Rossmann-toolbox: a deep learning-based protocol for the prediction and design of cofactor specificity in Rossmann fold proteins.

The Rossmann fold enzymes are involved in essential biochemical pathways such as nucleotide and amino acid metabolism. Their functioning relies on interaction with cofactors, small nucleoside-based compounds specifically recognized by a conserved ßαß motif shared by all Rossmann fold proteins. While Rossmann methyltransferases recognize only a single cofactor type, the S-adenosylmethionine, the oxidoreductases, depending on the family, bind nicotinamide (nicotinamide adenine dinucleotide, nicotinamide adenine dinucleotide phosphate) or flavin-based (flavin adenine dinucleotide) cofactors. In this study, we showed that despite its short length, the ßαß motif unambiguously defines the specificity towards the cofactor. Following this observation, we trained two complementary deep learning models for the prediction of the cofactor specificity based on the sequence and structural features of the ßαß motif. A benchmark on two independent test sets, one containing ßαß motifs bearing no resemblance to those of the training set, and the other comprising 38 experimentally confirmed cases of rational design of the cofactor specificity, revealed the nearly perfect performance of the two methods. The Rossmann-toolbox protocols can be accessed via the webserver at https://lbs.cent.uw.edu.pl/rossmann-toolbox and are available as a Python package at https://github.com/labstructbioinf/rossmann-toolbox.

pLM-BLAST: distant homology detection based on direct comparison of sequence representations from protein language models.

MOTIVATION: The detection of homology through sequence comparison is a typical first step in the study of protein function and evolution. In this work, we explore the applicability of protein language models to this task. RESULTS: We introduce pLM-BLAST, a tool inspired by BLAST, that detects distant homology by comparing single-sequence representations (embeddings) derived from a protein language model, ProtT5. Our benchmarks reveal that pLM-BLAST maintains a level of accuracy on par with HHsearch for both highly similar sequences (with >50% identity) and markedly divergent sequences (with <30% identity), while being significantly faster. Additionally, pLM-BLAST stands out among other embedding-based tools due to its ability to compute local alignments. We show that these local alignments, produced by pLM-BLAST, often connect highly divergent proteins, thereby highlighting its potential to uncover previously undiscovered homologous relationships and improve protein annotation. AVAILABILITY AND IMPLEMENTATION: pLM-BLAST is accessible via the MPI Bioinformatics Toolkit as a web server for searching precomputed databases (https://toolkit.tuebingen.mpg.de/tools/plmblast). It is also available as a standalone tool for building custom databases and performing batch searches (https://github.com/labstructbioinf/pLM-BLAST).

localpdb-a Python package to manage protein structures and their annotations.

MOTIVATION: The wealth of protein structures collected in the Protein Data Bank enabled large-scale studies of their function and evolution. Such studies, however, require the generation of customized datasets combining the structural data with miscellaneous accessory resources providing functional, taxonomic and other annotations. Unfortunately, the functionality of currently available tools for the creation of such datasets is limited and their usage frequently requires laborious surveying of various data sources and resolving inconsistencies between their versions. RESULTS: To address this problem, we developed localpdb, a versatile Python library for the management of protein structures and their annotations. The library features a flexible plugin system enabling seamless unification of the structural data with diverse auxiliary resources, full version control and powerful functionality of creating highly customized datasets. The localpdb can be used in a wide range of bioinformatic tasks, in particular those involving large-scale protein structural analyses and machine learning. AVAILABILITY AND IMPLEMENTATION: localpdb is freely available at https://github.com/labstructbioinf/localpdb. Documentation along with the usage examples can be accessed at https://labstructbioinf.github.io/localpdb/.

A library of coiled-coil domains: from regular bundles to peculiar twists.

MOTIVATION: Coiled coils are widespread protein domains involved in diverse processes ranging from providing structural rigidity to the transduction of conformational changes. They comprise two or more α-helices that are wound around each other to form a regular supercoiled bundle. Owing to this regularity, coiled-coil structures can be described with parametric equations, thus enabling the numerical representation of their properties, such as the degree and handedness of supercoiling, rotational state of the helices, and the offset between them. These descriptors are invaluable in understanding the function of coiled coils and designing new structures of this type. The existing tools for such calculations require manual preparation of input and are therefore not suitable for the high-throughput analyses. RESULTS: To address this problem, we developed SamCC-Turbo, a software for fully automated, per-residue measurement of coiled coils. By surveying Protein Data Bank with SamCC-Turbo, we generated a comprehensive atlas of ∼50 000 coiled-coil regions. This machine learning-ready dataset features precise measurements as well as decomposes coiled-coil structures into fragments characterized by various degrees of supercoiling. The potential applications of SamCC-Turbo are exemplified by analyses in which we reveal general structural features of coiled coils involved in functions requiring conformational plasticity. Finally, we discuss further directions in the prediction and modeling of coiled coils. AVAILABILITY AND IMPLEMENTATION: SamCC-Turbo is available as a web server (https://lbs.cent.uw.edu.pl/samcc_turbo) and as a Python library (https://github.com/labstructbioinf/samcc_turbo), whereas the results of the Protein Data Bank scan can be browsed and downloaded at https://lbs.cent.uw.edu.pl/ccdb. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

ß-sheet breakers with consecutive phenylalanines: Insights into mechanism of dissolution of ß-amyloid fibrils.

ß-sheet breakers (BSB) constitute a class of peptide inhibitors of amyloidogenesis, a process which is a hallmark of many diseases called amyloidoses, including Alzheimer's disease (AD); however, the molecular details of their action are still not fully understood. Here we describe the results of the computational investigation of the three BSBs, iaß6 (LPFFFD), iaß5 (LPFFD), and iaß6_Gly (LPFGFD), in complex with the fibril model of Aß42 and propose the kinetically probable mechanism of their action. The mechanism involves the binding of BSB to the central hydrophobic core (CHC) region (LVFFA) of Aß fibril and the π-stacking of its Phe rings both internally and with the Aß fibril. In the process, the Aß fibril undergoes distortion accumulating on the side of chain A (located on the odd tip of the fibril). In a single replica of extended molecular dynamics run of one of the iaß6 poses, the distortion concludes in a dissociation of chain A from the fibril model of Aß42. Altogether, we postulate that including consecutive Phe residues into BSBs docked around Phe 20 in the CHC region of Aß42 improve their potency for dissolution of fibrils.

Interplay between DsbA1, DsbA2 and C8J_1298 Periplasmic Oxidoreductases of Campylobacter jejuni and Their Impact on Bacterial Physiology and Pathogenesis.

The bacterial proteins of the Dsb family catalyze the formation of disulfide bridges between cysteine residues that stabilize protein structures and ensure their proper functioning. Here, we report the detailed analysis of the Dsb pathway of Campylobacter jejuni. The oxidizing Dsb system of this pathogen is unique because it consists of two monomeric DsbAs (DsbA1 and DsbA2) and one dimeric bifunctional protein (C8J_1298). Previously, we showed that DsbA1 and C8J_1298 are redundant. Here, we unraveled the interaction between the two monomeric DsbAs by in vitro and in vivo experiments and by solving their structures and found that both monomeric DsbAs are dispensable proteins. Their structures confirmed that they are homologs of EcDsbL. The slight differences seen in the surface charge of the proteins do not affect the interaction with their redox partner. Comparative proteomics showed that several respiratory proteins, as well as periplasmic transport proteins, are targets of the Dsb system. Some of these, both donors and electron acceptors, are essential elements of the C. jejuni respiratory process under oxygen-limiting conditions in the host intestine. The data presented provide detailed information on the function of the C. jejuni Dsb system, identifying it as a potential target for novel antibacterial molecules.

Self-analysis of repeat proteins reveals evolutionarily conserved patterns.

BACKGROUND: Protein repeats can confound sequence analyses because the repetitiveness of their amino acid sequences lead to difficulties in identifying whether similar repeats are due to convergent or divergent evolution. We noted that the patterns derived from traditional "dot plot" protein sequence self-similarity analysis tended to be conserved in sets of related repeat proteins and this conservation could be quantitated using a Jaccard metric. RESULTS: Comparison of these dot plots obviated the issues due to sequence similarity for analysis of repeat proteins. A high Jaccard similarity score was suggestive of a conserved relationship between closely related repeat proteins. The dot plot patterns decayed quickly in the absence of selective pressure with an expected loss of 50% of Jaccard similarity due to a loss of 8.2% sequence identity. To perform method testing, we assembled a standard set of 79 repeat proteins representing all the subgroups in RepeatsDB. Comparison of known repeat and non-repeat proteins from the PDB suggested that the information content in dot plots could be used to identify repeat proteins from pure sequence with no requirement for structural information. Analysis of the UniRef90 database suggested that 16.9% of all known proteins could be classified as repeat proteins. These 13.3 million putative repeat protein chains were clustered and a significant amount (82.9%) of clusters containing between 5 and 200 members were of a single functional type. CONCLUSIONS: Dot plot analysis of repeat proteins attempts to obviate issues that arise due to the sequence degeneracy of repeat proteins. These results show that this kind of analysis can efficiently be applied to analyze repeat proteins on a large scale.

DeepCoil-a fast and accurate prediction of coiled-coil domains in protein sequences.

MOTIVATION: Coiled coils are protein structural domains that mediate a plethora of biological interactions, and thus their reliable annotation is crucial for studies of protein structure and function. RESULTS: Here, we report DeepCoil, a new neural network-based tool for the detection of coiled-coil domains in protein sequences. In our benchmarks, DeepCoil significantly outperformed current state-of-the-art tools, such as PCOILS and Marcoil, both in the prediction of canonical and non-canonical coiled coils. Furthermore, in a scan of the human genome with DeepCoil, we detected many coiled-coil domains that remained undetected by other methods. This higher sensitivity of DeepCoil should make it a method of choice for accurate genome-wide detection of coiled-coil domains. AVAILABILITY AND IMPLEMENTATION: DeepCoil is written in Python and utilizes the Keras machine learning library. A web server is freely available at https://toolkit.tuebingen.mpg.de/#/tools/deepcoil and a standalone version can be downloaded at https://github.com/labstructbioinf/DeepCoil. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Interactions between motor domains in kinesin-14 Ncd - a molecular dynamics study.

Minus-end directed, non-processive kinesin-14 Ncd is a dimeric protein with C-terminally located motor domains (heads). Generation of the power-stroke by Ncd consists of a lever-like rotation of a long superhelical 'stalk' segment while one of the kinesin's heads is bound to the microtubule. The last ∼30 amino acids of Ncd head play a crucial but still poorly understood role in this process. Here, we used accelerated molecular dynamics simulations to explore the conformational dynamics of several systems built upon two crystal structures of Ncd, the asymmetrical T436S mutant in pre-stroke/post-stroke conformations of two partner subunits and the symmetrical wild-type protein in pre-stroke conformation of both subunits. The results revealed a new conformational state forming following the inward motion of the subunits and stabilized with several hydrogen bonds to residues located on the border or within the C-terminal linker, i.e. a modeled extension of the C-terminus by residues 675-683. Forming of this new, compact Ncd conformation critically depends on the length of the C-terminus extending to at least residue 681. Moreover, the associative motion leading to the compact conformation is accompanied by a partial lateral rotation of the stalk. We propose that the stable compact conformation of Ncd may represent an initial state of the working stroke.

Combining Rosetta with molecular dynamics (MD): A benchmark of the MD-based ensemble protein design.

Computational protein design is a set of procedures for computing amino acid sequences that will fold into a specified structure. Rosetta Design, a commonly used software for protein design, allows for the effective identification of sequences compatible with a given backbone structure, while molecular dynamics (MD) simulations can thoroughly sample near-native conformations. We benchmarked a procedure in which Rosetta design is started on MD-derived structural ensembles and showed that such a combined approach generates 20-30% more diverse sequences than currently available methods with only a slight increase in computation time. Importantly, the increase in diversity is achieved without a loss in the quality of the designed sequences assessed by their resemblance to natural sequences. We demonstrate that the MD-based procedure is also applicable to de novo design tasks started from backbone structures without any sequence information. In addition, we implemented a protocol that can be used to assess the stability of designed models and to select the best candidates for experimental validation. In sum our results demonstrate that the MD ensemble-based flexible backbone design can be a viable method for protein design, especially for tasks that require a large pool of diverse sequences.

Variability of the core geometry in parallel coiled-coil bundles.

In protein modelling and design, an understanding of the relationship between sequence and structure is essential. Using parallel, homotetrameric coiled-coil structures as a model system, we demonstrated that machine learning techniques can be used to predict structural parameters directly from the sequence. Coiled coils are regular protein structures, which are of great interest as building blocks for assembling larger nanostructures. They are composed of two or more alpha-helices wrapped around each other to form a supercoiled bundle. The coiled-coil bundles are defined by four basic structural parameters: topology (parallel or antiparallel), radius, degree of supercoiling, and the rotation of helices around their axes. In parallel coiled coils the latter parameter, describing the hydrophobic core packing geometry, was assumed to show little variation. However, we found that subtle differences between structures of this type were not artifacts of structure determination and could be predicted directly from the sequence. Using this information in modelling narrows the structural parameter space that must be searched and thus significantly reduces the required computational time. Moreover, the sequence-structure rules can be used to explain the effects of point mutations and to shed light on the relationship between hydrophobic core architecture and coiled-coil topology.

AlphaFold2 captures the conformational landscape of the HAMP signaling domain.

In this study, we present a conformational landscape of 5000 AlphaFold2 models of the Histidine kinases, Adenyl cyclases, Methyl-accepting proteins and Phosphatases (HAMP) domain, a short helical bundle that transduces signals from sensors to effectors in two-component signaling proteins such as sensory histidine kinases and chemoreceptors. The landscape reveals the conformational variability of the HAMP domain, including rotations, shifts, displacements, and tilts of helices, many combinations of which have not been observed in experimental structures. HAMP domains belonging to a single family tend to occupy a defined region of the landscape, even when their sequence similarity is low, suggesting that individual HAMP families have evolved to operate in a specific conformational range. The functional importance of this structural conservation is illustrated by poly-HAMP arrays, in which HAMP domains from families with opposite conformational preferences alternate, consistent with the rotational model of signal transduction. The only poly-HAMP arrays that violate this rule are predicted to be of recent evolutionary origin and structurally unstable. Finally, we identify a family of HAMP domains that are likely to be dynamic due to the presence of a conserved pi-helical bulge. All code associated with this work, including a tool for rapid sequence-based prediction of the rotational state in HAMP domains, is deposited at https://github.com/labstructbioinf/HAMPpred.

Siah-1-interacting protein regulates mutated huntingtin protein aggregation in Huntington's disease models.

BACKGROUND: Huntington's disease (HD) is a neurodegenerative disorder whereby mutated huntingtin protein (mHTT) aggregates when polyglutamine repeats in the N-terminal of mHTT exceeds 36 glutamines (Q). However, the mechanism of this pathology is unknown. Siah1-interacting protein (SIP) acts as an adaptor protein in the ubiquitination complex and mediates degradation of other proteins. We hypothesized that mHTT aggregation depends on the dysregulation of SIP activity in this pathway in HD. RESULTS: A higher SIP dimer/monomer ratio was observed in the striatum in young YAC128 mice, which overexpress mHTT. We found that SIP interacted with HTT. In a cellular HD model, we found that wildtype SIP increased mHTT ubiquitination, attenuated mHTT protein levels, and decreased HTT aggregation. We predicted mutations that should stabilize SIP dimerization and found that SIP mutant-overexpressing cells formed more stable dimers and had lower activity in facilitating mHTT ubiquitination and preventing exon 1 mHTT aggregation compared with wildtype SIP. CONCLUSIONS: Our data suggest that an increase in SIP dimerization in HD medium spiny neurons leads to a decrease in SIP function in the degradation of mHTT through a ubiquitin-proteasome pathway and consequently an increase in mHTT aggregation. Therefore, SIP could be considered a potential target for anti-HD therapy during the early stage of HD pathology.

TREM2 Gene Compound Heterozygosity in Neurodegenerative Disorders.

BACKGROUND: Homozygous variants of the TREM2 and TYROBP genes have been shown to be causative for multiple bone cysts and neurodegeneration leading to progressive dementia (NHD, Nasu-Hakola disease). OBJECTIVE: To determine if biallelic variants of these genes and/or oligogenic inheritance could be responsible for a wider spectrum of neurodegenerative conditions. METHODS: We analyzed 52 genes associated with neurodegenerative disorders using targeted next generation sequencing in a selected group of 29 patients (n = 14 Alzheimer's disease, n = 8 frontotemporal dementia, n = 7 amyotrophic lateral sclerosis) carrying diverse already determined rare variants in exon 2 of TREM2. Molecular modeling was used to get an insight into the potential effects of the mutation. RESULTS: We identified a novel mutation c.401_406delinsTCTAT; p.(Asp134Valfs*55) in exon 3 of TREM2 in an Alzheimer's disease patient also carrying the p.Arg62His TREM2 variant. Molecular modeling revealed that the identified mutation prevents anchoring of the TREM2 protein in the membrane, leaving the core of the Ig-like domain intact. CONCLUSION: Our results expand the spectrum of neurodegenerative diseases, where the carriers of biallelic mutations in TREM2 have been described for Alzheimer's disease, and highlight the impact of variant burden in other genes on phenotypic heterogeneity.

C8J_1298, a bifunctional thiol oxidoreductase of Campylobacter jejuni, affects Dsb (disulfide bond) network functioning.

Posttranslational generation of disulfide bonds catalyzed by bacterial Dsb (disulfide bond) enzymes is essential for the oxidative folding of many proteins. Although we now have a good understanding of the Escherichia coli disulfide bond formation system, there are significant gaps in our knowledge concerning the Dsb systems of other bacteria, including Campylobacter jejuni, a food-borne, zoonotic pathogen. We attempted to gain a more complete understanding of the process by thorough analysis of C8J_1298 functioning in vitro and in vivo. C8J_1298 is a homodimeric thiol-oxidoreductase present in wild type (wt) cells, in both reduced and oxidized forms. The protein was previously described as a homolog of DsbC, and thus potentially should be active in rearrangement of disulfides. Indeed, biochemical studies with purified protein revealed that C8J_1298 shares many properties with EcDsbC. However, its activity in vivo is dependent on the genetic background, namely, the set of other Dsb proteins present in the periplasm that determine the redox conditions. In wt C. jejuni cells, C8J_1298 potentially works as a DsbG involved in the control of the cysteine sulfenylation level and protecting single cysteine residues from oxidation to sulfenic acid. A strain lacking only C8J_1298 is indistinguishable from the wild type strain by several assays recognized as the criteria to determine isomerization or oxidative Dsb pathways. Remarkably, in C. jejuni strain lacking DsbA1, the protein involved in generation of disulfides, C8J_1298 acts as an oxidase, similar to the homodimeric oxidoreductase of Helicobater pylori, HP0231. In E. coli, C8J_1298 acts as a bifunctional protein, also resembling HP0231. These findings are strongly supported by phylogenetic data. We also showed that CjDsbD (C8J_0565) is a C8J_1298 redox partner.

PiPred - a deep-learning method for prediction of π-helices in protein sequences.

Canonical π-helices are short, relatively unstable secondary structure elements found in proteins. They comprise seven or more residues and are present in 15% of all known protein structures, often in functionally important regions such as ligand- and ion-binding sites. Given their similarity to α-helices, the prediction of π-helices is a challenging task and none of the currently available secondary structure prediction methods tackle it. Here, we present PiPred, a neural network-based tool for predicting π-helices in protein sequences. By performing a rigorous benchmark we show that PiPred can detect π-helices with a per-residue precision of 48% and sensitivity of 46%. Interestingly, some of the α-helices mispredicted by PiPred as π-helices exhibit a geometry characteristic of π-helices. Also, despite being trained only with canonical π-helices, PiPred can identify 6-residue-long α/π-bulges. These observations suggest an even higher effective precision of the method and demonstrate that π-helices, α/π-bulges, and other helical deformations may impose similar constraints on sequences. PiPred is freely accessible at: https://toolkit.tuebingen.mpg.de/#/tools/quick2d . A standalone version is available for download at: https://github.com/labstructbioinf/PiPred , where we also provide the CB6133, CB513, CASP10, and CASP11 datasets, commonly used for training and validation of secondary structure prediction methods, with correctly annotated π-helices.

Crystal structures of nematode (parasitic T. spiralis and free living C. elegans), compared to mammalian, thymidylate synthases (TS). Molecular docking and molecular dynamics simulations in search for nematode-specific inhibitors of TS.

Three crystal structures are presented of nematode thymidylate synthases (TS), including Caenorhabditis elegans (Ce) enzyme without ligands and its ternary complex with dUMP and Raltitrexed, and binary complex of Trichinella spiralis (Ts) enzyme with dUMP. In search of differences potentially relevant for the development of species-specific inhibitors of the nematode enzyme, a comparison was made of the present Ce and Ts enzyme structures, as well as binary complex of Ce enzyme with dUMP, with the corresponding mammalian (human, mouse and rat) enzyme crystal structures. To complement the comparison, tCONCOORD computations were performed to evaluate dynamic behaviors of mammalian and nematode TS structures. Finally, comparative molecular docking combined with molecular dynamics and free energy of binding calculations were carried out to search for ligands showing selective affinity to T. spiralis TS. Despite an overall strong similarity in structure and dynamics of nematode vs mammalian TSs, a pool of ligands demonstrating predictively a strong and selective binding to TsTS has been delimited. These compounds, the E63 family, locate in the dimerization interface of TsTS where they exert species-specific interactions with certain non-conserved residues, including hydrogen bonds with Thr174 and hydrophobic contacts with Phe192, Cys191 and Tyr152. The E63 family of ligands opens the possibility of future development of selective inhibitors of TsTS and effective agents against trichinellosis.

Phosphorylation of thymidylate synthase affects slow-binding inhibition by 5-fluoro-dUMP and N(4)-hydroxy-dCMP.

Endogenous thymidylate synthases, isolated from tissues or cultured cells of the same specific origin, have been reported to show differing slow-binding inhibition patterns. These were reflected by biphasic or linear dependence of the inactivation rate on time and accompanied by differing inhibition parameters. Considering its importance for chemotherapeutic drug resistance, the possible effect of thymidylate synthase inhibition by post-translational modification was tested, e.g. phosphorylation, by comparing sensitivities to inhibition by two slow-binding inhibitors, 5-fluoro-dUMP and N(4)-hydroxy-dCMP, of two fractions of purified recombinant mouse enzyme preparations, phosphorylated and non-phosphorylated, separated by metal oxide/hydroxide affinity chromatography on Al(OH)3 beads. The modification, found to concern histidine residues and influence kinetic properties by lowering Vmax, altered both the pattern of dependence of the inactivation rate on time from linear to biphasic, as well as slow-binding inhibition parameters, with each inhibitor studied. Being present on only one subunit of at least a great majority of phosphorylated enzyme molecules, it probably introduced dimer asymmetry, causing the altered time dependence of the inactivation rate pattern (biphasic with the phosphorylated enzyme) and resulting in asymmetric binding of each inhibitor studied. The latter is reflected by the ternary complexes, stable under denaturing conditions, formed by only the non-phosphorylated subunit of the phosphorylated enzyme with each of the two inhibitors and N(5,10)-methylenetetrahydrofolate. Inhibition of the phosphorylated enzyme by N(4)-hydroxy-dCMP was found to be strongly dependent on [Mg(2+)], cations demonstrated previously to also influence the activity of endogenous mouse TS isolated from tumour cells.