Quantitative fragmentomics allow affinity mapping of interactomes

Holdup Multiplex Assay for High-Throughput Measurement of Protein-Ligand Affinity Constants Using a Mass Spectrometry Readout

The accurate description and subsequent modeling of protein interactomes require quantification of their affinities at the proteome-wide scale. Here we develop and validate the Holdup Multiplex, a versatile assay with a mass spectrometry (MS) readout for profiling the affinities of a protein for large pools of peptides. The method can precisely quantify, in one single run, thousands of affinity constants over several orders of magnitude. The throughput, dynamic range, and sensitivity can be pushed to the performance limit of the MS readout. We applied the Holdup Multiplex to quantify in a few sample runs the affinities of the 14-3-3s, phosphoreader proteins highly abundant in humans, for 1000 different phosphopeptides. The seven human 14-3-3 isoforms were found to display similar specificities but staggered affinities, with 14-3-3γ being always the best binder and 14-3-3ε and σ being the weakest. Hundreds of new 14-3-3 binding sites were identified. We also identified dozens of 14-3-3 binding sites, some intervening in key signaling pathways, that were either stabilized or destabilized by the phytotoxin Fusicoccin-A. The results were corroborated by X-ray crystallography. Finally, we demonstrated the transferability of the Holdup Multiplex by quantifying the interactions of a PDZ domain for 5400 PBM peptides at once. The approach is applicable to any category of protein-binding ligands that can be quantifiable by mass spectrometry.

Accurate sequence-to-affinity models for SH2 domains from multi-round peptide binding assays coupled with free-energy regression

Short linear peptide motifs play important roles in cell signaling. They can act as modification sites for enzymes and as recognition sites for peptide binding domains. SH2 domains bind specifically to tyrosine-phosphorylated proteins, with the affinity of the interaction depending strongly on the flanking sequence. Quantifying this sequence specificity is critical for deciphering phosphotyrosine-dependent signaling networks. In recent years, protein display technologies and deep sequencing have allowed researchers to profile SH2 domain binding across thousands of candidate ligands. Here, we present a concerted experimental and computational strategy that improves the predictive power of SH2 specificity profiling. Through multi-round affinity selection and deep sequencing with large randomized phosphopeptide libraries, we produce suitable data to train an additive binding free energy model that covers the full theoretical ligand sequence space. Our models can be used to predict signaling network connectivity and the impact of missense variants in phosphoproteins on SH2 binding.

Why HPV16? Why, now, HPV42? How the discovery of HPV42 in rare cancers provides an opportunity to challenge our understanding about the transition between health and disease for common members of the healthy microbiota

In 2022, a bioinformatic, agnostic approach identified HPV42 as causative agent of a rare cancer, later confirmed experimentally. This unexpected association offers an opportunity to reconsider our understanding about papillomavirus infections and cancers. We have expanded our knowledge about the diversity of papillomaviruses and the diseases they cause. Yet, we still lack answers to fundamental questions, such as what makes HPV16 different from the closely related HPV31 or HPV33; or why the very divergent HPV13 and HPV32 cause focal epithelial hyperplasia, while HPV6 or HPV42 do not, despite their evolutionary relatedness. Certain members of the healthy skin microbiota are associated to rare clinical conditions. We propose that a focus on cellular phenotypes, most often transient and influenced by intrinsic and extrinsic factors, may help understand the continuum between health and disease. A conceptual switch is required towards an interpretation of biology as a diversity of states connected by transition probabilities, rather than quasi-deterministic programs. Under this perspective, papillomaviruses may only trigger malignant transformation when specific viral genotypes interact with precise cellular states. Drawing on Canguilhem's concepts of normal and pathological, we suggest that understanding the transition between fluid cellular states can illuminate how commensal-like infections transition from benign to malignant.

Structural biology of the human papillomavirus

Human papillomavirus (HPV), known for its oncogenic properties, is the primary cause of cervical cancer and significantly contributes to mortality rates. It also plays a considerable role in the globally rising incidences of head and neck cancers. These cancers pose a substantial health burden worldwide. Current limitations in diagnostic and treatment strategies, along with inadequate coverage of preventive vaccines in low- and middle-income countries, hinder the progress toward the World Health Organization (WHO) HPV prevention and control targets set for 2030. In response to these challenges, extensive research in structural virology has explored the properties of HPV proteins, yielding crucial insights into the mechanisms of HPV infection that are important for the development of prevention and therapeutic strategies. This review highlights recent advances in understanding the structures of HPV proteins and discusses achievements and future opportunities for HPV vaccine development.

HPV-18 E6 enhances the interaction between EMILIN2 and SNX27 to promote WNT signaling

Oncogenic HPV E6 proteins have a PDZ-binding motif (PBM) which plays important roles in both the viral life cycle and tumor development. The PBM confers interaction with a large number of different PDZ domain-containing substrates, one of which is Sorting Nexin 27. This protein is part of the retromer complex and plays an important role in endocytic sorting pathways. It has been shown that at least two SNX27 interacting partners, GLUT1 and TANC2, are aberrantly trafficked due to the E6 PBM-dependent interaction with SNX27. To investigate further which other components of the endocytic trafficking pathway might be affected by the SNX27-HPV E6 interaction, we analyzed the SNX27 proteome interaction profile in a previously described HeLa cell line expressing GFP-SNX27, both in the presence and absence of the HPV-18 E6 oncoprotein. In this study, we identify a novel interacting partner of SNX27, secreted glycoprotein EMILIN2, whose release is blocked by HPV18 E6 in a PBM-dependent manner. Mechanistically, E6 can block EMILIN2 interaction with the WNT1 ligand, thereby enhancing WNT1 signaling and promoting cell proliferation.

IMPORTANCE

This study demonstrates that HPV E6 blocks EMILIN2 inhibition of WNT1 signaling, thereby enhancing cell proliferation in HPV-positive tumor cells. This involves a novel mechanism whereby the E6 PBM actually contributes toward enhancing the interaction between SNX27 and EMILIN2, suggesting that the mode of recognition of SNX27 by E6 and EMILIN2 is different. This is the first example of the E6 PBM altering a PDZ domain-containing protein to enhance potential substrate recognition.

Uncovering the BIN1-SH3 interactome underpinning centronuclear myopathy

Truncation of the protein-protein interaction SH3 domain of the membrane remodeling Bridging Integrator 1 (BIN1, Amphiphysin 2) protein leads to centronuclear myopathy. Here, we assessed the impact of a set of naturally observed, previously uncharacterized BIN1 SH3 domain variants using conventional in vitro and cell-based assays monitoring the BIN1 interaction with dynamin 2 (DNM2) and identified potentially harmful ones that can be also tentatively connected to neuromuscular disorders. However, SH3 domains are typically promiscuous and it is expected that other, so far unknown partners of BIN1 exist besides DNM2, that also participate in the development of centronuclear myopathy. In order to shed light on these other relevant interaction partners and to get a holistic picture of the pathomechanism behind BIN1 SH3 domain variants, we used affinity interactomics. We identified hundreds of new BIN1 interaction partners proteome-wide, among which many appear to participate in cell division, suggesting a critical role of BIN1 in the regulation of mitosis. Finally, we show that the identified BIN1 mutations indeed cause proteome-wide affinity perturbation, signifying the importance of employing unbiased affinity interactomic approaches.

Phosphorylation Code of Human Nucleophosmin Includes Four Cryptic Sites for Hierarchical Binding of 14-3-3 Proteins

Nucleophosmin (NPM1) is the 46th most abundant human protein with many functions whose dysregulation leads to various cancers. Pentameric NPM1 resides in the nucleolus but can also shuttle to the cytosol. NPM1 is regulated by multisite phosphorylation, yet molecular consequences of site-specific NPM1 phosphorylation remain elusive. Here we identify four 14-3-3 protein binding sites in NPM1 concealed within its oligomerization and α-helical C-terminal domains that are found phosphorylated in vivo. By combining mutagenesis, in-cell phosphorylation and PermaPhos technology for site-directed incorporation of a non-hydrolyzable phosphoserine mimic, we show how phosphorylation promotes NPM1 monomerization and partial unfolding, to recruit 14-3-3 dimers with low-micromolar affinity. Using fluorescence anisotropy we quantified pairwise interactions of all seven human 14-3-3 isoforms with four recombinant NPM1 phosphopeptides and assessed their druggability by fusicoccin. This revealed a complex hierarchy of 14-3-3 affinities toward the primary (S48, S293) and secondary (S106, S260) sites, differentially modulated by the small molecule. As three of these 14-3-3 binding phosphosites in NPM1 reside within signal sequences, this work suggests a mechanism of NPM1 regulation by which NPM1 phosphorylation can promote 14-3-3 binding to affect NPM1 shuttling between cell compartments. It also provides further evidence that phosphorylation-induced structural rearrangements of globular proteins serve to expose otherwise cryptic 14-3-3-binding sites that are important for cellular function.

PDZome-wide and structural characterization of the PDZ-binding motif of VANGL2

VANGL2 is a core component of the non-canonical Wnt/Planar Cell Polarity signaling pathway that uses its highly conserved carboxy-terminal type 1 PDZ-binding motif (PBM) to bind a variety of PDZ proteins. In this study, we characterize and quantitatively assess the largest VANGL2 PDZome-binding profile documented so far, using orthogonal methods. The results of our holdup approach support VANGL2 interactions with a large panel of both long-recognized and unprecedented PDZ domains. Truncation and point mutation analyses of the VANGL2 PBM establish that, beyond the strict requirement of the P-0 / V521 and P-2 / T519 amino acids, upstream residues, including E518, Q516 and R514 at, respectively, P-3, P-5 and P-7 further contribute to the robustness of VANGL2 interactions with two distinct PDZ domains, SNX27 and SCRIBBLE-PDZ3. In agreement with these data, incremental amino-terminal deletions of the VANGL2 PBM causes its overall affinity to progressively decline. Moreover, the holdup data establish that the PDZome binding repertoire of VANGL2 starts to diverge significantly with the truncation of E518. A structural analysis of the SYNJ2BP-PDZ/VANGL2 interaction with truncated PBMs identifies a major conformational change in the binding direction of the PBM peptide after the P-2 position. Finally, we report that the PDZome binding profile of VANGL2 is dramatically rearranged upon phosphorylation of S517, T519 and S520. Our crystallographic approach illustrates how SYNJ2BP accommodates a S520-phosphorylated PBM peptide through the ideal positioning of two basic residues, K48 and R86. Altogether our data provides a comprehensive view of the VANGL2 PDZ network and how this network specifically responds to the post-translation modification of distinct PBM residues. These findings should prove useful in guiding future functional and molecular studies of the key PCP component VANGL2.

"It's Only a Model": When Protein Structure Predictions Need Experimental Validation, the Case of the HTLV-1 Tax Protein

Human T-cell Leukemia Virus type 1 (HTLV-1) is a human retrovirus responsible for leukaemia in 5 to 10% of infected individuals. Among the viral proteins, Tax has been described as directly involved in virus-induced leukemogenesis. Tax is therefore an interesting therapeutic target. However, its 3D structure is still unknown and this hampers the development of drug-design-based therapeutic strategies. Several algorithms are available that can be used to predict the structure of proteins, particularly with the recent appearance of artificial intelligence (AI)-driven pipelines. Here, we review how the structure of Tax is predicted by several algorithms using distinct modelling strategies. We discuss the consequences for the understanding of Tax structure/function relationship, and more generally for the use of structure models for modular and/or flexible proteins, which are frequent in retroviruses.

Fast and scalable querying of eukaryotic linear motifs with gget elm

MOTIVATION

Eukaryotic linear motifs (ELMs), or Short Linear Motifs, are protein interaction modules that play an essential role in cellular processes and signaling networks and are often involved in diseases like cancer. The ELM database is a collection of manually curated motif knowledge from scientific papers. It has become a crucial resource for investigating motif biology and recognizing candidate ELMs in novel amino acid sequences. Users can search amino acid sequences or UniProt Accessions on the ELM resource web interface. However, as with many web services, there are limitations in the swift processing of large-scale queries through the ELM web interface or API calls, and, therefore, integration into protein function analysis pipelines is limited.

RESULTS

To allow swift, large-scale motif analyses on protein sequences using ELMs curated in the ELM database, we have extended the gget suite of Python and command line tools with a new module, gget elm, which does not rely on the ELM server for efficiently finding candidate ELMs in user-submitted amino acid sequences and UniProt Accessions. gget elm increases accessibility to the information stored in the ELM database and allows scalable searches for motif-mediated interaction sites in the amino acid sequences.

AVAILABILITY AND IMPLEMENTATION

The manual and source code are available at https://github.com/pachterlab/gget.

Phosphorylation code of human nucleophosmin includes four cryptic sites for hierarchical binding of 14-3-3 proteins

Nucleophosmin (NPM1) is the 46th most abundant human protein with many functions whose dysregulation leads to various cancers. Pentameric NPM1 resides in the nucleolus but can also shuttle to the cytosol. NPM1 is regulated by multisite phosphorylation, yet molecular consequences of site-specific NPM1 phosphorylation remain elusive. Here we identify four 14-3-3 protein binding sites in NPM1 concealed within its oligomerization and α-helical C-terminal domains that are found phosphorylated in vivo. By combining mutagenesis, in-cell phosphorylation and PermaPhos technology for site-directed incorporation of a non-hydrolyzable phosphoserine mimic, we show how phosphorylation promotes NPM1 monomerization and partial unfolding, to recruit 14-3-3 dimers with low-micromolar affinity. Using fluorescence anisotropy we quantified pairwise interactions of all seven human 14-3-3 isoforms with four recombinant NPM1 phosphopeptides and assessed their druggability by fusicoccin. This revealed a complex hierarchy of 14-3-3 affinities toward the primary (S48, S293) and secondary (S106, S260) sites, differentially modulated by the small molecule. As three of these 14-3-3 binding phospho-sites in NPM1 reside within signal sequences, this work highlights a key mechanism of NPM1 regulation by which NPM1 phosphorylation promotes 14-3-3 binding to control nucleocytoplasmic shuttling. It also provides further evidence that phosphorylation-induced structural rearrangements of globular proteins serve to expose otherwise cryptic 14-3-3-binding sites that are important for cellular function.

Comparative analysis of PDZ-binding motifs in the diacylglycerol kinase family

Diacylglycerol kinases (DGKs) control local and temporal amounts of diacylglycerol (DAG) and phosphatidic acid (PA) by converting DAG to PA through phosphorylation in cells. Certain DGK enzymes possess C-terminal sequences that encode potential PDZ-binding motifs (PBMs), which could be involved in their recruitment into supramolecular signaling complexes. In this study, we used two different interactomic approaches, quantitative native holdup (nHU) and qualitative affinity purification (AP), both coupled to mass spectrometry (MS) to investigate the PDZ partners associated with the potential PBMs of DGKs. Complementing these results with site-specific affinity interactomic data measured on isolated PDZ domain fragments and PBM motifs, as well as evolutionary conservation analysis of the PBMs of DGKs, we explored functional differences within different DGK groups. All our results indicate that putative PBM sequences of type II enzymes, namely DGKδ, DGKη, and DGKκ, are likely to be nonfunctional. In contrast, type IV enzymes, namely DGKζ and DGKι, possess highly promiscuous PBMs that interact with a set of PDZ proteins with very similar affinity interactomes. The combination of various interactomic assays and evolutionary analyses provides a useful strategy for identifying functional domains and motifs within diverse enzyme families.

Viral manipulation of cell polarity signalling

Cell polarity refers to the asymmetric distribution of biomacromolecules that enable the correct orientation of a cell in a particular direction. It is thus an essential component for appropriate tissue development and function. Viral infections can lead to dysregulation of polarity. This is associated with a poor prognosis due to viral interference with core cell polarity regulatory scaffolding proteins that often feature PDZ (PSD-95, DLG, and ZO-1) domains including Scrib, Dlg, Pals1, PatJ, Par3 and Par6. PDZ domains are also promiscuous, binding to several different partners through their C-terminal region which contain PDZ-binding motifs (PBM). Numerous viruses encode viral effector proteins that target cell polarity regulators for their benefit and include papillomaviruses, flaviviruses and coronaviruses. A better understanding of the mechanisms of action utilised by viral effector proteins to subvert host cell polarity sigalling will provide avenues for future therapeutic intervention, while at the same time enhance our understanding of cell polarity regulation and its role tissue homeostasis.

The Human PDZome 2.0: Characterization of a New Resource to Test for PDZ Interactions by Yeast Two-Hybrid

PSD95-disc large-zonula occludens (PDZ) domains are globular modules of 80-90 amino acids that co-evolved with multicellularity. They commonly bind to carboxy-terminal sequences of a plethora of membrane-associated proteins and influence their trafficking and signaling. We previously built a PDZ resource (PDZome) allowing us to unveil human PDZ interactions by Yeast two-hybrid. Yet, this resource is incomplete according to the current knowledge on the human PDZ proteome. Here we built the PDZome 2.0 library for Yeast two-hybrid, based on a PDZ library manually curated from online resources. The PDZome2.0 contains 305 individual clones (266 PDZ domains in isolation and 39 tandems), for which all boundaries were designed based on available PDZ structures. Using as bait the E6 oncoprotein from HPV16, a known promiscuous PDZ interactor, we show that PDZome 2.0 outperforms the previous resource.

The next wave of interactomics: Mapping the SLiM-based interactions of the intrinsically disordered proteome

Short linear motifs (SLiMs) are a unique and ubiquitous class of protein interaction modules that perform key regulatory functions and drive dynamic complex formation. For decades, interactions mediated by SLiMs have accumulated through detailed low-throughput experiments. Recent methodological advances have opened this previously underexplored area of the human interactome to high-throughput protein-protein interaction discovery. In this article, we discuss that SLiM-based interactions represent a significant blind spot in the current interactomics data, introduce the key methods that are illuminating the elusive SLiM-mediated interactome of the human cell on a large scale, and discuss the implications for the field.

Phosphorylation-dependent recognition of diverse protein targets by the cryptic GK domain of MAGI MAGUKs

Dynamic signal transduction requires the rapid assembly and disassembly of signaling complexes, often mediated by phosphoprotein binding modules. The guanylate kinase-like (GK) domain of the membrane-associated guanylate kinases (MAGUKs) is such a module orchestrating signaling at cellular junctions. The MAGI subfamily of MAGUKs contains a truncated GK domain with unknown structure and function, although they participate in diverse physiological and pathological processes. Here, we demonstrate that the truncated GK domain of MAGI2 interacts with its adjacent PDZ0 domain to form a structural supramodule capable of recognizing phosphoproteins. A conserved phosphorylation-dependent binding motif for PDZ0-GK is delineated, which leads to identification of a set of previously unknown binding partners. We explore the structure and function of the MAGI2-target complex with an inhibitory peptide derived from the consensus motif. Our work reveals an action mechanism of the cryptic MAGI GKs and broadens our understanding of the target recognition rules of phosphoprotein binding modules.

Targeting peptide-mediated interactions in omics

Peptide-mediated interactions (PMIs) play a crucial role in cell signaling network, which are responsible for about half of cellular protein-protein associations in the human interactome and have recently been recognized as a new kind of promising druggable target for drug development and disease therapy. In this article, we give a systematic review regarding the proteome-wide discovery of PMIs and targeting druggable PMIs (dPMIs) with chemical drugs, self-inhibitory peptides (SIPs) and protein agents, particularly focusing on their implications and applications for therapeutic purpose in omics. We also introduce computational peptidology strategies used to model, analyze, and design PMI-targeted molecular entities and further extend the concepts of protein context, direct/indirect readout, and enthalpy/entropy effect involved in PMIs. Current issues and future perspective on this topic are discussed. There is still a long way to go before establishment of efficient therapeutic strategies to target PMIs on the omics scale.

Evaluation of affinity-purification coupled to mass spectrometry approaches for capture of short linear motif-based interactions

Low affinity and transient protein-protein interactions, such as short linear motif (SLiM)-based interactions, require dedicated experimental tools for discovery and validation. Here, we evaluated and compared biotinylated peptide pulldown and protein interaction screen on peptide matrix (PRISMA) coupled to mass-spectrometry (MS) using a set of peptides containing interaction motifs. Eight different peptide sequences that engage in interactions with three distinct protein domains (KEAP1 Kelch, MDM2 SWIB, and TSG101 UEV) with a wide range of affinities were tested. We found that peptide pulldown can be an effective approach for SLiM validation, however, parameters such as protein abundance and competitive interactions can prevent the capture of known interactors. The use of tandem peptide repeats improved the capture and preservation of some interactions. When testing PRISMA, it failed to provide comparable results for model peptides that successfully pulled down known interactors using biotinylated peptide pulldown. Overall, in our hands, we find that albeit more laborious, biotin-peptide pulldown was more successful in terms of validation of known interactions. Our results highlight that the tested affinity-capture MS-based methods for validation of SLiM-based interactions from cell lysates are suboptimal, and we identified parameters for consideration for method development.

Growing ecosystem of deep learning methods for modeling protein-protein interactions

Numerous cellular functions rely on protein-protein interactions. Efforts to comprehensively characterize them remain challenged however by the diversity of molecular recognition mechanisms employed within the proteome. Deep learning has emerged as a promising approach for tackling this problem by exploiting both experimental data and basic biophysical knowledge about protein interactions. Here, we review the growing ecosystem of deep learning methods for modeling protein interactions, highlighting the diversity of these biophysically informed models and their respective trade-offs. We discuss recent successes in using representation learning to capture complex features pertinent to predicting protein interactions and interaction sites, geometric deep learning to reason over protein structures and predict complex structures, and generative modeling to design de novo protein assemblies. We also outline some of the outstanding challenges and promising new directions. Opportunities abound to discover novel interactions, elucidate their physical mechanisms, and engineer binders to modulate their functions using deep learning and, ultimately, unravel how protein interactions orchestrate complex cellular behaviors.

Interactions of the protein tyrosine phosphatase PTPN3 with viral and cellular partners through its PDZ domain: insights into structural determinants and phosphatase activity

The human protein tyrosine phosphatase non-receptor type 3 (PTPN3) is a phosphatase containing a PDZ (PSD-95/Dlg/ZO-1) domain that has been found to play both tumor-suppressive and tumor-promoting roles in various cancers, despite limited knowledge of its cellular partners and signaling functions. Notably, the high-risk genital human papillomavirus (HPV) types 16 and 18 and the hepatitis B virus (HBV) target the PDZ domain of PTPN3 through PDZ-binding motifs (PBMs) in their E6 and HBc proteins respectively. This study focuses on the interactions between the PTPN3 PDZ domain (PTPN3-PDZ) and PBMs of viral and cellular protein partners. We solved the X-ray structures of complexes between PTPN3-PDZ and PBMs of E6 of HPV18 and the tumor necrosis factor-alpha converting enzyme (TACE). We provide new insights into key structural determinants of PBM recognition by PTPN3 by screening the selectivity of PTPN3-PDZ recognition of PBMs, and by comparing the PDZome binding profiles of PTPN3-recognized PBMs and the interactome of PTPN3-PDZ. The PDZ domain of PTPN3 was known to auto-inhibit the protein's phosphatase activity. We discovered that the linker connecting the PDZ and phosphatase domains is involved in this inhibition, and that the binding of PBMs does not impact this catalytic regulation. Overall, the study sheds light on the interactions and structural determinants of PTPN3 with its cellular and viral partners, as well as on the inhibitory role of its PDZ domain on its phosphatase activity.

Native holdup (nHU) to measure binding affinities from cell extracts

Characterizing macromolecular interactions is essential for understanding cellular processes, yet most methods currently used to detect protein interactions from cells are qualitative. Here, we introduce the native holdup (nHU) approach to estimate equilibrium binding constants of protein interactions directly from cell extracts. Compared to other pull-down-based assays, nHU requires less sample preparation and can be coupled to any analytical methods as readouts, such as Western blotting or mass spectrometry. We use nHU to explore interactions of SNX27, a cargo adaptor of the retromer complex and find good agreement between in vitro affinities and those measured directly from cell extracts using nHU. We discuss the strengths and limitations of nHU and provide simple protocols that can be implemented in most laboratories.