DEMO2: Assemble multi-domain protein structures by coupling analogous template alignments with deep-learning inter-domain restraint prediction

Structure and function of engineered stromal cell-derived factor-1α

To design biologically active, collagen-based scaffolds for bone tissue engineering, we have synthesized chimeric proteins consisting of stromal cell-derived factor-1α (SDF) and the von Willebrand factor A3 collagen-binding domain (CBD). The chimeric proteins were used to evaluate the effect of domain linkage and its order on the structure and function of the SDF and CBD. The structure of the chimeric proteins was analyzed by circular dichroism spectroscopy, while functional analysis was performed by a cell migration assay for the SDF domain and a collagen-binding assay for the CBD domain. Furthermore, computational structural prediction was conducted for the chimeric proteins to examine the consistency with the results of structural and functional analyses. Our structural and functional analyses as well as structural prediction revealed that linking two domains can affect their functions. However, their order had minor effects on the three-dimensional structure of CBD and SDF in the chimeric proteins.

Recent Progress of Protein Tertiary Structure Prediction

The prediction of three-dimensional (3D) protein structure from amino acid sequences has stood as a significant challenge in computational and structural bioinformatics for decades. Recently, the widespread integration of artificial intelligence (AI) algorithms has substantially expedited advancements in protein structure prediction, yielding numerous significant milestones. In particular, the end-to-end deep learning method AlphaFold2 has facilitated the rise of structure prediction performance to new heights, regularly competitive with experimental structures in the 14th Critical Assessment of Protein Structure Prediction (CASP14). To provide a comprehensive understanding and guide future research in the field of protein structure prediction for researchers, this review describes various methodologies, assessments, and databases in protein structure prediction, including traditionally used protein structure prediction methods, such as template-based modeling (TBM) and template-free modeling (FM) approaches; recently developed deep learning-based methods, such as contact/distance-guided methods, end-to-end folding methods, and protein language model (PLM)-based methods; multi-domain protein structure prediction methods; the CASP experiments and related assessments; and the recently released AlphaFold Protein Structure Database (AlphaFold DB). We discuss their advantages, disadvantages, and application scopes, aiming to provide researchers with insights through which to understand the limitations, contexts, and effective selections of protein structure prediction methods in protein-related fields.

Modeling and Analysis of HIV-1 Pol Polyprotein as a Case Study for Predicting Large Polyprotein Structures

Acquired immunodeficiency syndrome (AIDS) is caused by human immunodeficiency virus (HIV). HIV protease, reverse transcriptase, and integrase are targets of current drugs to treat the disease. However, anti-viral drug-resistant strains have emerged quickly due to the high mutation rate of the virus, leading to the demand for the development of new drugs. One attractive target is Gag-Pol polyprotein, which plays a key role in the life cycle of HIV. Recently, we found that a combination of M50I and V151I mutations in HIV-1 integrase can suppress virus release and inhibit the initiation of Gag-Pol autoprocessing and maturation without interfering with the dimerization of Gag-Pol. Additional mutations in integrase or RNase H domain in reverse transcriptase can compensate for the defect. However, the molecular mechanism is unknown. There is no tertiary structure of the full-length HIV-1 Pol protein available for further study. Therefore, we developed a workflow to predict the tertiary structure of HIV-1 NL4.3 Pol polyprotein. The modeled structure has comparable quality compared with the recently published partial HIV-1 Pol structure (PDB ID: 7SJX). Our HIV-1 NL4.3 Pol dimer model is the first full-length Pol tertiary structure. It can provide a structural platform for studying the autoprocessing mechanism of HIV-1 Pol and for developing new potent drugs. Moreover, the workflow can be used to predict other large protein structures that cannot be resolved via conventional experimental methods.

DEMO-EM2: assembling protein complex structures from cryo-EM maps through intertwined chain and domain fitting

The breakthrough in cryo-electron microscopy (cryo-EM) technology has led to an increasing number of density maps of biological macromolecules. However, constructing accurate protein complex atomic structures from cryo-EM maps remains a challenge. In this study, we extend our previously developed DEMO-EM to present DEMO-EM2, an automated method for constructing protein complex models from cryo-EM maps through an iterative assembly procedure intertwining chain- and domain-level matching and fitting for predicted chain models. The method was carefully evaluated on 27 cryo-electron tomography (cryo-ET) maps and 16 single-particle EM maps, where DEMO-EM2 models achieved an average TM-score of 0.92, outperforming those of state-of-the-art methods. The results demonstrate an efficient method that enables the rapid and reliable solution of challenging cryo-EM structure modeling problems.

Recent Advances and Challenges in Protein Structure Prediction

Artificial intelligence has made significant advances in the field of protein structure prediction in recent years. In particular, DeepMind's end-to-end model, AlphaFold2, has demonstrated the capability to predict three-dimensional structures of numerous unknown proteins with accuracy levels comparable to those of experimental methods. This breakthrough has opened up new possibilities for understanding protein structure and function as well as accelerating drug discovery and other applications in the field of biology and medicine. Despite the remarkable achievements of artificial intelligence in the field, there are still some challenges and limitations. In this Review, we discuss the recent progress and some of the challenges in protein structure prediction. These challenges include predicting multidomain protein structures, protein complex structures, multiple conformational states of proteins, and protein folding pathways. Furthermore, we highlight directions in which further improvements can be conducted.

Multi-domain and complex protein structure prediction using inter-domain interactions from deep learning

Accurately capturing domain-domain interactions is key to understanding protein function and designing structure-based drugs. Although AlphaFold2 has made a breakthrough on single domain, it should be noted that the structure modeling for multi-domain protein and complex remains a challenge. In this study, we developed a multi-domain and complex structure assembly protocol, named DeepAssembly, based on domain segmentation and single domain modeling algorithms. Firstly, DeepAssembly uses a population-based evolutionary algorithm to assemble multi-domain proteins by inter-domain interactions inferred from a developed deep learning network. Secondly, protein complexes are assembled by means of domains rather than chains using DeepAssembly. Experimental results show that on 219 multi-domain proteins, the average inter-domain distance precision by DeepAssembly is 22.7% higher than that of AlphaFold2. Moreover, DeepAssembly improves accuracy by 13.1% for 164 multi-domain structures with low confidence deposited in AlphaFold database. We apply DeepAssembly for the prediction of 247 heterodimers. We find that DeepAssembly successfully predicts the interface (DockQ ≥ 0.23) for 32.4% of the dimers, suggesting a lighter way to assemble complex structures by treating domains as assembly units and using inter-domain interactions learned from monomer structures.

Integrating deep learning, threading alignments, and a multi-MSA strategy for high-quality protein monomer and complex structure prediction in CASP15

We report the results of the "UM-TBM" and "Zheng" groups in CASP15 for protein monomer and complex structure prediction. These prediction sets were obtained using the D-I-TASSER and DMFold-Multimer algorithms, respectively. For monomer structure prediction, D-I-TASSER introduced four new features during CASP15: (i) a multiple sequence alignment (MSA) generation protocol that combines multi-source MSA searching and a structural modeling-based MSA ranker; (ii) attention-network based spatial restraints; (iii) a multi-domain module containing domain partition and arrangement for domain-level templates and spatial restraints; (iv) an optimized I-TASSER-based folding simulation system for full-length model creation guided by a combination of deep learning restraints, threading alignments, and knowledge-based potentials. For 47 free modeling targets in CASP15, the final models predicted by D-I-TASSER showed average TM-score 19% higher than the standard AlphaFold2 program. We thus showed that traditional Monte Carlo-based folding simulations, when appropriately coupled with deep learning algorithms, can generate models with improved accuracy over end-to-end deep learning methods alone. For protein complex structure prediction, DMFold-Multimer generated models by integrating a new MSA generation algorithm (DeepMSA2) with the end-to-end modeling module from AlphaFold2-Multimer. For the 38 complex targets, DMFold-Multimer generated models with an average TM-score of 0.83 and Interface Contact Score of 0.60, both significantly higher than those of competing complex prediction tools. Our analyses on complexes highlighted the critical role played by MSA generating, ranking, and pairing in protein complex structure prediction. We also discuss future room for improvement in the areas of viral protein modeling and complex model ranking.

In silico structural elucidation of Nipah virus L protein and targeting RNA-dependent RNA polymerase domain by nucleoside analogs

The large (L) protein of Mononegavirales is a multi-domain protein that performs transcription and genome replication. One of the important domains in L is the RNA-dependent RNA polymerase (RdRp), a promising target for antiviral drugs. In this work, we employed rigorous computational comparative modeling to predict the structure of L protein of Nipah virus (NiV). The RdRp domain was targeted by a panel of nucleotide analogs, previously reported to inhibit different viral RNA polymerases, using molecular docking. Best binder compounds were subjected to molecular dynamics simulation to validate their binding. Molecular mechanics/generalized-born surface area (MM/GBSA) calculations estimated the binding free energy. The predicted model of NiV L has an excellent quality as judged by physics- and knowledge-based validation tests. Galidesivir, AT-9010 and Norov-29 scored the top nucleotide analogs to bind to the RdRp. Their binding free energies obtained by MM/GBSA (-31.01 ± 3.9 to -38.37 ± 4.8 kcal/mol) ranked Norov-29 as the best potential inhibitor. Purine nucleotide analogs are expected to harbor the scaffold for an effective drug against NiV. Finally, this study is expected to provide a start point for medicinal chemistry and drug discovery campaigns toward identification of effective chemotherapeutic agent(s) against NiV.Communicated by Ramaswamy H. Sarma.

Narrow funnel-like interaction energy distribution is an indicator of specific protein interaction partner

Protein interaction networks underlie countless biological mechanisms. However, most protein interaction predictions are based on biological evidence that are biased to well-known protein interaction or physical evidence that exhibits low accuracy for weak interactions and requires high computational power. In this study, a novel method has been suggested to predict protein interaction partners by investigating narrow funnel-like interaction energy distribution. In this study, it was demonstrated that various protein interactions including kinases and E3 ubiquitin ligases have narrow funnel-like interaction energy distribution. To analyze protein interaction distribution, modified scores of iRMS and TM-score are introduced. Then, using these scores, algorithm and deep learning model for prediction of protein interaction partner and substrate of kinase and E3 ubiquitin ligase were developed. The prediction accuracy was similar to or even better than that of yeast two-hybrid screening. Ultimately, this knowledge-free protein interaction prediction method will broaden our understanding of protein interaction networks.

Protein structure and folding pathway prediction based on remote homologs recognition using PAthreader

Recognition of remote homologous structures is a necessary module in AlphaFold2 and is also essential for the exploration of protein folding pathways. Here, we propose a method, PAthreader, to recognize remote templates and explore folding pathways. Firstly, we design a three-track alignment between predicted distance profiles and structure profiles extracted from PDB and AlphaFold DB, to improve the recognition accuracy of remote templates. Secondly, we improve the performance of AlphaFold2 using the templates identified by PAthreader. Thirdly, we explore protein folding pathways based on our conjecture that dynamic folding information of protein is implicitly contained in its remote homologs. The results show that the average accuracy of PAthreader templates is 11.6% higher than that of HHsearch. In terms of structure modelling, PAthreader outperform AlphaFold2 and ranks first on the CAMEO blind test for the latest three months. Furthermore, we predict protein folding pathways for 37 proteins, in which the results of 7 proteins are almost consistent with those of biological experiments, and the other 30 human proteins have yet to be verified by biological experiments, revealing that folding information can be exploited from remote homologous structures.

Structural modelling of human complement FHR1 and two of its synthetic derivatives provides insight into their in-vivo functions

Human complement is the first line of defence against invading pathogens and is involved in tissue homeostasis. Complement-targeted therapies to treat several diseases caused by a dysregulated complement are highly desirable. Despite huge efforts invested in their development, only very few are currently available, and a deeper understanding of the numerous interactions and complement regulation mechanisms is indispensable. Two important complement regulators are human Factor H (FH) and Factor H-related protein 1 (FHR1). MFHR1 and MFHR13, two promising therapeutic candidates based on these regulators, combine the dimerization and C5-regulatory domains of FHR1 with the central C3-regulatory and cell surface-recognition domains of FH. Here, we used AlphaFold2 to model the structure of these two synthetic regulators. Moreover, we used AlphaFold-Multimer (AFM) to study possible interactions of C3 fragments and membrane attack complex (MAC) components C5, C7 and C9 in complex with FHR1, MFHR1, MFHR13 as well as the best-known MAC regulators vitronectin (Vn), clusterin and CD59, whose experimental structures remain undetermined. AFM successfully predicted the binding interfaces of FHR1 and the synthetic regulators with C3 fragments and suggested binding to C3. The models revealed structural differences in binding to these ligands through different interfaces. Additionally, AFM predictions of Vn, clusterin or CD59 with C7 or C9 agreed with previously published experimental results. Because the role of FHR1 as MAC regulator has been controversial, we analysed possible interactions with C5, C7 and C9. AFM predicted interactions of FHR1 with proteins of the terminal complement complex (TCC) as indicated by experimental observations, and located the interfaces in FHR1_1-2 and FHR1_4-5. According to AFM prediction, FHR1 might partially block the C3b binding site in C5, inhibiting C5 activation, and block C5b-7 complex formation and C9 polymerization, with similar mechanisms of action as clusterin and vitronectin. Here, we generate hypotheses and give the basis for the design of rational approaches to understand the molecular mechanism of MAC inhibition, which will facilitate the development of further complement therapeutics.