Rosetta:MSF:NN: Boosting performance of multi-state computational protein design with a neural network

Rational protein design aims at the targeted modification of existing proteins. To reach this goal, software suites like Rosetta propose sequences to introduce the desired properties. Challenging design problems necessitate the representation of a protein by means of a structural ensemble. Thus, Rosetta multi-state design (MSD) protocols have been developed wherein each state represents one protein conformation. Computational demands of MSD protocols are high, because for each of the candidate sequences a costly three-dimensional (3D) model has to be created and assessed for all states. Each of these scores contributes one data point to a complex, design-specific energy landscape. As neural networks (NN) proved well-suited to learn such solution spaces, we integrated one into the framework Rosetta:MSF instead of the so far used genetic algorithm with the aim to reduce computational costs. As its predecessor, Rosetta:MSF:NN administers a set of candidate sequences and their scores and scans sequence space iteratively. During each iteration, the union of all candidate sequences and their Rosetta scores are used to re-train NNs that possess a design-specific architecture. The enormous speed of the NNs allows an extensive assessment of alternative sequences, which are ranked on the scores predicted by the NN. Costly 3D models are computed only for a small fraction of best-scoring sequences; these and the corresponding 3D-based scores replace half of the candidate sequences during each iteration. The analysis of two sets of candidate sequences generated for a specific design problem by means of a genetic algorithm confirmed that the NN predicted 3D-based scores quite well; the Pearson correlation coefficient was at least 0.95. Applying Rosetta:MSF:NN:enzdes to a benchmark consisting of 16 ligand-binding problems showed that this protocol converges ten-times faster than the genetic algorithm and finds sequences with comparable scores.

Screening Collagenase Activity in Bacterial Lysate for Directed Enzyme Applications

Collagenases are essential enzymes capable of digesting triple-helical collagen under physiological conditions. These enzymes play a key role in diverse physiological and pathophysiological processes. Collagenases are used for diverse biotechnological applications, and it is thus of major interest to identify new enzyme variants with improved characteristics such as expression yield, stability, or activity. The engineering of new enzyme variants often relies on either rational protein design or directed enzyme evolution. The latter includes screening of a large randomized or semirational genetic library, both of which require an assay that enables the identification of improved variants. Moreover, the assay should be tailored for microplates to allow the screening of hundreds or thousands of clones. Herein, we repurposed the previously reported fluorogenic assay using 3,4-dihydroxyphenylacetic acid for the quantitation of collagen, and applied it in the detection of bacterial collagenase activity in bacterial lysates. This enabled the screening of hundreds of E. coli colonies expressing an error-prone library of collagenase G from C. histolyticum, in 96-well deep-well plates, by measuring activity directly in lysates with collagen. As a proof-of-concept, a single variant exhibiting higher activity than the starting-point enzyme was expressed, purified, and characterized biochemically and computationally. This showed the feasibility of this method to support medium-high throughput screening based on direct evaluation of collagenase activity.

How coiled-coil assemblies accommodate multiple aromatic residues

ABSTRACTRational protein design requires understanding the contribution of each amino acid to a targeted protein fold. For a subset of protein structures, namely the α;-helical coiled coils (CCs), knowledge is sufficiently advanced to allow the rational de novo design of many structures, including entirely new protein folds. However, current CC design rules center on using aliphatic hydrophobic residues predominantly to drive the folding and assembly of amphipathic α helices. The consequences of using aromatic residues—which would be useful for introducing structural probes, and binding and catalytic functionalities—into these interfaces is not understood. There are specific examples of designed CCs containing such aromatic residues, e.g., phenylalanine-rich sequences, and the use of polar aromatic residues to make buried hydrogen-bond networks. However, it is not known generally if sequences rich in tyrosine can form CCs, or what CC assemblies these would lead to. Here we explore tyrosine-rich sequences in a general CC-forming background and resolve new CC structures. In one of these, an antiparallel tetramer, the tyrosine residues are solvent accessible and pack at the interface between the core and the surface. In the other more-complex structure, the residues are buried and form an extended hydrogen-bond network.

A safety cap protects hydrogenase from oxygen attack

Abstract[FeFe]-hydrogenases are efficient H2-catalysts, yet upon contact with dioxygen their catalytic cofactor (H-cluster) is irreversibly inactivated. Here, we combine X-ray crystallography, rational protein design, direct electrochemistry, and Fourier-transform infrared spectroscopy to describe a protein morphing mechanism that controls the reversible transition between the catalytic Hox-state and the inactive but oxygen-resistant Hinact-state in [FeFe]-hydrogenase CbA5H of Clostridium beijerinckii. The X-ray structure of air-exposed CbA5H reveals that a conserved cysteine residue in the local environment of the active site (H-cluster) directly coordinates the substrate-binding site, providing a safety cap that prevents O2-binding and consequently, cofactor degradation. This protection mechanism depends on three non-conserved amino acids situated approximately 13 Å away from the H-cluster, demonstrating that the 1st coordination sphere chemistry of the H-cluster can be remote-controlled by distant residues.

Non-native fold of the putative VPS39 zinc finger domain

Background: The multi-subunit homotypic fusion and vacuole protein sorting (HOPS) membrane-tethering complex is involved in regulating the fusion of late endosomes and autophagosomes with lysosomes in eukaryotes. The C-terminal regions of several HOPS components have been shown to be required for correct complex assembly, including the C-terminal really interesting new gene (RING) zinc finger domains of HOPS components VPS18 and VPS41. We sought to structurally characterise the putative C-terminal zinc finger domain of VPS39, which we hypothesised may be important for binding of VPS39 to cellular partners or to other HOPS components. Methods: We recombinantly expressed, purified and solved the crystal structure of the proposed zinc-binding region of VPS39. Results: In the structure, this region forms an anti-parallel β-hairpin that is incorporated into a homotetrameric eight-stranded β-barrel. However, the fold is stabilised by coordination of zinc ions by residues from the purification tag and an intramolecular disulphide bond between two predicted zinc ligands. Conclusions: We solved the structure of the VPS39 C-terminal domain adopting a non-native fold. Our work highlights the risk of non-native folds when purifying small zinc-containing domains with hexahistidine tags. However, the non-native structure we observe may have implications for rational protein design.

Non-native fold of the putative VPS39 zinc finger domain

Background: The multi-subunit homotypic fusion and vacuole protein sorting (HOPS) membrane-tethering complex is involved in regulating the fusion of late endosomes and autophagosomes with lysosomes in eukaryotes. The C-terminal regions of several HOPS components have been shown to be required for correct complex assembly, including the C-terminal really interesting new gene (RING) zinc finger domains of HOPS components VPS18 and VPS41. We sought to structurally characterise the putative C-terminal zinc finger domain of VPS39, which we hypothesised may be important for binding of VPS39 to cellular partners or to other HOPS components. Methods: We recombinantly expressed, purified and solved the crystal structure of the proposed zinc-binding region of VPS39. Results: In the structure, this region forms an anti-parallel β-hairpin that is incorporated into a homotetrameric eight-stranded β-barrel. However, the fold is stabilised by coordination of zinc ions by residues from the purification tag and an intramolecular disulphide bond between two predicted zinc ligands. Conclusions: We solved the structure of the VPS39 C-terminal domain adopting a non-native fold. Our work highlights the risk of non-native folds when purifying small zinc-containing domains with hexahistidine tags. However, the non-native structure we observe may have implications for rational protein design.

Common substructures and sequence characteristics of sandwich-like proteins from 42 different folds

This study addresses the following fundamental question: Do sequences of protein domains with sandwich architecture have common sequence characteristics even though they belong to different superfamilies and folds? The analysis was carried out in two stages: determination of substructures in the domains that are common to all sandwich proteins; and detection of common sequence characteristics within the substructures. Analysis of supersecondary structures in domains of proteins revealed two types of four-strand substructures that are common to sandwich proteins. At least one of these common substructures was found in proteins of 42 sandwich-like folds (as per structural classification in the CATH database). Comparison of the sequence fragments corresponding to strands that make up the common substructures revealed specific rules of distribution of hydrophobic residues within these strands. These rules can be conceptualized as grammatical rules of beta protein linguistics. Understanding of the structural and sequence commonalities of sandwich proteins may also be useful for rational protein design.