are minimal linear peptides with short repetitive sequences. TLP-3 is a trimeric fibril with a diameter of approximately 4 nm, in which three strands of tripeptide repeats twist around each other. Each repeat, tentatively assigned as Gly-Gly-Ser, features an O-glycosylated Ser/Thr modified with a branched glycan structure containing eight sugar residues. Helical stabilization arises from both backbone hydrogen bonds between neighboring chains and interlocking glycans from adjacent repeats.

TLP-2 contains an even simpler scaffold: a single chain of dipeptide repeats (tentatively assigned as Ala-Ser) with a glycosylated phosphoserine in each repeat. The cryo-EM density was consistent with phosphoglycosylation, such as Ser-linked GlcNAc-1-PO₄, a modification previously described in Dictyostelium discoideum Proteinase I. Glycan-mediated inter-repeat interactions between repeats n and n+1, and between n and n+3, collectively stabilize the helical assembly of TLP-2.

TLP-0: A Fibril Composed Entirely of Glycans

The most remarkable discovery in this study is TLP-0, a fibril that lacks any protein component and is composed exclusively of Glycans, with helical parameters of 2.6 Å rise and −80.0° twist. The core of TLP-0 consists of a spiral tri-glycosyl repeat with the composition Fucalpha1-3GalNAcalpha1-3Man, modeled at a local resolution of nearly 2.8 A. Three glycan branches are attached to each core repeat: one facing the interior and two extending outward. In total, 18 sugar residues were built onto each tri-glycosyl repeat.

The assembly of TLP-0 is maintained entirely by Glycan-mediated Interactions. On the interior, a sugar residue from branch 1 in repeat n+3 stacks against the Man residue of repeat n. On the exterior, branch 2 of repeat n engages in multiple interactions with branch 2 of repeats n-1 and n+4. The all-sugar composition of TLP-0, which pushes the limits of conventional glycan profiling, is reminiscent of starch, cellulose, and glycogen, but with a more complex monosaccharide composition. This discovery raises fundamental questions about the structural autonomy of glycans and their capacity to form highly ordered assemblies in the absence of protein templates.

Fig. 2 TLP-0 is formed entirely by glycans. (Li, et al. 2025)

Implications: Forward Structural Biology

From a structural biology and Glycoproteomics perspective, these findings highlight the underappreciated role of glycan-mediated interactions in determining high-order structural assembly. In contrast to the conventional paradigm in which proteins dominate architecture and glycans play supporting roles, the glycofibrils described here — particularly TLP-0 — demonstrate that glycans can independently direct structural folding and fibril formation.

Although the physiological functions of these glycofibrils remain unclear, they may represent a strategy for extracellular carbon storage. Unlike starch and glycogen, which accumulate intracellularly, or cellulose, which defines the cell wall, glycofibrils extend into the extracellular milieu, offering potentially unlimited space for carbon sequestration. Such fibrils could serve as nutrient reservoirs or mediate host-parasite interactions in specific ecological niches. In a broader sense, the CryoSeek strategy may represent "forward structural biology" — a structure-first paradigm for biological discovery.

Related Services & Products

• Glycan Release Service
• Glycan Separation and Purification Service
• Glycan Derivation Service
• Glycomic Characterization Service
• Glycomic Quantitation
• Glycoprotein Enrichment
• Glycoprotein Structure Analysis
• Glyco-biomarker Detection Service
• Custom Glycosylation Service

Reference

1. Li, Z., et al. (2025). CryoSeek identification of glycofibrils with diverse compositions and structural assemblies. Cell Chemical Biology. DOI: 10.1016/j.chembiol.2026.03.008.