Predicting protein domain interactions from coevolution of conserved regions

The knowledge of protein and domain interactions provide crucial insights into their function within a cell. Several computational methods have been proposed to detect interactions between proteins and their constitutive domains. In this work, we focus on approaches based on correlated evolution (coevolution) of sequences of interacting proteins. In this type of approach, often referred to as the mirrortree method, a high correlation of evolutionary histories of two proteins is used as an indicator to predict protein interactions. Recently, it has been observed that subtracting the underlying speciation process by separating coevolution due to common speciation divergence from that due to common function of interacting pairs greatly improves the predictive power of the mirrortree approach. In this article, we investigate possible improvements and limitations of this method. In particular, we demonstrate that the performance of the mirrortree method that can be further improved by restricting the coevolution analysis to the relatively conserved regions in the protein domain sequences (disregarding highly divergent regions). We provide a theoretical validation of our results leading to new insights into the interplay between coevolution and speciation of interacting proteins.     

Predicting Protein-Protein Interaction by the Mirrortree Method: Possibilities and Limitations

Molecular co-evolution analysis as a sequence-only based method has been used to predict protein-protein interactions. In co-evolution analysis, Pearson''s correlation within the mirrortree method is a well-known way of quantifying the correlation between protein pairs. Here we studied the mirrortree method on both known interacting protein pairs and sets of presumed non-interacting protein pairs, to evaluate the utility of this correlation analysis method for predicting protein-protein interactions within eukaryotes. We varied metrics for computing evolutionary distance and evolutionary span of the species analyzed. We found the differences between co-evolutionary correlation scores of the interacting and non-interacting proteins, normalized for evolutionary span, to be significantly predictive for proteins conserved over a wide range of eukaryotic clades (from mammals to fungi). On the other hand, for narrower ranges of evolutionary span, the predictive power was much weaker.     

Domain rearrangements in protein evolution

Most eukaryotic proteins are multi-domain proteins that are created from fusions of genes, deletions and internal repetitions. An investigation of such evolutionary events requires a method to find the domain architecture from which each protein originates. Therefore, we defined a novel measure, domain distance, which is calculated as the number of domains that differ between two domain architectures. Using this measure the evolutionary events that distinguish a protein from its closest ancestor have been studied and it was found that indels are more common than internal repetition and that the exchange of a domain is rare. Indels and repetitions are common at both the N and C-terminals while they are rare between domains. The evolution of the majority of multi-domain proteins can be explained by the stepwise insertions of single domains, with the exception of repeats that sometimes are duplicated several domains in tandem. We show that domain distances agree with sequence similarity and semantic similarity based on gene ontology annotations. In addition, we demonstrate the use of the domain distance measure to build evolutionary trees. Finally, the evolution of multi-domain proteins is exemplified by a closer study of the evolution of two protein families, non-receptor tyrosine kinases and RhoGEFs.     

Co-evolution of proteins with their interaction partners

The divergent evolution of proteins in cellular signaling pathways requires ligands and their receptors to co-evolve, creating new pathways when a new receptor is activated by a new ligand. However, information about the evolution of binding specificity in ligand-receptor systems is difficult to glean from sequences alone. We have used phosphoglycerate kinase (PGK), an enzyme that forms its active site between its two domains, to develop a standard for measuring the co-evolution of interacting proteins. The N-terminal and C-terminal domains of PGK form the active site at their interface and are covalently linked. Therefore, they must have co-evolved to preserve enzyme function. By building two phylogenetic trees from multiple sequence alignments of each of the two domains of PGK, we have calculated a correlation coefficient for the two trees that quantifies the co-evolution of the two domains. The correlation coefficient for the trees of the two domains of PGK is 0. 79, which establishes an upper bound for the co-evolution of a protein domain with its binding partner. The analysis is extended to ligands and their receptors, using the chemokines as a model. We show that the correlation between the chemokine ligand and receptor trees' distances is 0.57. The chemokine family of protein ligands and their G-protein coupled receptors have co-evolved so that each subgroup of chemokine ligands has a matching subgroup of chemokine receptors. The matching subfamilies of ligands and their receptors create a framework within which the ligands of orphan chemokine receptors can be more easily determined. This approach can be applied to a variety of ligand and receptor systems.     

The Role of Insulin C-Peptide in the Coevolution Analyses of the Insulin Signaling Pathway: A Hint for Its Functions

As the linker between the A chain and B chain of proinsulin, C-peptide displays high variability in length and amino acid composition, and has been considered as an inert byproduct of insulin synthesis and processing for many years. Recent studies have suggested that C-peptide can act as a bioactive hormone, exerting various biological effects on the pathophysiology and treatment of diabetes. In this study, we analyzed the coevolution of insulin molecules among vertebrates, aiming at exploring the evolutionary characteristics of insulin molecule, especially the C-peptide. We also calculated the correlations of evolutionary rates between the insulin and the insulin receptor (IR) sequences as well as the domain-domain pairs of the ligand and receptor by the mirrortree method. The results revealed distinctive features of C-peptide in insulin intramolecular coevolution and correlated residue substitutions, which partly supported the idea that C-peptide can act as a bioactive hormone, with significant sequence features, as well as a linker assisting the formation of mature insulin during synthesis. Interestingly, the evolution of C-peptide exerted the highest correlation with that of the insulin receptor and its ligand binding domain (LBD), implying a potential relationship with the insulin signaling pathway.     

LIM proteins: association with the actin cytoskeleton

The LIM domain is an evolutionary conserved double-zinc finger motif found in a variety of proteins exhibiting diverse biological roles. LIM domains have been observed to act as modular protein-binding interfaces mediating protein-protein interactions in the cytoplasm and the nucleus. Interaction of LIM domains with specific protein partners is now known to influence its subcellular localization and activity; however, no single binding motif has been identified as a common target for LIM domains. Several LIM domain-containing proteins associated with the actin cytoskeleton have been identified, playing a role in signal transduction and organization of the actin filaments during various cellular processes.     

Two‐domain analysis of JmjN‐JmjC and PHD‐JmjC lysine demethylases: Detecting an inter‐domain evolutionary stress

Residues at different positions of a multiple sequence alignment sometimes evolve together, due to a correlated structural or functional stress at these positions. Co‐evolution has thus been evidenced computationally in multiple proteins or protein domains. Here, we wish to study whether an evolutionary stress is exerted on a sequence alignment across protein domains, i.e., on longer sequence separations than within a single protein domain. JmjC‐containing lysine demethylases were chosen for analysis, as a follow‐up to previous studies; these proteins are important multidomain epigenetic regulators. In these proteins, the JmjC domain is responsible for the demethylase activity, and surrounding domains interact with histones, DNA or partner proteins. This family of enzymes was analyzed at the sequence level, in order to determine whether the sequence of JmjC‐domains was affected by the presence of a neighboring JmjN domain or PHD finger in the protein. Multiple positions within JmjC sequences were shown to have their residue distributions significantly altered by the presence of the second domain. Structural considerations confirmed the relevance of the analysis for JmjN‐JmjC proteins, while among PHD‐JmjC proteins, the length of the linker region could be correlated to the residues observed at the most affected positions. The correlation of domain architecture with residue types at certain positions, as well as that of overall architecture with protein function, is discussed. The present results thus evidence the existence of an across‐domain evolutionary stress in JmjC‐containing demethylases, and provide further insights into the overall domain architecture of JmjC domain‐containing proteins.     

Conservation among HSP60 sequences in relation to structure, function, and evolution

The chaperonin HSP60 (GroEL) proteins are essential in eubacterial genomes and in eukaryotic organelles. Functional regions inferred from mutation studies and the Escherichia coli GroEL 3D crystal complexes are evaluated in a multiple alignment across 43 diverse HSP60 sequences, centering on ATP/ADP and Mg2+ binding sites, on residues interacting with substrate, on GroES contact positions, on interface regions between monomers and domains, and on residues important in allosteric conformational changes. The most evolutionary conserved residues relate to the ATP/ADP and Mg2+ binding sites. Hydrophobic residues that contribute in substrate binding are also significantly conserved. A large number of charged residues line the central cavity of the GroEL-GroES complex in the substrate-releasing conformation. These span statistically significant intra- and inter-monomer three-dimensional (3D) charge clusters that are highly conserved among sequences and presumably play an important role interacting with the substrate. Unaligned short segments between blocks of alignment are generally exposed at the outside wall of the Anfinsen cage complex. The multiple alignment reveals regions of divergence common to specific evolutionary groups. For example, rickettsial sequences diverge in the ATP/ADP binding domain and gram-positive sequences diverge in the allosteric transition domain. The evolutionary information of the multiple alignment proffers attractive sites for mutational studies.     

Identification of a positively evolving putative binding region with increased variability in posttranslational motifs in zonadhesin MAM domain 2

Positive selection has been shown to be pervasive in sex-related proteins of many metazoan taxa. However, we are only beginning to understand molecular evolutionary processes on the lineage to humans. To elucidate the evolution of proteins involved in human reproduction, we studied the sequence evolution of MAM domains of the sperm-ligand zonadhesin in respect to single amino acid sites, solvent accessibility, and posttranslational modification. GenBank-data were supplemented by new cDNA-sequences of a representative non-human primate panel. Solvent accessibility predictions identified a probably exposed fragment of 30 amino acids belonging to MAM domain 2 (i.e., MAM domain 3 in mouse). The fragment is characterized by significantly increased rate of positively selected amino acid sites and exhibits high variability in predicted posttranslational modification, and, thus, might represent a binding region in the mature protein. At the same time, there is a significant coincidence of positively selected amino acid sites and non-conserved posttranslational motifs. We conclude that the binding specificity of zonadhesin MAM domains, especially of the presumed epitope, is achieved by positive selection at the level of single amino acid sites and posttranslational modifications, respectively.     

Network evolution: rewiring and signatures of conservation in signaling

The analysis of network evolution has been hampered by limited availability of protein interaction data for different organisms. In this study, we investigate evolutionary mechanisms in Src Homology 3 (SH3) domain and kinase interaction networks using high-resolution specificity profiles. We constructed and examined networks for 23 fungal species ranging from Saccharomyces cerevisiae to Schizosaccharomyces pombe. We quantify rates of different rewiring mechanisms and show that interaction change through binding site evolution is faster than through gene gain or loss. We found that SH3 interactions evolve swiftly, at rates similar to those found in phosphoregulation evolution. Importantly, we show that interaction changes are sufficiently rapid to exhibit saturation phenomena at the observed timescales. Finally, focusing on the SH3 interaction network, we observe extensive clustering of binding sites on target proteins by SH3 domains and a strong correlation between the number of domains that bind a target protein (target in-degree) and interaction conservation. The relationship between in-degree and interaction conservation is driven by two different effects, namely the number of clusters that correspond to interaction interfaces and the number of domains that bind to each cluster leads to sequence specific conservation, which in turn results in interaction conservation. In summary, we uncover several network evolution mechanisms likely to generalize across peptide recognition modules.     

Quantification of the elevated rate of domain rearrangements in metazoa

Most eukaryotic proteins consist of multiple domains created through gene fusions or internal duplications. The most frequent change of a domain architecture (DA) is insertion or deletion of a domain at the N or C terminus. Still, the mechanisms underlying the evolution of multidomain proteins are not very well studied.Here, we have studied the evolution of multidomain architectures (MDA), guided by evolutionary information in the form of a phylogenetic tree. Our results show that Pfam domain families and MDAs have been created with comparable rates (0.1-1 per million years (My)). The major changes in DA evolution have occurred in the process of multicellularization and within the metazoan lineage. In contrast, creation of domains seems to have been frequent already in the early evolution. Furthermore, most of the architectures have been created from older domains or architectures, whereas novel domains are mainly found in single-domain proteins. However, a particular group of exon-bordering domains may have contributed to the rapid evolution of novel multidomain proteins in metazoan organisms. Finally, MDAs have evolved predominantly through insertions of domains, whereas domain deletions are less common.In conclusion, the rate of creation of multidomain proteins has accelerated in the metazoan lineage, which may partly be explained by the frequent insertion of exon-bordering domains into new architectures. However, our results indicate that other factors have contributed as well.     

Molecular evolution of ACTIN RELATED PROTEIN 6, a component of SWR1 complex in Arabidopsis

To date, it has been assumed that the evolution of a protein complex is different from that of other proteins. However, there have been few evidences to support this assumption. To understand how protein complexes evolve, we analyzed the evolutionary constraints on ACTIN RELATED PROTEIN 6 (ARP6), a component of the SWR1 complex. Interspecies complementation experiments using transgenic plants that ectopically express transARP6s (ARP6s from other organisms) showed that the function of ARP6s is conserved in plants. In addition, a yeast two-hybrid analysis revealed that this functional conservation depends on its ability to bind with both PIE1 and AtSWC6. ARP6 consists of 4 domains similar to actin. Functional analysis of chimericARP6s (domain-swapped ARP6s between Arabidopsis and mouse) demonstrated that each domain of ARP6s imposes differential evolutionary constraints. Domains 1 and 3 of ARP6 were found to interact with SWC6 and PIE1, respectively, and domain 4 provides a nuclear localization signal. Moreover, domains 1 and 3 showed a slower evolution rate than domain 4, indicating that the interacting domains have higher evolutionary constraints than non-interacting domains do. These findings suggest that the components of this protein complex have evolved coordinately to preserve their interactions.     

Evolution of Chitin-Binding Proteins in Invertebrates

Analysis of a group of invertebrate proteins, including chitinases and peritrophic matrix proteins, reveals the presence of chitin-binding domains that share significant amino acid sequence similarity. The data suggest that these domains evolved from a common ancestor which may be a protein containing a single chitin-binding domain. The duplication and transposition of this chitin-binding domain may have contributed to the functional diversification of chitin-binding proteins. Sequence comparisons indicated that invertebrate and plant chitin binding domains do not share significant amino acid sequence similarity, suggesting that they are not coancestral. However, both the invertebrate and the plant chitin-binding domains are cysteine-rich and have several highly conserved aromatic residues. In plants, cysteines have been elucidated in maintaining protein folding and aromatic amino acids in interacting with saccharides [Wright HT, Sanddrasegaram G, Wright CS (1991) J Mol Evol 33:283–294]. It is likely that these residues perform similar functions in invertebrates. We propose that the invertebrate and the plant chitin-binding domains share similar mechanisms for folding and saccharide binding and that they evolved by convergent evolution. Furthermore, we propose that the disulfide bonds and aromatic residues are hallmarks for saccharide-binding proteins. Received: 2 March 1998 / Accepted: 17 July 1998     

Different evolutionary histories of kringle and protease domains in serine proteases: A typical example of domain evolution

With the aim of elucidating the evolutionary processes of the kringle and protease domains in serine proteases which are involved with the system of blood coagulation and fibrinolysis, we constructed phylogenetic trees for the kringle and protease domains, separately, by use of amino acid sequence data. The phylogenetic trees constructed clearly showed that the topologies were different between the kringle and protease domains. Because both domains are coded by single peptides of serine proteases, this strongly suggests that the kringle and protease domains must have undergone different evolutionary processes. Thus, these observations imply that serine proteases evolve in a way such that each domain is a unit of evolution, exemplifying a typical mode of domain evolution. A possible relationship between the domain evolution and the exon shuffling theory is also discussed from the viewpoint of gene evolution.     

Molecular basis for regulatory subunit diversity in cAMP-dependent protein kinase: crystal structure of the type II beta regulatory subunit

BACKGROUND: Cyclic AMP binding domains possess common structural features yet are diversely coupled to different signaling modules. Each cAMP binding domain receives and transmits a cAMP signal; however, the signaling networks differ even within the same family of regulatory proteins as evidenced by the long-standing biochemical and physiological differences between type I and type II regulatory subunits of cAMP-dependent protein kinase. RESULTS: We report the first type II regulatory subunit crystal structure, which we determined to 2.45 A resolution and refined to an R factor of 0.176 with a free R factor of 0.198. This new structure of the type II beta regulatory subunit of cAMP-dependent protein kinase demonstrates that the relative orientations of the two tandem cAMP binding domains are very different in the type II beta as compared to the type I alpha regulatory subunit. Each structural unit for binding cAMP contains the highly conserved phosphate binding cassette that can be considered the "signature" motif of cAMP binding domains. This motif is coupled to nonconserved regions that link the cAMP signal to diverse structural and functional modules. CONCLUSIONS: Both the diversity and similarity of cAMP binding sites are demonstrated by this new type II regulatory subunit structure. The structure represents an intramolecular paradigm for the cooperative triad that links two cAMP binding sites through a domain interface to the catalytic subunit of cAMP-dependent protein kinase. The domain interface surface is created by the binding of only one cAMP molecule and is enabled by amino acid sequence variability within the peptide chain that tethers the two domains together.     

Exploring metazoan evolution through dynamic and holistic changes in protein families and domains

ABSTRACT: BACKGROUND: Proteins convey the majority of biochemical and cellular activities in organisms. Over the course of evolution, proteins undergo normal sequence mutations as well as large scale mutations involving domain duplication and/or domain shuffling. These events result in the generation of new proteins and protein families. Processes that affect proteome evolution drive species diversity and adaptation. Herein, change over the course of metazoan evolution, as defined by birth/death and duplication/deletion events within protein families and domains, was examined using the proteomes of 9 metazoan and two outgroup species. RESULTS: In studying members of the three major metazoan groups, the vertebrates, arthropods, and nematodes, we found that the number of protein families increased at the majority of lineages over the course of metazoan evolution where the magnitude of these increases was greatest at the lineages leading to mammals. In contrast, the number of protein domains decreased at most lineages and at all terminal lineages. This resulted in a weak correlation between protein family birth and domain birth; however, the correlation between domain birth and domain member duplication was quite strong. These data suggest that domain birth and protein family birth occur via different mechanisms, and that domain shuffling plays a role in the formation of protein families. The ratio of protein family birth to protein domain birth (domain shuffling index) suggests that shuffling had a more demonstrable effect on protein families in nematodes and arthropods than in vertebrates. Through the contrast of high and low domain shuffling indices at the lineages of Trichinella spiralis and Gallus gallus, we propose a link between protein redundancy and evolutionary changes controlled by domain shuffling; however, the speed of adaptation among the different lineages was relatively invariant. Evaluating the functions of protein families that appeared or disappeared at the last common ancestors (LCAs) of the three metazoan clades supports a correlation with organism adaptation. Furthermore, bursts of new protein families and domains in the LCAs of metazoans and vertebrates are consistent with whole genome duplications. CONCLUSION: Metazoan speciation and adaptation were explored by birth/death and duplication/deletion events among protein families and domains. Our results provide insights into protein evolution and its bearing on metazoan evolution.     

Binding constraints on the evolution of enzymes and signalling proteins: the important role of negative pleiotropy

A number of biophysical and population-genetic processes influence amino acid substitution rates. It is commonly recognized that proteins must fold into a native structure with preference over an unfolded state, and must bind to functional interacting partners favourably to function properly. What is less clear is how important folding and binding specificity are to amino acid substitution rates. A hypothesis of the importance of binding specificity in constraining sequence and functional evolution is presented. Examples include an evolutionary simulation of a population of SH2 sequences evolved by threading through the structure and binding to a native ligand, as well as SH3 domain signalling in yeast and selection for specificity in enzymatic reactions. An example in vampire bats where negative pleiotropy appears to have been adaptive is presented. Finally, considerations of compartmentalization and macromolecular crowding on negative pleiotropy are discussed.     

Crystal Structure and Functional Analysis of the Protein Disulfide Isomerase-Related Protein ERp29

The protein disulfide isomerase-related protein ERp29 is a putative chaperone involved in processing and secretion of secretory proteins. Until now, however, both the structure and the exact nature of interacting substrates remained unclear. We provide for the first time a crystal structure of human ERp29, refined to 2.9 Å, and show that the protein has considerable structural homology to its Drosophila homolog Wind. We show that ERp29 binds directly not only to thyroglobulin and thyroglobulin-derived peptides in vitro but also to the Wind client protein Pipe and Pipe-derived peptides, although it fails to process Pipe in vivo. A monomeric mutant of ERp29 and a D domain mutant in which the second peptide binding site is inactivated also bind protein substrates, indicating that the monomeric thioredoxin domain is sufficient for client protein binding. Indeed, the b domains of ERp29 or Wind, expressed alone, are sufficient for binding proteins and peptides. Interacting peptides have in common two or more aromatic residues, with stronger binding for sequences with overall basic character. Thus, the data allow a view of the two putative peptide binding sites of ERp29 and indicate that the apparent, different processing activity of the human and Drosophila proteins in vivo does not stem from differences in peptide binding properties.