A human protein-protein interaction network: a resource for annotating the proteome

Protein-protein interaction maps provide a valuable framework for a better understanding of the functional organization of the proteome. To detect interacting pairs of human proteins systematically, a protein matrix of 4456 baits and 5632 preys was screened by automated yeast two-hybrid (Y2H) interaction mating. We identified 3186 mostly novel interactions among 1705 proteins, resulting in a large, highly connected network. Independent pull-down and co-immunoprecipitation assays validated the overall quality of the Y2H interactions. Using topological and GO criteria, a scoring system was developed to define 911 high-confidence interactions among 401 proteins. Furthermore, the network was searched for interactions linking uncharacterized gene products and human disease proteins to regulatory cellular pathways. Two novel Axin-1 interactions were validated experimentally, characterizing ANP32A and CRMP1 as modulators of Wnt signaling. Systematic human protein interaction screens can lead to a more comprehensive understanding of protein function and cellular processes.     

Peeling the yeast protein network

A set of highly connected proteins (or hubs) plays an important role for the integrity of the protein interaction network of Saccharomyces cerevisae by connecting the network's intrinsic modules. The importance of the hubs' central placement is further confirmed by their propensity to be lethal. However, although highly emphasized, little is known about the topological coherence among the hubs. Applying a core decomposition method which allows us to identify the inherent layer structure of the protein interaction network, we find that the probability of nodes both being essential and evolutionary conserved successively increases toward the innermost cores. While connectivity alone is often not a sufficient criterion to assess a protein's functional, evolutionary and topological relevance, we classify nodes as globally and locally central depending on their appearance in the inner or outer cores. The observation that globally central proteins participate in a substantial number of protein complexes which display an elevated degree of evolutionary conservation allows us to hypothesize that globally central proteins serve as the evolutionary backbone of the proteome. Even though protein interaction data are extensively flawed, we find that our results are very robust against inaccurately determined protein interactions.     

Stratus not altocumulus: a new view of the yeast protein interaction network

Systems biology approaches can reveal intermediary levels of organization between genotype and phenotype that often underlie biological phenomena such as polygenic effects and protein dispensability. An important conceptualization is the module, which is loosely defined as a cohort of proteins that perform a dedicated cellular task. Based on a computational analysis of limited interaction datasets in the budding yeast Saccharomyces cerevisiae, it has been suggested that the global protein interaction network is segregated such that highly connected proteins, called hubs, tend not to link to each other. Moreover, it has been suggested that hubs fall into two distinct classes: "party" hubs are co-expressed and co-localized with their partners, whereas "date" hubs interact with incoherently expressed and diversely localized partners, and thereby cohere disparate parts of the global network. This structure may be compared with altocumulus clouds, i.e., cotton ball-like structures sparsely connected by thin wisps. However, this organization might reflect a small and/or biased sample set of interactions. In a multi-validated high-confidence (HC) interaction network, assembled from all extant S. cerevisiae interaction data, including recently available proteome-wide interaction data and a large set of reliable literature-derived interactions, we find that hub-hub interactions are not suppressed. In fact, the number of interactions a hub has with other hubs is a good predictor of whether a hub protein is essential or not. We find that date hubs are neither required for network tolerance to node deletion, nor do date hubs have distinct biological attributes compared to other hubs. Date and party hubs do not, for example, evolve at different rates. Our analysis suggests that the organization of global protein interaction network is highly interconnected and hence interdependent, more like the continuous dense aggregations of stratus clouds than the segregated configuration of altocumulus clouds. If the network is configured in a stratus format, cross-talk between proteins is potentially a major source of noise. In turn, control of the activity of the most highly connected proteins may be vital. Indeed, we find that a fluctuation in steady-state levels of the most connected proteins is minimized.     

Characterization of protein hubs by inferring interacting motifs from protein interactions

The characterization of protein interactions is essential for understanding biological systems. While genome-scale methods are available for identifying interacting proteins, they do not pinpoint the interacting motifs (e.g., a domain, sequence segments, a binding site, or a set of residues). Here, we develop and apply a method for delineating the interacting motifs of hub proteins (i.e., highly connected proteins). The method relies on the observation that proteins with common interaction partners tend to interact with these partners through a common interacting motif. The sole input for the method are binary protein interactions; neither sequence nor structure information is needed. The approach is evaluated by comparing the inferred interacting motifs with domain families defined for 368 proteins in the Structural Classification of Proteins (SCOP). The positive predictive value of the method for detecting proteins with common SCOP families is 75% at sensitivity of 10%. Most of the inferred interacting motifs were significantly associated with sequence patterns, which could be responsible for the common interactions. We find that yeast hubs with multiple interacting motifs are more likely to be essential than hubs with one or two interacting motifs, thus rationalizing the previously observed correlation between essentiality and the number of interacting partners of a protein. We also find that yeast hubs with multiple interacting motifs evolve slower than the average protein, contrary to the hubs with one or two interacting motifs. The proposed method will help us discover unknown interacting motifs and provide biological insights about protein hubs and their roles in interaction networks.     

Dynamic Hubs Show Competitive and Static Hubs Non-Competitive Regulation of Their Interaction Partners

Date hub proteins have 1 or 2 interaction interfaces but many interaction partners. This raises the question of whether all partner proteins compete for the interaction interface of the hub or if the cell carefully regulates aspects of this process? Here, we have used real-time rendering of protein interaction networks to analyse the interactions of all the 1 or 2 interface hubs of Saccharomyces cerevisiae during the cell cycle. By integrating previously determined structural and gene expression data, and visually hiding the nodes (proteins) and their edges (interactions) during their troughs of expression, we predict when interactions of hubs and their partners are likely to exist. This revealed that 20 out of all 36 one- or two- interface hubs in the yeast interactome fell within two main groups. The first was dynamic hubs with static partners, which can be considered as ‘competitive hubs’. Their interaction partners will compete for the interaction interface of the hub and the success of any interaction will be dictated by the kinetics of interaction (abundance and affinity) and subcellular localisation. The second was static hubs with dynamic partners, which we term ‘non-competitive hubs’. Regulatory mechanisms are finely tuned to lessen the presence and/or effects of competition between the interaction partners of the hub. It is possible that these regulatory processes may also be used by the cell for the regulation of other, non-cell cycle processes.     

A genome-wide Ras-effector interaction network

Here using structural information and protein design tools we have drawn the network of interactions between 20 Ras subfamily proteins with 50 putative Ras binding domains. To validate this network we have cloned six poorly characterized Ras binding domains (RBD) and two Ras proteins (RERG, DiRas1). These, together with previously described RBD domains, Ras and Rap proteins have been analyzed in 70 pull-down experiments. Comparing our interaction network with these and previous pull-down experiments (total of 150 cases) shows a very high accuracy for distinguishing between binders and non-binders ( approximately 0.80). Bioinformatics information was integrated to distinguish those in vitro interactions that are more likely to be relevant in vivo. We proposed several new interactions between Ras family members and effector domains that are of relevance in understanding the physiological role of these proteins. More broadly our results demonstrate that (domain-domain) interaction specificities between members of protein families can be accurately predicted using structural information.     

Mapping signal transduction pathways by phage display

Rapid identification of proteins that interact with a novel gene product is an important element of functional genomics. Here we describe a phage display-based technique for interaction screening of complex cDNA libraries using proteins or synthetic peptides as baits. Starting with the epidermal growth factor receptor (EGFR) cytoplasmic tail, we identified known protein interactions that link EGFR to the Ras/MAP kinase signal transduction cascade and several novel interactions. This approach can be used as a rapid and efficient tool for elucidating protein networks and mapping intracellular signal transduction pathways.     

Allosteric Regulation of the Hsp90 Dynamics and Stability by Client Recruiter Cochaperones: Protein Structure Network Modeling

The fundamental role of the Hsp90 chaperone in supporting functional activity of diverse protein clients is anchored by specific cochaperones. A family of immune sensing client proteins is delivered to the Hsp90 system with the aid of cochaperones Sgt1 and Rar1 that act cooperatively with Hsp90 to form allosterically regulated dynamic complexes. In this work, functional dynamics and protein structure network modeling are combined to dissect molecular mechanisms of Hsp90 regulation by the client recruiter cochaperones. Dynamic signatures of the Hsp90-cochaperone complexes are manifested in differential modulation of the conformational mobility in the Hsp90 lid motif. Consistent with the experiments, we have determined that targeted reorganization of the lid dynamics is a unifying characteristic of the client recruiter cochaperones. Protein network analysis of the essential conformational space of the Hsp90-cochaperone motions has identified structurally stable interaction communities, interfacial hubs and key mediating residues of allosteric communication pathways that act concertedly with the shifts in conformational equilibrium. The results have shown that client recruiter cochaperones can orchestrate global changes in the dynamics and stability of the interaction networks that could enhance the ATPase activity and assist in the client recruitment. The network analysis has recapitulated a broad range of structural and mutagenesis experiments, particularly clarifying the elusive role of Rar1 as a regulator of the Hsp90 interactions and a stability enhancer of the Hsp90-cochaperone complexes. Small-world organization of the interaction networks in the Hsp90 regulatory complexes gives rise to a strong correspondence between highly connected local interfacial hubs, global mediator residues of allosteric interactions and key functional hot spots of the Hsp90 activity. We have found that cochaperone-induced conformational changes in Hsp90 may be determined by specific interaction networks that can inhibit or promote progression of the ATPase cycle and thus control the recruitment of client proteins.     

Quantitative assessment of the structural bias in protein-protein interaction assays

With recent publications of several large-scale protein-protein interaction (PPI) studies, the realization of the full yeast interaction network is getting closer. Here, we have analysed several yeast protein interaction datasets to understand their strengths and weaknesses. In particular, we investigate the effect of experimental biases on some of the protein properties suggested to be enriched in highly connected proteins. Finally, we use support vector machines (SVM) to assess the contribution of these properties to protein interactivity. We find that protein abundance is the most important factor for detecting interactions in tandem affinity purifications (TAP), while it is of less importance for Yeast Two Hybrid (Y2H) screens. Consequently, sequence conservation and/or essentiality of hubs may be related to their high abundance. Further, proteins with disordered structure are over-represented in Y2H screens and in one, but not the other, large-scale TAP assay. Hence, disordered regions may be important both in transient interactions and interactions in complexes. Finally, a few domain families seem to be responsible for a large part of all interactions. Most importantly, we show that there are method-specific biases in PPI experiments. Thus, care should be taken before drawing strong conclusions based on a single dataset.     

Why do hubs in the yeast protein interaction network tend to be essential: reexamining the connection between the network topology and essentiality

The centrality-lethality rule, which notes that high-degree nodes in a protein interaction network tend to correspond to proteins that are essential, suggests that the topological prominence of a protein in a protein interaction network may be a good predictor of its biological importance. Even though the correlation between degree and essentiality was confirmed by many independent studies, the reason for this correlation remains illusive. Several hypotheses about putative connections between essentiality of hubs and the topology of protein-protein interaction networks have been proposed, but as we demonstrate, these explanations are not supported by the properties of protein interaction networks. To identify the main topological determinant of essentiality and to provide a biological explanation for the connection between the network topology and essentiality, we performed a rigorous analysis of six variants of the genomewide protein interaction network for Saccharomyces cerevisiae obtained using different techniques. We demonstrated that the majority of hubs are essential due to their involvement in Essential Complex Biological Modules, a group of densely connected proteins with shared biological function that are enriched in essential proteins. Moreover, we rejected two previously proposed explanations for the centrality-lethality rule, one relating the essentiality of hubs to their role in the overall network connectivity and another relying on the recently published essential protein interactions model.     

A densely interconnected genome-wide network of microRNAs and oncogenic pathways revealed using gene expression signatures

MicroRNAs (miRNAs) are important components of cellular signaling pathways, acting either as pathway regulators or pathway targets. Currently, only a limited number of miRNAs have been functionally linked to specific signaling pathways. Here, we explored if gene expression signatures could be used to represent miRNA activities and integrated with genomic signatures of oncogenic pathway activity to identify connections between miRNAs and oncogenic pathways on a high-throughput, genome-wide scale. Mapping >300 gene expression signatures to >700 primary tumor profiles, we constructed a genome-wide miRNA-pathway network predicting the associations of 276 human miRNAs to 26 oncogenic pathways. The miRNA-pathway network confirmed a host of previously reported miRNA/pathway associations and uncovered several novel associations that were subsequently experimentally validated. Globally, the miRNA-pathway network demonstrates a small-world, but not scale-free, organization characterized by multiple distinct, tightly knit modules each exhibiting a high density of connections. However, unlike genetic or metabolic networks typified by only a few highly connected nodes ("hubs"), most nodes in the miRNA-pathway network are highly connected. Sequence-based computational analysis confirmed that highly-interconnected miRNAs are likely to be regulated by common pathways to target similar sets of downstream genes, suggesting a pervasive and high level of functional redundancy among coexpressed miRNAs. We conclude that gene expression signatures can be used as surrogates of miRNA activity. Our strategy facilitates the task of discovering novel miRNA-pathway connections, since gene expression data for multiple normal and disease conditions are abundantly available.     

Revisiting Date and Party Hubs: Novel Approaches to Role Assignment in Protein Interaction Networks

The idea of “date” and “party” hubs has been influential in the study of protein–protein interaction networks. Date hubs display low co-expression with their partners, whilst party hubs have high co-expression. It was proposed that party hubs are local coordinators whereas date hubs are global connectors. Here, we show that the reported importance of date hubs to network connectivity can in fact be attributed to a tiny subset of them. Crucially, these few, extremely central, hubs do not display particularly low expression correlation, undermining the idea of a link between this quantity and hub function. The date/party distinction was originally motivated by an approximately bimodal distribution of hub co-expression; we show that this feature is not always robust to methodological changes. Additionally, topological properties of hubs do not in general correlate with co-expression. However, we find significant correlations between interaction centrality and the functional similarity of the interacting proteins. We suggest that thinking in terms of a date/party dichotomy for hubs in protein interaction networks is not meaningful, and it might be more useful to conceive of roles for protein-protein interactions rather than for individual proteins.     

Relative quantification of protein-protein interactions using a dual luciferase reporter pull-down assay system

The identification and quantitative analysis of protein-protein interactions are essential to the functional characterization of proteins in the post-proteomics era. The methods currently available are generally time-consuming, technically complicated, insensitive and/or semi-quantitative. The lack of simple, sensitive approaches to precisely quantify protein-protein interactions still prevents our understanding of the functions of many proteins. Here, we develop a novel dual luciferase reporter pull-down assay by combining a biotinylated Firefly luciferase pull-down assay with a dual luciferase reporter assay. The biotinylated Firefly luciferase-tagged protein enables rapid and efficient isolation of a putative Renilla luciferase-tagged binding protein from a relatively small amount of sample. Both of these proteins can be quantitatively detected using the dual luciferase reporter assay system. Protein-protein interactions, including Fos-Jun located in the nucleus; MAVS-TRAF3 in cytoplasm; inducible IRF3 dimerization; viral protein-regulated interactions, such as MAVS-MAVS and MAVS-TRAF3; IRF3 dimerization; and protein interaction domain mapping, are studied using this novel assay system. Herein, we demonstrate that this dual luciferase reporter pull-down assay enables the quantification of the relative amounts of interacting proteins that bind to streptavidin-coupled beads for protein purification. This study provides a simple, rapid, sensitive, and efficient approach to identify and quantify relative protein-protein interactions. Importantly, the dual luciferase reporter pull-down method will facilitate the functional determination of proteins.     

Functional dissection of an intrinsically disordered protein: Understanding the roles of different domains of Knr4 protein in protein�Cprotein interactions

Knr4, recently characterized as an intrinsically disordered Saccharomyces cerevisiae protein, participates in cell wall formation and cell cycle regulation. It is constituted of a functional central globular core flanked by a poorly structured N‐terminal and large natively unfolded C‐terminal domains. Up to now, about 30 different proteins have been reported to physically interact with Knr4. Here, we used an in vivo two‐hybrid system approach and an in vitro surface plasmon resonance (BIAcore) technique to compare the interaction level of different Knr4 deletion variants with given protein partners. We demonstrate the indispensability of the N‐terminal domain of Knr4 for the interactions. On the other hand, presence of the unstructured C‐terminal domain has a negative effect on the interaction strength. In protein interactions networks, the most highly connected proteins or “hubs” are significantly enriched in unstructured regions, and among them the transient hub proteins contain the largest and most highly flexible regions. The results presented here of our analysis of Knr4 protein suggest that these large disordered regions are not always involved in promoting the protein–protein interactions of hub proteins, but in some cases, might rather inhibit them. We propose that this type of regions could prevent unspecific protein interactions, or ensure the correct timing of occurrence of transient interactions, which may be of crucial importance for different signaling and regulation processes.     

Modification of gene duplicability during the evolution of protein interaction network

Duplications of genes encoding highly connected and essential proteins are selected against in several species but not in human, where duplicated genes encode highly connected proteins. To understand when and how gene duplicability changed in evolution, we compare gene and network properties in four species (Escherichia coli, yeast, fly, and human) that are representative of the increase in evolutionary complexity, defined as progressive growth in the number of genes, cells, and cell types. We find that the origin and conservation of a gene significantly correlates with the properties of the encoded protein in the protein-protein interaction network. All four species preserve a core of singleton and central hubs that originated early in evolution, are highly conserved, and accomplish basic biological functions. Another group of hubs appeared in metazoans and duplicated in vertebrates, mostly through vertebrate-specific whole genome duplication. Such recent and duplicated hubs are frequently targets of microRNAs and show tissue-selective expression, suggesting that these are alternative mechanisms to control their dosage. Our study shows how networks modified during evolution and contributes to explaining the occurrence of somatic genetic diseases, such as cancer, in terms of network perturbations.     

Understanding gene essentiality by finely characterizing hubs in the yeast protein interaction network

The centrality-lethality rule, i.e., high-degree proteins or hubs tend to be more essential than low-degree proteins in the yeast protein interaction network, reveals that a protein’s central position indicates its important function, but whether and why hubs tend to be more essential have been heavily debated. Here, we integrated gene expression and functional module data to classify hubs into four types: non-co-expressed non-co-cluster hubs, non-co-expressed co-cluster hubs, co-expressed non-co-cluster hubs and co-expressed co-cluster hubs. We found that all the four hub types are more essential than non-hubs, but they also show different enrichments in essential proteins. Non-co-expressed non-co-cluster hubs play key role in organizing different modules formed by the other three hub types, but they are less important to the survival of the yeast cell. Among the four hub types, co-expressed co-cluster hubs, which likely correspond to the core components of stable protein complexes, are the most essential. These results demonstrated that our classification of hubs into four types could better improve the understanding of gene essentiality.