AlphaFold2 for Protein Structure Prediction: Best Practices and Critical Analyses.

AlphaFold2 (AF2) has emerged in recent years as a groundbreaking innovation that has revolutionized several scientific fields, in particular structural biology, drug design, and the elucidation of disease mechanisms. Many scientists now use AF2 on a daily basis, including non-specialist users. This chapter is aimed at the latter. Tips and tricks for getting the most out of AF2 to produce a high-quality biological model are discussed here. We suggest to non-specialist users how to maintain a critical perspective when working with AF2 models and provide guidelines on how to properly evaluate them. After showing how to perform our own structure prediction using ColabFold, we list several ways to improve AF2 models by adding information that is missing from the original AF2 model. By using software such as AlphaFill to add cofactors and ligands to the models, or MODELLER to add disulfide bridges between cysteines, we guide users to build a high-quality biological model suitable for applications such as drug design, protein interaction, or molecular dynamics studies.

Utilization of an optimized AlphaFold protein model for structure-based design of a selective HDAC11 inhibitor with anti-neuroblastoma activity.

AlphaFold is an artificial intelligence approach for predicting the three-dimensional (3D) structures of proteins with atomic accuracy. One challenge that limits the use of AlphaFold models for drug discovery is the correct prediction of folding in the absence of ligands and cofactors, which compromises their direct use. We have previously described the optimization and use of the histone deacetylase 11 (HDAC11) AlphaFold model for the docking of selective inhibitors such as FT895 and SIS17. Based on the predicted binding mode of FT895 in the optimized HDAC11 AlphaFold model, a new scaffold for HDAC11 inhibitors was designed, and the resulting compounds were tested in vitro against various HDAC isoforms. Compound 5a proved to be the most active compound with an IC50 of 365 nM and was able to selectively inhibit HDAC11. Furthermore, docking of 5a showed a binding mode comparable to FT895 but could not adopt any reasonable poses in other HDAC isoforms. We further supported the docking results with molecular dynamics simulations that confirmed the predicted binding mode. 5a also showed promising activity with an EC50 of 3.6 µM on neuroblastoma cells.

Constructing networks for comparison of collagen types.

Collagens are structural proteins that are predominantly found in the extracellular matrix of multicellular animals, where they are mainly responsible for the stability and structural integrity of various tissues. All collagens contain polypeptide strands (α-chains). There are several types of collagens, some of which differ significantly in form, function, and tissue specificity. Because of their importance in clinical research, they are grouped into subdivisions, the so-called collagen families, and their sequences are often analysed. However, problems arise with highly homologous sequence segments. To increase the accuracy of collagen classification and prediction of their functions, the structure of these collagens and their expression in different tissues could result in a better focus on sequence segments of interest. Here, we analyse collagen families with different levels of conservation. As a result, clusters with high interconnectivity can be found, such as the fibrillar collagens, the COL4 network-forming collagens, and the COL9 FACITs. Furthermore, a large cluster between network-forming, FACIT, and COL28a1 α-chains is formed with COL6a3 as a major hub node. The formation of clusters also signifies, why it is important to always analyse the α-chains and why structural changes can have a wide range of effects on the body.

All-RNA-mediated targeted gene integration in mammalian cells with rationally engineered R2 retrotransposons.

All-RNA-mediated targeted gene integration methods, rendering reduced immunogenicity, effective deliverability with non-viral vehicles, and a low risk of random mutagenesis, are urgently needed for next-generation gene addition technologies. Naturally occurring R2 retrotransposons hold promise in this context due to their site-specific integration profile. Here, we systematically analyzed the biodiversity of R2 elements and screened several R2 orthologs capable of full-length gene insertion in mammalian cells. Robust R2 system gene integration efficiency was attained using combined donor RNA and protein engineering. Importantly, the all-RNA-delivered engineered R2 system showed effective integration activity, with efficiency over 60% in mouse embryos. Unbiased high-throughput sequencing demonstrated that the engineered R2 system exhibited high on-target integration specificity (99%). In conclusion, our study provides engineered R2 tools for applications based on hit-and-run targeted DNA integration and insights for further optimization of retrotransposon systems.

A random mutagenesis screen enriched for missense mutations in bacterial effector proteins.

To remodel their hosts and escape immune defenses, many pathogens rely on large arsenals of proteins (effectors) that are delivered to the host cell using dedicated translocation machinery. Effectors hold significant insight into the biology of both the pathogens that encode them and the host pathways that they manipulate. One of the most powerful systems biology tools for studying effectors is the model organism, Saccharomyces cerevisiae. For many pathogens, the heterologous expression of effectors in yeast is growth inhibitory at a frequency much higher than housekeeping genes, an observation ascribed to targeting conserved eukaryotic proteins. Abrogation of yeast growth inhibition has been used to identify bacterial suppressors of effector activity, host targets, and functional residues and domains within effector proteins. We present here a yeast-based method for enriching for informative, in-frame, missense mutations in a pool of random effector mutants. We benchmark this approach against three effectors from Legionella pneumophila, an intracellular bacterial pathogen that injects a staggering >330 effectors into the host cell. For each protein, we show how in silico protein modeling (AlphaFold2) and missense-directed mutagenesis can be combined to reveal important structural features within effectors. We identify known active site residues within the metalloprotease RavK, the putative active site in SdbB, and previously unidentified functional motifs within the C-terminal domain of SdbA. We show that this domain has structural similarity with glycosyltransferases and exhibits in vitro activity consistent with this predicted function.

Everything AlphaFold tells us about protein knots.

Recent advances in Machine Learning methods in structural biology opened up new perspectives for protein analysis. Utilizing these methods allows us to go beyond the limitations of empirical research, and take advantage of the vast amount of generated data. We use a complete set of potentially knotted protein models identified in all high-quality predictions from the AlphaFold Database to search for any common trends that describe them. We show that the vast majority of knotted proteins have 31 knot and that the presence of knots is preferred in neither Bacteria, Eukaryota, or Archaea domains. On the contrary, the percentage of knotted proteins in any given proteome is around 0.4%, regardless of the taxonomical group. We also verified that the organism's living conditions do not impact the number of knotted proteins in its proteome, as previously expected. We did not encounter an organism without a single knotted protein. What is more, we found four universally present families of knotted proteins in Bacteria, consisting of SAM synthase, and TrmD, TrmH, and RsmE methyltransferases.

Hidden Structural States of Proteins Revealed by Conformer Selection with AlphaFold-NMR.

Recent advances in molecular modeling using deep learning can revolutionize our understanding of dynamic protein structures. NMR is particularly well-suited for determining dynamic features of biomolecular structures. The conventional process for determining biomolecular structures from experimental NMR data involves its representation as conformation-dependent restraints, followed by generation of structural models guided by these spatial restraints. Here we describe an alternative approach: generating a distribution of realistic protein conformational models using artificial intelligence-(AI-) based methods and then selecting the sets of conformers that best explain the experimental data. We applied this conformational selection approach to redetermine the solution NMR structure of the enzyme Gaussia luciferase. First, we generated a diverse set of conformer models using AlphaFold2 (AF2) with an enhanced sampling protocol. The models that best-fit NOESY and chemical shift data were then selected with a Bayesian scoring metric. The resulting models include features of both the published NMR structure and the standard AF2 model generated without enhanced sampling. This "AlphaFold-NMR" protocol also generated an alternative "open" conformational state that fits nearly as well to the overall NMR data but accounts for some NOESY data that is not consistent with first "closed" conformational state; while other NOESY data consistent with this second state are not consistent with the first conformational state. The structure of this "open" structural state differs from that of the "closed" state primarily by the position of a thumb-shaped loop between α-helices H5 and H6, revealing a cryptic surface pocket. These alternative conformational states of Gluc are supported by "double recall" analysis of NOESY data and AF2 models. Additional structural states are also indicated by backbone chemical shift data indicating partially-disordered conformations for the C-terminal segment. Considered as a multistate ensemble, these multiple states of Gluc together fit the NOESY and chemical shift data better than the "restraint-based" NMR structure and provide novel insights into its structure-dynamic-function relationships. This study demonstrates the potential of AI-based modeling with enhanced sampling to generate conformational ensembles followed by conformer selection with experimental data as an alternative to conventional restraint satisfaction protocols for protein NMR structure determination.

δ-Conotoxin Structure Prediction and Analysis through Large-Scale Comparative and Deep Learning Modeling Approaches.

The δ-conotoxins, a class of peptides produced in the venom of cone snails, are of interest due to their ability to inhibit the inactivation of voltage-gated sodium channels causing paralysis and other neurological responses, but difficulties in their isolation and synthesis have made structural characterization challenging. Taking advantage of recent breakthroughs in computational algorithms for structure prediction that have made modeling especially useful when experimental data is sparse, this work uses both the deep-learning-based algorithm AlphaFold and comparative modeling method RosettaCM to model and analyze 18 previously uncharacterized δ-conotoxins derived from piscivorous, vermivorous, and molluscivorous cone snails. The models provide useful insights into the structural aspects of these peptides and suggest features likely to be significant in influencing their binding and different pharmacological activities against their targets, with implications for drug development. Additionally, the described protocol provides a roadmap for the modeling of similar disulfide-rich peptides by these complementary methods.

Using site-directed mutagenesis to further the understanding of insulin receptor-insulin like growth factor-1 receptor heterodimer structure.

Type 2 diabetes is characterised by the disruption of insulin and insulin-like growth factor (IGF) signalling. The key hubs of these signalling cascades - the Insulin receptor (IR) and Insulin-like growth factor 1 receptor (IGF1R) - are known to form functional IR-IGF1R hybrid receptors which are insulin resistant. However, the mechanisms underpinning IR-IGF1R hybrid formation are not fully understood, hindering the ability to modulate this for future therapies targeting this receptor. To pinpoint suitable sites for intervention, computational hotspot prediction was utilised to identify promising epitopes for targeting with point mutagenesis. Specific IGF1R point mutations F450A, R391A and D555A show reduced affinity of the hybrid receptor in a BRET based donor-saturation assay, confirming hybrid formation could be modulated at this interface. These data provide the basis for rational design of more effective hybrid receptor modulators, supporting the prospect of identifying a small molecule that specifically interacts with this target.

Identification, characterization and classification of prokaryotic nucleoid-associated proteins.

Common throughout life is the need to compact and organize the genome. Possible mechanisms involved in this process include supercoiling, phase separation, charge neutralization, macromolecular crowding, and nucleoid-associated proteins (NAPs). NAPs are special in that they can organize the genome at multiple length scales, and thus are often considered as the architects of the genome. NAPs shape the genome by either bending DNA, wrapping DNA, bridging DNA, or forming nucleoprotein filaments on the DNA. In this mini-review, we discuss recent advancements of unique NAPs with differing architectural properties across the tree of life, including NAPs from bacteria, archaea, and viruses. To help the characterization of NAPs from the ever-increasing number of metagenomes, we recommend a set of cheap and simple in vitro biochemical assays that give unambiguous insights into the architectural properties of NAPs. Finally, we highlight and showcase the usefulness of AlphaFold in the characterization of novel NAPs.

LazyAF, a pipeline for accessible medium-scale in silico prediction of protein-protein interactions.

Artificial intelligence has revolutionized the field of protein structure prediction. However, with more powerful and complex software being developed, it is accessibility and ease of use rather than capability that is quickly becoming a limiting factor to end users. LazyAF is a Google Colaboratory-based pipeline which integrates the existing ColabFold BATCH software to streamline the process of medium-scale protein-protein interaction prediction. LazyAF was used to predict the interactome of the 76 proteins encoded on the broad-host-range multi-drug resistance plasmid RK2, demonstrating the ease and accessibility the pipeline provides.

AlphaFold-guided redesign of a plant pectin methylesterase inhibitor for broad-spectrum disease resistance.

Plant cell walls are a critical site where plants and pathogens continuously struggle for physiological dominance. Here we show that dynamic remodeling of pectin methylesterification of plant cell walls is a component of the physiological and co-evolutionary struggles between hosts and pathogens. A Phytophthora sojae secreted pectin methylesterase (PsPME1) decreases the degree of pectin methylesterification, thus synergizing with an endo-polygalacturonase (PsPG1) to weaken plant cell walls. To counter PsPME1-mediated susceptibility, a plant-derived pectin methylesterase inhibitor protein, GmPMI1, protects pectin to maintain a high methylesterification status. GmPMI1 protects plant cell walls from enzymatic degradation by inhibiting both soybean and P. sojae pectin methylesterases during infection. However, constitutive expression of GmPMI1 disrupted the tradeoff between host growth and defense responses. So, we used AlphaFold structure tools to design a modified form of GmPMI1 (GmPMI1R) which specifically targets and inhibits pectin methylesterases secreted from pathogens but not from the plants. Transient expression of GmPMI1R enhanced plant resistance to oomycetes and fungal pathogens. In summary, our work highlights biochemical modification of the cell wall as an important focal point in the physiological and co-evolutionary conflict between the hosts and microbes and serves as an important proof-of-concept for how rapid advancements in AI-driven structure-based tools can accelerate the prediction of new strategies for plant protection.

C(P)XCG Proteins of Haloferax volcanii with Predicted Zinc Finger Domains: The Majority Bind Zinc, but Several Do Not.

In recent years, interest in very small proteins (µ-proteins) has increased significantly, and they were found to fulfill important functions in all prokaryotic and eukaryotic species. The halophilic archaeon Haloferax volcanii encodes about 400 µ-proteins of less than 70 amino acids, 49 of which contain at least two C(P)XCG motifs and are, thus, predicted zinc finger proteins. The determination of the NMR solution structure of HVO_2753 revealed that only one of two predicted zinc fingers actually bound zinc, while a second one was metal-free. Therefore, the aim of the current study was the homologous production of additional C(P)XCG proteins and the quantification of their zinc content. Attempts to produce 31 proteins failed, underscoring the particular difficulties of working with µ-proteins. In total, 14 proteins could be produced and purified, and the zinc content was determined. Only nine proteins complexed zinc, while five proteins were zinc-free. Three of the latter could be analyzed using ESI-MS and were found to contain another metal, most likely cobalt or nickel. Therefore, at least in haloarchaea, the variability of predicted C(P)XCG zinc finger motifs is higher than anticipated, and they can be metal-free, bind zinc, or bind another metal. Notably, AlphaFold2 cannot correctly predict whether or not the four cysteines have the tetrahedral configuration that is a prerequisite for metal binding.

Novel Insights into Stereoselective Reproductive Toxicity Induced by Mefentrifluconazole in Earthworms (Eisenia fetida): First Report of Estrogenic Effects.

Widespread use of the new chiral triazole fungicide mefentrifluconazole (MFZ) poses a threat to soil organisms. Although triazole fungicides have been reported to induce reproductive disorders in vertebrates, significant research gaps remain regarding their impact on the reproductive health of soil invertebrates. Here, reproduction-related toxicity end points were explored in earthworms (Eisenia fetida) after exposure for 28 d to soil containing 4 mg/kg racemic MFZ, R-(-)-MFZ, and S-(+)-MFZ. The S-(+)-MFZ treatment resulted in a more pronounced reduction in the number of cocoons and juveniles compared to R-(-)-MFZ treatment, and the expression of annetocin gene was significantly downregulated following exposure to both enantiomers. This reproductive toxicity has been attributed to the disruption of ovarian steroidogenesis at the transcriptional level. Further studies revealed that MFZ enantiomers were able to activate the estrogen receptor (ER). Indirect evidence for this estrogenic effect is provided by the introduction of 17ß-estradiol, which also induces reproductive disorders through ER activation.

Structures prediction and replica exchange molecular dynamics simulations of α-synuclein: A case study for intrinsically disordered proteins.

In recent years, a variety of three-dimensional structure prediction tools, including AlphaFold2, AlphaFold3, I-TASSER, C-I-TASSER, Phyre2, ESMFold, and RoseTTAFold, have been employed in the investigation of intrinsically disordered proteins. However, a comprehensive validation of these tools specifically for intrinsically disordered proteins has yet to be conducted. In this study, we utilize AlphaFold2, AlphaFold3, I-TASSER, C-I-TASSER, Phyre2, ESMFold, and RoseTTAFold to predict the structure of a model intrinsically disordered α-synuclein protein. Additionally, extensive replica exchange molecular dynamics simulations of the intrinsically disordered protein are conducted. The resulting structures from both structure prediction tools and replica exchange molecular dynamics simulations are analyzed for radius of gyration, secondary and tertiary structure properties, as well as Cα and Hα chemical shift values. A comparison of the obtained results with experimental data reveals that replica exchange molecular dynamics simulations provide results in excellent agreement with experimental observations. However, none of the structure prediction tools utilized in this study can fully capture the structural characteristics of the model intrinsically disordered protein. This study shows that a cluster of ensembles are required for intrinsically disordered proteins. Artificial-intelligence based structure prediction tools such as AlphaFold3 and C-I-TASSER could benefit from stochastic sampling or Monte Carlo simulations for generating an ensemble of structures for intrinsically disordered proteins.

GATSol, an enhanced predictor of protein solubility through the synergy of 3D structure graph and large language modeling.

BACKGROUND: Protein solubility is a critically important physicochemical property closely related to protein expression. For example, it is one of the main factors to be considered in the design and production of antibody drugs and a prerequisite for realizing various protein functions. Although several solubility prediction models have emerged in recent years, many of these models are limited to capturing information embedded in one-dimensional amino acid sequences, resulting in unsatisfactory predictive performance. RESULTS: In this study, we introduce a novel Graph Attention network-based protein Solubility model, GATSol, which represents the 3D structure of proteins as a protein graph. In addition to the node features of amino acids extracted by the state-of-the-art protein large language model, GATSol utilizes amino acid distance maps generated using the latest AlphaFold technology. Rigorous testing on independent eSOL and the Saccharomyces cerevisiae test datasets has shown that GATSol outperforms most recently introduced models, especially with respect to the coefficient of determination R2, which reaches 0.517 and 0.424, respectively. It outperforms the current state-of-the-art GraphSol by 18.4% on the S. cerevisiae_test set. CONCLUSIONS: GATSol captures 3D dimensional features of proteins by building protein graphs, which significantly improves the accuracy of protein solubility prediction. Recent advances in protein structure modeling allow our method to incorporate spatial structure features extracted from predicted structures into the model by relying only on the input of protein sequences, which simplifies the entire graph neural network prediction process, making it more user-friendly and efficient. As a result, GATSol may help prioritize highly soluble proteins, ultimately reducing the cost and effort of experimental work. The source code and data of the GATSol model are freely available at https://github.com/binbinbinv/GATSol .

AlphaFold2 as a replacement for solution NMR structure determination of small proteins: Not so fast!

The determination of a protein's structure is often a first step towards the development of a mechanistic understanding of its function. Considerable advances in computational protein structure prediction have been made in recent years, with AlphaFold2 (AF2) emerging as the primary tool used by researchers for this purpose. While AF2 generally predicts accurate structures of folded proteins, we present here a case where AF2 incorrectly predicts the structure of a small, folded and compact protein with high confidence. This protein, pro-interleukin-18 (pro-IL-18), is the precursor of the cytokine IL-18. Interestingly, the structure of pro-IL-18 predicted by AF2 matches that of the mature cytokine, and not the corresponding experimentally determined structure of the pro-form of the protein. Thus, while computational structure prediction holds immense promise for addressing problems in protein biophysics, there is still a need for experimental structure determination, even in the context of small well-folded, globular proteins.

Assessing the functional roles of coevolving PHD finger residues.

Although in silico folding based on coevolving residue constraints in the deep-learning era has transformed protein structure prediction, the contributions of coevolving residues to protein folding, stability, and other functions in physical contexts remain to be clarified and experimentally validated. Herein, the PHD finger module, a well-known histone reader with distinct subtypes containing subtype-specific coevolving residues, was used as a model to experimentally assess the contributions of coevolving residues and to clarify their specific roles. The results of the assessment, including proteolysis and thermal unfolding of wildtype and mutant proteins, suggested that coevolving residues have varying contributions, despite their large in silico constraints. Residue positions with large constraints were found to contribute to stability in one subtype but not others. Computational sequence design and generative model-based energy estimates of individual structures were also implemented to complement the experimental assessment. Sequence design and energy estimates distinguish coevolving residues that contribute to folding from those that do not. The results of proteolytic analysis of mutations at positions contributing to folding were consistent with those suggested by sequence design and energy estimation. Thus, we report a comprehensive assessment of the contributions of coevolving residues, as well as a strategy based on a combination of approaches that should enable detailed understanding of the residue contributions in other large protein families.

Discovery of periplasmic solute binding proteins with specificity for ketone bodies: ß-hydroxybutyrate binding proteins.

Polyhydroxyalkanoates are a class of biodegradable, thermoplastic polymers which represent a major carbon source for various bacteria. Proteins which mediate the translocation of polyhydroxyalkanoate breakdown products, such as ß-hydroxybutyrate (BHB)-a ketone body which in humans serves as an important biomarker, have not been well characterized. In our investigation to screen a solute-binding protein (SBP) which can act as a suitable recognition element for BHB, we uncovered insights at the intersection of bacterial metabolism and diagnostics. Herein, we identify SBPs associated with putative ATP-binding cassette transporters that specifically recognize BHB, with the potential to serve as recognition elements for continuous quantification of this analyte. Through bioinformatic analysis, we identified candidate SBPs from known metabolizers of polyhydroxybutyrate-including proteins from Cupriavidus necator, Ensifer meliloti, Paucimonas lemoignei, and Thermus thermophilus. After recombinant expression in Escherichia coli, we demonstrated with intrinsic tryptophan fluorescence spectroscopy that four candidate proteins interacted with BHB, ranging from nanomolar to micromolar affinity. Tt.2, an intrinsically thermostable protein from Thermus thermophilus, was observed to have the tightest binding and specificity for BHB, which was confirmed by isothermal calorimetry. Structural analyses facilitated by AlphaFold2, along with molecular docking and dynamics simulations, were used to hypothesize key residues in the binding pocket and to model the conformational dynamics of substrate unbinding. Overall, this study provides strong evidence identifying the cognate ligands of SBPs which we hypothesize to be involved in prokaryotic cellular translocation of polyhydroxyalkanoate breakdown products, while highlighting these proteins' promising biotechnological application.