Protein-Selbstorganisation durch das Design autokatalytischer Isopeptidbindungen
Biologische und Biomimetische Chemie
Biomaterialien
Physikalische Chemie von Molekülen, Flüssigkeiten und Grenzflächen, Biophysikalische Chemie
Strukturbiologie
Zusammenfassung der Projektergebnisse
Proteins are structurally and chemically complex building blocks, yet this complexity enables (artificial) protein assemblies to form structures with unprecedented function and unusual properties. To drive their self-assembly, scientist have applied different strategies, taking inspiration from nature. Autocatalytically formed isopeptide bonds (IPB) are a unique structural feature identified in only a limited set of protein families so far. With the extraordinary stability they provide and due to their orthogonal nature to alternative covalent protein-protein crosslinks (e.g., disulfide bonds), it would be invaluable to add them to our repertoire of designable elements. Yet, IPB are intrinsically difficult to study, and their de novo design has not been accomplished. We show that the de novo design of IPB is possible, building on two recently developed deep learning-based protein design tools (RFdiffusion and ProteinMPNN). Two new IPB forming proteins with minimal sequence identity to existing IPB-forming proteins were designed. Both proteins utilize an Asn/Lys/Glu triad to autocatalytically form the covalent bond. This proof of principle shows that IPB-forming motifs can be “transplanted” into different protein contexts and opens avenues toward designing custom proteins and protein assemblies in which IPB provide a key stabilizing function. These new designs will additionally serve as functional models, driving mechanistic studies of IPB formation. Furthermore, learnings from initial failed IPB design attempts led us to discover a method to design a stable apo-form of an otherwise heme-containing protein. While at first, the IPB design was complicated by an unexpected behavior of the chosen four-helix-bundle model protein, transferring the knowledge that we gained during this process to another four-helix bundle protein (cytochrome cb562) helped us solve a long-standing challenge: Cytochrome cb562 is an exquisite building block for the formation of discrete and extended protein assemblies. However, the presence of its heme cofactor complicates protein expression and downstream enzymatic and spectroscopic assays. Being able to remove a mandatory metal cofactor without compromising the protein’s stability streamlines the design of new protein assemblies and potentially the study of other metal cofactor-containing proteins. Finally, building on this new apo-cytochrome (ApoCyt), we designed a first-in-class molecular protein weave with surprising physicochemical properties. Molecular weaves are periodically and regularly entangled molecular strands. As opposed to crosslinked polymers, weaves do not require direct physical crosslinking between the strands to remain structurally intact. Instead, they are mechanically interlocked, such that they can move freely within the topological limits set by their mechanically bonded partners. Yet breaking of a covalent bond is necessary for their full dissociation. By combining metal-induced oligomerization, protein crystallization, and installing key covalent bonds, we were able to form woven protein crystals. Due to their unique topological constraints, these crystals were able to reversibly expand and contract in response to the surrounding medium and exhibit stiffness values (Young’s moduli) spanning two orders of magnitude during the expansion/contraction process. Overall, we are very satisfied with the results of the project, as the central goal of the project (de novo design of IPB) was achieved. Even if, due to unexpected complications, not all experiments worked exactly as proposed, we were able to utilize the knowledge gained during the process to develop new methods (design of ApoCyt) and unlock the path to a new kind of dynamic protein material (molecular protein weave).
Projektbezogene Publikationen (Auswahl)
-
Computationally Guided Redesign of a
Heme-free Cytochrome with Native-like Structure and Stability. Biochemistry 2022, 61,
2063-2072
Alexander M Hoffnagle, Vanessa H Eng , Ulrich Markel , F Akif Tezcan
-
De Novo Design of Proteins for Autocatalytic Isopeptide Bond Formation, J. Am. Chem. Soc. 2025
Suppachai Srisantitham,Alyssa L. Walker,Ulrich Markel and F.Akif Tezcan
