Topology concerns the connectivity and spatial relationship of a molecule, and is an important molecular attribute for macromolecules. The additional covalent connectivity and mechanical interlocking in topological proteins impose considerable conformational constraints on their unfolded states, thereby enhancing the stability of their native states in different ways to varying degrees. Therefore, expanding the protein fold space beyond linear chains is of fundamental significance. The team led by Prof. Wen-Bin Zhang at the College of Chemistry and Molecular Engineering, Peking University, has reported the creation of seven topological isoforms (i.e., linear, cyclic, knot, lasso, pseudorotaxane, and catenane) from a single protein fold precursor. The research results were recently published online in Proceedings of the National Academy of Sciences, USAon October 15, 2024 (https://www.pnas.org/doi/10.1073/pnas.2407355121) with the title “Folds from Fold: Exploring Topological Isoforms of a Single Domain Protein”.
The SpyTag-SpyCatcher complex is robust and can tolerate extensive protein engineering. It has been successfully split into three parts, namely, SpyTag, BDTag, and SpyStapler. Upon reconstitution, SpyStapler can still catalyze the isopeptide bonding between SpyTag and BDTag. By rewiring the connectivity of their N- and C-terminus (N1/C1 of SpyTag, N2/C2 of BDTag, and N3/C3 of SpyStapler) to introduce entanglement and mutating the reactive residue on SpyTag to abolish the isopeptide bonding, a series of protein folds with distinct chemical topology can be designed (Fig. 1).
Fig. 1 Design of seven topological isoforms from a single protein fold precursor by rewiring the connectivity of secondary structure elements of the SpyTag-SpyCatcher complex and mutating the reactive residue on SpyTag to abolish the isopeptide bonding.
Two series of samples were designed for synthesis, one with relatively short linkers (the s-series) and the other with sufficiently long linkers (the l-series). The former actually induced domain swapping and led to oligomerization (Fig. 2A and 2B). Fortunately, the latter ensured intramolecular reconstitution producing the expected single-domain topological protein as the major product (Fig. 2C). Within each series, although the overall amino acid compositions differed slightly (mostly at the new linker region), the overall number of amino acids was kept identical for consistency.
Fig. 2 (A) Intramolecular reconstitution and domain swapping are influenced by the length of the linker. A short linker length promotes domain swapping and a long linker length promotes intramolecular reconstitution. (B) SDS-PAGE analysis and SEC trace of purified topological proteins with short linkers. (C) SDS-PAGE analysis and SEC trace of purified topological proteins with long linkers.
To reveal the structure-property relationship, the l-series of single-domain topological isoforms were used for further study. The CD spectrometry and 2D 1H-15N heteronuclear single quantum coherence (HSQC) NMR spectroscopy were used to verify that the samples remained well-folded. As shown in Fig. 3A, all samples exhibited a small bump near 230 nm characteristic of the β-turn structure of the SpyTag-SpyCatcher complex, suggesting that their secondary structure elements remained largely undisturbed. Additionally, all samples displayed well-dispersed and sharp peaks in HSQC-NMR spectra (Fig. 3B) similar to those of the SpyTag-SpyCatcher complex. The characteristic cross-peaks at an up-field 1H chemical shift (∼6.0 ppm) in l-cyclic, l-lasso, l-catenane, and l-rotaxane were consistent with the isopeptide bond formation. The results suggest well-folded structures inall these topological isoforms. The melting temperatures (Tm) of these topological isoforms were determined by DSC. As shown in Fig. 3C, the Tm values of the proteins with isopeptide bonding were typically over 30 °C higher than those of their DA mutants . Consistent with the experimental results, the DA mutants also exhibited much more dynamic motions than the isopeptide-bonded topologies in MD simulation, as evidenced by the constantly more fluctuating distance between the Cα of Lys31 and Ala117 (Fig. 3D). The results suggest that the isopeptide bonds stabilize the entire protein domain by maintaining valid contacts and constraining the segments’ relative motions, which is in agreement with their higher Tm values.
The topological isoforms exhibit distinct folding/unfolding behaviors characteristic of different free energy landscapes. To promote unfolding, MD simulation was conducted at 498 K and 26 atm to speed up the unfolding process within a relatively short simulation time. The clear color boundary of dynamic cross-correlation matrixes of catenane indicates a negative correlation between the two individual rings and a positive self-correlation within each ring (Fig. 3E). This independence of the two chains, combined with their mutual interaction through mechanical bonds, significantly enriches the protein’s dynamic behaviors. During unfolding, catenane also retains a high level of secondary structure contents. There is considerable transient recovery of secondary structures as shown by cycles of β-sheet unfolding and refolding during 20-80 ns, which suggests excellent resilience (Fig. 3F).
Fig. 3 (A) CD spectra of topological proteins at 303 K. (B) Overlay of HSQC NMR spectra of topological proteins. (C) DSC curves of seven topological proteins. (D) Distance of residues at the isopeptide bonding site during simulation with stick illustration of the Lys and mutated Ala residues. (E) Dynamic cross-correlation matrix (DCCM) of protein catenane at 498 K and 26 atm, shows the relative motion between the two rings. (F) Fraction of secondary structure of catenane during simulation at 498 K and 26 atm, and representative structures captured from unfolding trajectories showing the cycles of unfolding and refolding during simulation.
Dr. Zhiyu Qu at Peking University is the first author of this paper. Prof. Wen-Bin Zhang from Peking University is the corresponding author. Dr. Lianjie Xu, Mr. Fengyi Jiang, and Dr. Yuan Liu from Peking University also contributed to this work. This research was jointly supported by the National Natural Science Foundation of China, the National Key R&D Program of China, and Beijing National Laboratory for Molecular Sciences.
Original link for the paper: https://www.pnas.org/doi/10.1073/pnas.2407355121
