Correlation between gene expression profiles and protein-protein interactions within and across genomes

MOTIVATION: Function annotation of an unclassified protein on the basis of its interaction partners is well documented in the literature. Reliable predictions of interactions from other data sources such as gene expression measurements would provide a useful route to function annotation. We investigate the global relationship of protein-protein interactions with gene expression. This relationship is studied in four evolutionarily diverse species, for which substantial information regarding their interactions and expression is available: human, mouse, yeast and Escherichia coli. RESULTS: In E.coli the expression of interacting pairs is highly correlated in comparison to random pairs, while in the other three species, the correlation of expression of interacting pairs is only slightly stronger than that of random pairs. To strengthen the correlation, we developed a protocol to integrate ortholog information into the interaction and expression datasets. In all four genomes, the likelihood of predicting protein interactions from highly correlated expression data is increased using our protocol. In yeast, for example, the likelihood of predicting a true interaction, when the correlation is > 0.9, increases from 1.4 to 9.4. The improvement demonstrates that protein interactions are reflected in gene expression and the correlation between the two is strengthened by evolution information. The results establish that co-expression of interacting protein pairs is more conserved than that of random ones.     

Combining gene expression profiles and protein-protein interaction data to infer gene functions

The ever-increasing flow of gene expression profiles and protein-protein interactions has catalyzed many computational approaches for inference of gene functions. Despite all the efforts, there is still room for improvement, for the information enriched in each biological data source has not been exploited to its fullness. A composite method is proposed for classifying unannotated genes based on expression data and protein-protein interaction (PPI) data, which extracts information from both data sources in novel ways. With the noise nature of expression data taken into consideration, importance is attached to the consensus expression patterns of gene classes instead of the actual expression profiles of individual genes, thus characterizing the composite method with enhanced robustness against microarray data variation. With regard to the PPI network, the traditional clear-cut binary attitude towards inter- and intra-functional interactions is abandoned, whereas a more objective perspective into the PPI network structure is formed through incorporating the varied function-function interaction probabilities into the algorithm. The composite method was implemented in two numerical experiments, where its improvement over single-data-source based methods was observed and the superiority of the novel data handling operations was discussed.     

Regulation of HIV gene expression by RNA-protein interactions

Human immunodeficiency virus (HIV) gene expression is tightly controlled through the interaction of trans-acting regulatory proteins with the many cis-acting elements present in viral DNA and RNA. Two proteins encoded by HIV, referred to as Tat and Rev, are essential positive regulators of gene expression. Recent work shows that these proteins control HIV gene expression through interaction with RNA target elements present within the 5' untranslated leader sequence and envelope gene, respectively. There is evidence that these interactions in themselves are not sufficient to confer regulation without the presence of additional host cell factors.     

Path PPI: an integrated dataset of human pathways and protein-protein interactions

Integration of pathway and protein-protein interaction(PPI) data can provide more information that could lead to new biological insights. PPIs are usually represented by a simple binary model, whereas pathways are represented by more complicated models. We developed a series of rules for transforming protein interactions from pathway to binary model, and the protein interactions from seven pathway databases, including PID, Bio Carta, Reactome, Net Path, INOH, SPIKE and KEGG, were transformed based on these rules. These pathway-derived binary protein interactions were integrated with PPIs from other five PPI databases including HPRD, Int Act, Bio GRID, MINT and DIP, to develop integrated dataset(named Path PPI). More detailed interaction type and modification information on protein interactions can be preserved in Path PPI than other existing datasets. Comparison analysis results indicate that most of the interaction overlaps values(OAB) among these pathway databases were less than 5%, and these databases must be used conjunctively. The Path PPI data was provided at http://proteomeview. hupo.org.cn/Path PPI/Path PPI.html.     

Tissue differences revealed by gene expression profiles of various cell lines

Mechanisms through which tissues are formed and maintained remain unknown but are fundamental aspects in biology. Tissue-specific gene expression is a valuable tool to study such mechanisms. But in many biomedical studies, cell lines, rather than human body tissues, are used to investigate biological mechanisms Whether or not cell lines maintain their tissue-specific characteristics after they are isolated and cultured outside the human body remains to be explored. In this study, we applied a novel computational method to identify core genes that contribute to the differentiation of cell lines from various tissues. Several advanced computational techniques, such as Monte Carlo feature selection method, incremental feature selection method, and support vector machine (SVM) algorithm, were incorporated in the proposed method, which extensively analyzed the gene expression profiles of cell lines from different tissues. As a result, we extracted a group of functional genes that can indicate the differences of cell lines in different tissues and built an optimal SVM classifier for identifying cell lines in different tissues. In addition, a set of rules for classifying cell lines were also reported, which can give a clearer picture of cell lines in different issues although its performance was not better than the optimal SVM classifier. Finally, we compared such genes with the tissue-specific genes identified by the Genotype-tissue Expression project. Results showed that most expression patterns between tissues remained in the derived cell lines despite some uniqueness that some genes show tissue specificity.     

IDQD算法预测人类蛋白质相互作用

蛋白质是一切生命活动的物质基础,研究蛋白质的相互作用有助于理解生物过程的分子机制,阐明疾病的分子机理。本文依据蛋白质序列组分特征,应用基于多样性增量的二次判别分析方法,对人类的1 963对蛋白质相互作用进行了预测。自洽检验的各项预测指标均在79%以上,且交叉检验的总精度也大于60%,表明本算法可以用于蛋白质相互作用预测。     

Mapping protein-protein interactions between MutL and MutH by cross-linking

Strand discrimination in Escherichia coli DNA mismatch repair requires the activation of the endonuclease MutH by MutL. There is evidence that MutH binds to the N-terminal domain of MutL in an ATP-dependent manner; however, the interaction sites and the molecular mechanism of MutH activation have not yet been determined. We used a combination of site-directed mutagenesis and site-specific cross-linking to identify protein interaction sites between the proteins MutH and MutL. Unique cysteine residues were introduced in cysteine-free variants of MutH and MutL. The introduced cysteines were modified with the cross-linking reagent 4-maleimidobenzophenone. Photoactivation resulted in cross-links verified by mass spectrometry of some of the single cysteine variants to their respective Cys-free partner proteins. Moreover, we mapped the site of interaction by cross-linking different combinations of single cysteine MutH and MutL variants with thiol-specific homobifunctional cross-linkers of varying length. These results were used to model the MutH.MutL complex and to explain the ATP dependence of this interaction.     

Arabidopsis gene expression changes during cyst nematode parasitism revealed by statistical analyses of microarray expression profiles

With the availability of microarray technology, the expression profiles of thousands of genes can be monitored simultaneously to help determine the mechanisms of these biological processes. We conducted Affymetrix GeneChip microarray analyses of the Arabidopsis-cyst nematode interaction and employed a statistical procedure to analyze the resultant data, which allowed us to identify significant gene expression changes. Quantitative real-time RT-PCR assays were used to confirm the microarray analyses. The results of the expression profiling revealed 128 genes with altered steady-state mRNA levels following infection by the sugar beet cyst nematode (Heterodera schachtii; BCN), in contrast to only 12 genes that had altered expression following infection by the soybean cyst nematode (H. glycines; SCN). The expression of these 12 genes also changed following infection by BCN, i.e. we did not identify any genes regulated exclusively by SCN. The identification of 116 genes whose expression changes during successful cyst nematode parasitism by BCN suggests a potential involvement of these genes in the infection events starting with successful syncytium induction. Further characterization of these genes will permit the formulation of testable hypotheses to explain successful cyst nematode parasitism.     

A systems biology analysis of protein-protein interactions between yeast superoxide dismutases and DNA repair pathways

Superoxide dismutases (SODs) are widely distributed in eukaryotic and prokaryotic species and are responsible for O(2)(.-) scavenging and dismutation to H(2)O(2) and O(2). Mutations in the cytoplasmic (Sod1p) or mitochondrial (Sod2p) form of SODs result in aging, neurodegenerative diseases, and carcinogenesis. Diminished activity of SODs leads to reduced activity of DNA repair pathways, and overexpression of SODs in cells defective for DNA repair increases their level of chromatin damage. Unfortunately, little is understood regarding the interplay between SODs and DNA repair proteins and their role in protecting the genome from oxidative damage. To elucidate the association between yeast SODs and DNA repair mechanisms, a systems biology study was performed employing algorithms of literature data mining and the construction of physical protein-protein interactions from large yeast protein databases. The results obtained in this work allow us to draw two models suggesting that yeast SODs act as O(2)(.-) sensors under conditions of redox imbalance, activating and controlling specific DNA repair mechanisms (e.g., recombinational and excision repair pathways), chromatin remodeling, and synthesis of dNTPs.     

Similarity between protein-protein and protein-carbohydrate interactions, revealed by two crystal structures of lectins from the roots of pokeweed

The roots of pokeweed (Phytolacca americana) are known to contain the lectins designated PL-A, PL-B, PL-C, PL-D1, and PL-D2. Of these lectins, the crystal structures of two PLs, the ligand-free PL-C and the complex of PL-D2 with tri-N-acetylchitotriose, have been determined at 1.8A resolution. The polypeptide chains of PL-C and PL-D2 form three and two repetitive chitin-binding domains, respectively. In the crystal structure of the PL-D2 complex, one trisaccharide molecule is shared mainly between two neighboring molecules related to each other by a crystallographic 2(1)-screw axis, and infinite helical chains of complexed molecules are generated by the sharing of ligand molecules. The crystal structure of PL-C reveals that the molecule is a dimer of two identical subunits, whose polypeptide chains are located in a head-to-tail fashion by a molecular 2-fold axis. Three putative carbohydrate-binding sites in each subunit are located in the dimer interface. The dimerization of PL-C is performed through the hydrophobic interactions between the carbohydrate-binding sites of the opposite domains in the dimer, leading to a distinct dimerization mode from that of wheat-germ agglutinin. Three aromatic residues in each carbohydrate-binding site of PL-C are involved in the dimerization. These residues correspond to the residues that interact mainly with the trisaccharide in the PL-D2 complex and appear to mimic the saccharide residues in the complex. Consequently, the present structure of the PL-C dimer has no room for accommodating carbohydrate. The quaternary structure of PL-C formed through these putative carbohydrate-binding residues may lead to the lack of hemagglutinating activity.     

Host-pathogen interactions revealed by human genome-wide surveys

Genome-wide association studies (GWAS) have now convincingly shown that the diverse outcomes (such as the resolution of infection, clinical deterioration to severe disease, or progression from acute infection to persistent infection) that occur following microbial infection can be at least partly explained by human genetic variation. Unbiased whole-genome approaches have revealed unprecedentedly robust associations between genetic markers and susceptibility to disease, providing clear insights into our understanding of infectious disease biology by revealing the crucial host-pathogen interaction sites. Further work characterizing both the host causative variations and pathogenic microbial strains with distinct host interactions and disease outcomes is now required to provide potential new intervention strategies.     

Heat shock protein 90 mediates protein-protein interactions between human aminoacyl-tRNA synthetases

Heat shock protein 90 (hsp90) is a molecular chaperone responsible for protein folding and maturation in vivo. Interaction of hsp90 with human glutamyl-prolyl-tRNA synthetase (EPRS) was found by genetic screening, co-immunoprecipitation, and in vitro binding experiments. This interaction was sensitive to the hsp90 inhibitor, geldanamycin, and also ATP, suggesting that the chaperone activity of hsp90 is required for interaction with EPRS. Interaction of EPRS with hsp90 was targeted to the region of three tandem repeats linking the two catalytic domains of EPRS that is also responsible for the interaction with isoleucyl-tRNA synthetase (IRS). Interaction of EPRS and IRS also depended on the activity of hsp90, implying that their association was mediated by hsp90. EPRS and IRS form a macromolecular protein complex with at least six other tRNA synthetases and three cofactors. hsp90 preferentially binds to most of the complex-forming enzymes rather than those that are not found in the complex. In addition, inactivation of hsp90 interfered with the in vivo incorporation of the nascent aminoacyl-tRNA synthetases into the multi-ARS complex. Thus, hsp90 appears to mediate protein-protein interactions of mammalian tRNA synthetases.     

Defining and solving the essential protein-protein interactions in HIV infection

The structure determination of macromolecular complexes is entering a new era. The methods of optical microscopy, electron microscopy, X-ray crystallography, and nuclear magnetic resonance increasingly are being combined in hybrid method approaches to achieve an integrated view of macromolecular complexes that span from cellular context to atomic detail. A particularly important application of these hybrid method approaches is the structural analysis of the Human Immunodeficiency Virus (HIV) proteins with their cellular binding partners. High resolution structure determination of essential HIV - host cell protein complexes and correlative analysis of these complexes in the live cell can serve as critical guides in the design of a broad, new class of therapeutics that function by disrupting such complexes. Here, with the hope of stimulating some discussion, we will briefly review some of the literature in the context of what could be done to further apply structural methods to HIV research. We have chosen to focus our attention on certain aspects of the HIV replication cycle where we think that structural information would contribute substantially to the development of new therapeutic and vaccine targets for HIV.     

Evolutionary rate depends on number of protein-protein interactions independently of gene expression level

Background

Whether or not a protein's number of physical interactions with other proteins plays a role in determining its rate of evolution has been a contentious issue. A recent analysis suggested that the observed correlation between number of interactions and evolutionary rate may be due to experimental biases in high-throughput protein interaction data sets.

Discussion

The number of interactions per protein, as measured by some protein interaction data sets, shows no correlation with evolutionary rate. Other data sets, however, do reveal a relationship. Furthermore, even when experimental biases of these data sets are taken into account, a real correlation between number of interactions and evolutionary rate appears to exist.

Summary

A strong and significant correlation between a protein's number of interactions and evolutionary rate is apparent for interaction data from some studies. The extremely low agreement between different protein interaction data sets indicates that interaction data are still of low coverage and/or quality. These limitations may explain why some data sets reveal no correlation with evolutionary rates.

     

Probabilistic prediction and ranking of human protein-protein interactions

Background  

Although the prediction of protein-protein interactions has been extensively investigated for yeast, few such datasets exist for the far larger proteome in human. Furthermore, it has recently been estimated that the overall average false positive rate of available computational and high-throughput experimental interaction datasets is as high as 90%.