Detecting Protein Complexes in Protein Interaction Networks Modeled as Gene Expression Biclusters

Developing suitable methods for the detection of protein complexes in protein interaction networks continues to be an intriguing area of research. The importance of this objective originates from the fact that protein complexes are key players in most cellular processes. The more complexes we identify, the better we can understand normal as well as abnormal molecular events. Up till now, various computational methods were designed for this purpose. However, despite their notable performance, questions arise regarding potential ways to improve them, in addition to ameliorative guidelines to introduce novel approaches. A close interpretation leads to the assent that the way in which protein interaction networks are initially viewed should be adjusted. These networks are dynamic in reality and it is necessary to consider this fact to enhance the detection of protein complexes. In this paper, we present “DyCluster”, a framework to model the dynamic aspect of protein interaction networks by incorporating gene expression data, through biclustering techniques, prior to applying complex-detection algorithms. The experimental results show that DyCluster leads to higher numbers of correctly-detected complexes with better evaluation scores. The high accuracy achieved by DyCluster in detecting protein complexes is a valid argument in favor of the proposed method. DyCluster is also able to detect biologically meaningful protein groups. The code and datasets used in the study are downloadable from https://github.com/emhanna/DyCluster.     

Genetic Architecture of Physiological Phenotypes: Empirical Evidence for Coadapted Gene Complexes

Physiological phenotypes are the result of the coordinated functionof many genes, some of which may be differentiated between conspecificpopulations. Within any one population, natural selection willfavor evolution of a coadapted set of alleles which optimizesphysiological performance and reproductive success. The existenceof such coadapted gene complexes may be assessed by assayingphenotypes of interpopulation hybrids: inferior performanceof hybrids suggests that the allelic combinations present inthe parental populations are coadapted. This approach has beenused to examine the genetic architecture of physiological traitsin the copepod Tigriopus californicus, a species characterizedby sharp genetic differentiation of populations. Developmentaltime and response to osmotic stress both show pronounced F₂hybrid breakdown, a result consistent with genetic coadaptationwithin populations. To better understand the biochemical andmolecular mechanisms underlying hybrid breakdown, we are investigatinga specific biochemical phenotype, the activity of the enzymecytochrome c oxidase (COX). COX (encoded by multiple nuclearand mitochondrial genes) catalyzes the oxidation of cytochromec (encoded by a nuclear gene). Two approaches are being usedto address the extent of coadaptation (both among nuclear genesand between nuclear and mitochondrial genes) underlying COXfunction: (1) studies of the DNA (and inferred amino acid) sequencesof component genes among populations in search of coordinatepatterns of amino acid substitution across loci, and (2) directstudies of COX function in interpopulation hybrids and backcrosses.These approaches provide evidence for the existence of nuclear/nuclearand/or nuclear/mitochondrial coadaptation within natural populationsof T. californicus.     

Genetic Evidence for Function of the bHLH-PAS Protein Gce/Met As a Juvenile Hormone Receptor

Juvenile hormones (JHs) play a major role in controlling development and reproduction in insects and other arthropods. Synthetic JH-mimicking compounds such as methoprene are employed as potent insecticides against significant agricultural, household and disease vector pests. However, a receptor mediating effects of JH and its insecticidal mimics has long been the subject of controversy. The bHLH-PAS protein Methoprene-tolerant (Met), along with its Drosophila melanogaster paralog germ cell-expressed (Gce), has emerged as a prime JH receptor candidate, but critical evidence that this protein must bind JH to fulfill its role in normal insect development has been missing. Here, we show that Gce binds a native D. melanogaster JH, its precursor methyl farnesoate, and some synthetic JH mimics. Conditional on this ligand binding, Gce mediates JH-dependent gene expression and the hormone''s vital role during development of the fly. Any one of three different single amino acid mutations in the ligand-binding pocket that prevent binding of JH to the protein block these functions. Only transgenic Gce capable of binding JH can restore sensitivity to JH mimics in D. melanogaster Met-null mutants and rescue viability in flies lacking both Gce and Met that would otherwise die at pupation. Similarly, the absence of Gce and Met can be compensated by expression of wild-type but not mutated transgenic D. melanogaster Met protein. This genetic evidence definitively establishes Gce/Met in a JH receptor role, thus resolving a long-standing question in arthropod biology.     

Aggregation As a Determinant of Protein Fate in Post-Golgi Compartments: Role of the Luminal Domain of Furin in Lysosomal Targeting

The mammalian endopeptidase furin is a type 1 integral membrane protein that is predominantly localized to the TGN and is degraded in lysosomes with a t_1/2 = 2–4 h. Whereas the localization of furin to the TGN is largely mediated by sorting signals in the cytosolic tail of the protein, we show here that targeting of furin to lysosomes is a function of the luminal domain of the protein. Inhibition of lysosomal degradation results in the accumulation of high molecular weight aggregates of furin; aggregation is also dependent on the luminal domain of furin. Temperature and pharmacologic manipulations suggest that furin aggregation occurs in the TGN and thus precedes delivery to lysosomes. These findings are consistent with a model in which furin becomes progressively aggregated in the TGN, an event that leads to its transport to lysosomes. Our observations indicate that changes in the aggregation state of luminal domains can be potent determinants of biosynthetic targeting to lysosomes and suggest the possible existence of quality control mechanisms for disposal of aggregated proteins in compartments of the secretory pathway other than the endoplasmic reticulum.     

Covariation of Mitochondrial Genome Size with Gene Lengths: Evidence for Gene Length Reduction During Mitochondrial Evolution

Reduction of genome size and gene shortening have been observed in a number of parasitic and mutualistic intracellular symbionts. Reduction of coding capacity is also a unifying principle in the evolutionary history of mitochondria, but little is known about the evolution of gene length in mitochondria. The genes for cytochrome c oxidase subunits I–III, cytochrome b, and the large and small subunit rRNAs are, with very few exceptions, always found on the mitochondrial genome. These resident mitochondrial genes can therefore be used to test whether the reduction in gene lengths observed in a number of intracellular symbionts is also seen in mitochondria. Here we show that resident mitochondrial gene products are shorter than their corresponding counterparts in -proteobacteria and, furthermore, that the reduction of mitochondrial genome size is correlated with a reduction in the length of the corresponding resident gene products. We show that relative genomic AT content, which has been identified as a factor influencing gene lengths in other systems, cannot explain gene length/genome size covariance observed in mitochondria. Our data are therefore in agreement with the idea that gene length evolves as a consequence of selection for smaller genomes, either to avoid accumulation of deleterious mutations or triggered by selection for a replication advantage.     

Green Fluorescent Protein As a Cell-Labeling Tool and a Reporter of Gene Expression in Transgenic Rainbow Trout

Green fluorescent protein (GFP) has been used as an indicator of transgene expression in living cells and organisms. For testing the utility of GFP in rainbow trout, we microinjected fertilized eggs with four types of supercoiled constructs containing two variants of GFP complementary DNA (S65T and EGFP), driven by two ubiquitous regulatory elements, human cytomegalovirus immediate early enhancer-promoter (CMV) and Xenopus laevis elongation factor 1α enhancer-promoter (EF1). Green fluorescence was first observed at 3 days postfertilization, when the embryo was in the mid-blastula stage. Fluorescence could be detected mosaically in various types of embryonic cells and tissues of swim-up fry. Both the percentage of fluorescent cells and the fluorescence intensity of GFP-expressing cells on blastoderms, measured with a microscopic photometry system, were highest in CMV-EGFP-microinjected embryos. We conclude that GFP is capable of producing detectable fluorescence in rainbow trout, and can be a powerful tool as a cell marker and reporter gene for cold-water fish, and that analysis of GFP expression in living cells is useful for characterizing the activity of cis-elements in vivo. Received December 21, 1998; accepted March 31, 1999.     

Rank of Correlation Coefficient as a Comparable Measure for Biological Significance of Gene Coexpression

Information regarding gene coexpression is useful to predict gene function. Several databases have been constructed for gene coexpression in model organisms based on a large amount of publicly available gene expression data measured by GeneChip platforms. In these databases, Pearson''s correlation coefficients (PCCs) of gene expression patterns are widely used as a measure of gene coexpression. Although the coexpression measure or GeneChip summarization method affects the performance of the gene coexpression database, previous studies for these calculation procedures were tested with only a small number of samples and a particular species. To evaluate the effectiveness of coexpression measures, assessments with large-scale microarray data are required. We first examined characteristics of PCC and found that the optimal PCC threshold to retrieve functionally related genes was affected by the method of gene expression database construction and the target gene function. In addition, we found that this problem could be overcome when we used correlation ranks instead of correlation values. This observation was evaluated by large-scale gene expression data for four species: Arabidopsis, human, mouse and rat.     

Protein Complexes in Bacteria

Large-scale analyses of protein complexes have recently become available for Escherichia coli and Mycoplasma pneumoniae, yielding 443 and 116 heteromultimeric soluble protein complexes, respectively. We have coupled the results of these mass spectrometry-characterized protein complexes with the 285 “gold standard” protein complexes identified by EcoCyc. A comparison with databases of gene orthology, conservation, and essentiality identified proteins conserved or lost in complexes of other species. For instance, of 285 “gold standard” protein complexes in E. coli, less than 10% are fully conserved among a set of 7 distantly-related bacterial “model” species. Complex conservation follows one of three models: well-conserved complexes, complexes with a conserved core, and complexes with partial conservation but no conserved core. Expanding the comparison to 894 distinct bacterial genomes illustrates fractional conservation and the limits of co-conservation among components of protein complexes: just 14 out of 285 model protein complexes are perfectly conserved across 95% of the genomes used, yet we predict more than 180 may be partially conserved across at least half of the genomes. No clear relationship between gene essentiality and protein complex conservation is observed, as even poorly conserved complexes contain a significant number of essential proteins. Finally, we identify 183 complexes containing well-conserved components and uncharacterized proteins which will be interesting targets for future experimental studies.     

Subunit Exchange in Protein Complexes

Over the past 50?years, protein complexes have been studied with techniques such as X-ray crystallography and electron microscopy, generating images which although detailed are static and homogeneous. More recently, limited application of in vivo fluorescence and other techniques has revealed that many complexes previously thought stable and compositionally uniform are dynamically variable, continually exchanging components with a freely circulating pool of “spares.” Here, we consider the purpose and prevalence of protein exchange, first reviewing the ongoing story of exchange in the bacterial flagella motor, before surveying reports of exchange in complexes across all domains of life, together highlighting great diversity in timescales and functions. Finally, we put this in the context of high-throughput proteomic studies which hint that exchange might be the norm, rather than an exception.     

Identification of a Novel Determinant for Membrane Association in Hepatitis C Virus Nonstructural Protein 4B

Nonstructural protein 4B (NS4B) plays an essential role in the formation of the hepatitis C virus (HCV) replication complex. It is a relatively poorly characterized integral membrane protein predicted to comprise four transmembrane segments in its central portion. Here, we describe a novel determinant for membrane association represented by amino acids (aa) 40 to 69 in the N-terminal portion of NS4B. This segment was sufficient to target and tightly anchor the green fluorescent protein to cellular membranes, as assessed by fluorescence microscopy as well as membrane extraction and flotation analyses. Circular dichroism and nuclear magnetic resonance structural analyses showed that this segment comprises an amphipathic α-helix extending from aa 42 to 66. Attenuated total reflection infrared spectroscopy and glycosylation acceptor site tagging revealed that this amphipathic α-helix has the potential to traverse the phospholipid bilayer as a transmembrane segment, likely upon oligomerization. Alanine substitution of the fully conserved aromatic residues on the hydrophobic helix side abrogated membrane association of the segment comprising aa 40 to 69 and disrupted the formation of a functional replication complex. These results provide the first atomic resolution structure of an essential membrane-associated determinant of HCV NS4B.With 120 to 180 million chronically infected individuals worldwide, hepatitis C virus (HCV) infection represents a major cause of chronic hepatitis, liver cirrhosis, and hepatocellular carcinoma (38). HCV contains a 9.6-kb positive-strand RNA genome that encodes a polyprotein of about 3,000 amino acids (reviewed in references 36 and 51). The polyprotein precursor is co- and posttranslationally processed by cellular and viral proteases to yield the mature structural and nonstructural proteins. The structural proteins include the core and the envelope glycoproteins E1 and E2. The nonstructural proteins include the p7 ion channel polypeptide, the NS2-3 and NS3-4A proteases, an RNA helicase located in the C-terminal two-thirds of NS3, the NS4B and NS5A proteins, and the NS5B RNA-dependent RNA polymerase. HCV replication takes place in a membrane-associated complex, composed of viral proteins, replicating RNA, altered cellular membranes, and other host factors (7, 18, 31, 43). Determinants for membrane association of the HCV nonstructural proteins have been mapped and a likely endoplasmic reticulum (ER)-derived membrane alteration, designated the membranous web, was found to harbor the HCV replication complex (7, 18; reviewed in reference 36).NS4B is a 27-kDa integral ER membrane protein (21). The expression of NS4B alone induces the formation of the membranous web (7). Thus, an essential function of NS4B is the induction of the specific membrane alteration that serves as a scaffold for the HCV replication complex. In addition, a nucleotide-binding motif has been proposed to reside in the middle of NS4B (8), and RNA binding properties have recently been reported for NS4B (9).Both the N and the C termini of NS4B are believed to be oriented toward the cytosol, and prediction algorithms indicate the presence of four putative transmembrane segments in the central portion of the protein (21, 27, 28, 42). The cytosolic orientation of the bulk of the protein was confirmed experimentally (21), but a more refined membrane topology is so far elusive. The introduction of glycosylation acceptor sites at various positions in NS4B validated the prediction of ER luminal loops around amino acid positions 112 and 161 (27, 28). Intriguingly, the N terminus of NS4B was reported to be translocated into the ER lumen at least partially, presumably by a posttranslational mechanism (28). Interestingly, the coexpression of the other HCV proteins appears to limit this translocation (27). A recent report indicates that NS4B is palmitoylated at C-terminal residues Cys 257 and Cys 261 and forms oligomers (59).In order to define the membrane topology of NS4B we have analyzed a comprehensive panel of green fluorescent protein (GFP) fusion constructs comprising different segments of NS4B (N. Arora, V. Castet, and D. Moradpour, unpublished data). In the course of these studies, we unexpectedly found that a fusion construct comprising the N-terminal 74 amino acids (aa) of NS4B was associated with membranes, while most prediction methods located the beginning of the first transmembrane segment around aa 74 (21, 28, 42). Here, we demonstrate that an amphipathic α-helix extending from aa 42 to 66 (α-helix 42-66) in the N-terminal portion of NS4B mediates this membrane association and plays an essential role in the formation of the HCV replication complex.     

Analysis of Gene Coexpression by B-Spline Based CoD Estimation

The gene coexpression study has emerged as a novel holistic approach for microarray data analysis. Different indices have been used in exploring coexpression relationship, but each is associated with certain pitfalls. The Pearson's correlation coefficient, for example, is not capable of uncovering nonlinear pattern and directionality of coexpression. Mutual information can detect nonlinearity but fails to show directionality. The coefficient of determination (CoD) is unique in exploring different patterns of gene coexpression, but so far only applied to discrete data and the conversion of continuous microarray data to the discrete format could lead to information loss. Here, we proposed an effective algorithm, CoexPro, for gene coexpression analysis. The new algorithm is based on B-spline approximation of coexpression between a pair of genes, followed by CoD estimation. The algorithm was justified by simulation studies and by functional semantic similarity analysis. The proposed algorithm is capable of uncovering both linear and a specific class of nonlinear relationships from continuous microarray data. It can also provide suggestions for possible directionality of coexpression to the researchers. The new algorithm presents a novel model for gene coexpression and will be a valuable tool for a variety of gene expression and network studies. The application of the algorithm was demonstrated by an analysis on ligand-receptor coexpression in cancerous and noncancerous cells. The software implementing the algorithm is available upon request to the authors.     

Further Evidence that Cytoplasmic Acidosis Is a Determinant of Flooding Intolerance in Plants

We present two pieces of evidence that regulation of cytoplasmic pH near neutrality is a prerequisite for survival of root tips during hypoxia. First, blackeye peas and navy beans show earlier cytoplasmic acidosis under hypoxia than soybeans or pumpkin or maize, and die earlier. Second, when cytoplasmic acidosis in maize root tips is greatly retarded by treatment with 25 millimolar Ca(NO₃)₂, they remain viable under hypoxia for a much longer period of time than untreated hypoxic root tips. We also show that viability of maize root tips is unaffected by the supply of exogenous sugar (and so on the rate of ethanolic fermentation) for at least 16 hours of hypoxia.     

Isolation of a Gene for a Metallothionein-Like Protein from Soybean

Using a synthetic oligonucleotide that corresponded to the consensusnucleotide sequence of the N-terminal region of mammalian metallothioneinas probe, we isolated a cDNA clone from a soybean library. Theclone had an ORF that encode a protein of 79 amino acids whichshowed significant homology to both N- and C-terminal regionsof mammalian and Neurospora crassa metallothioneins ⁴Present address: Department of Biosciences, Teikyo University,Toyosatodai, Utsunomiya, Tochigi, 320 Japan (Received March 13, 1991; Accepted June 17, 1991)     

Molecular Analysis of the O-Antigen Gene Cluster of Escherichia coli O86:B7 and Characterization of the Chain Length Determinant Gene (wzz)

Escherichia coli O86:B7 has long been used as a model bacterial strain to study the generation of natural blood group antibody in humans, and it has been shown to possess high human blood B activity. The O-antigen structure of O86:B7 was solved recently in our laboratory. Comparison with the published structure of O86:H2 showed that both O86 subtypes shared the same O unit, yet each of the O antigens is polymerized from a different terminal sugar in a different glycosidic linkage. To determine the genetic basis for the O-antigen differences between the two O86 strains, we report the complete sequence of O86:B7 O-antigen gene cluster between galF and hisI, each gene was identified based on homology to other genes in the GenBank databases. Comparison of the two O86 O-antigen gene clusters revealed that the encoding regions between galF and gnd are identical, including wzy genes. However, deletion of the two wzy genes revealed that wzy in O86:B7 is responsible for the polymerization of the O antigen, while the deletion of wzy in O86:H2 has no effect on O-antigen biosynthesis. Therefore, we proposed that there must be another functional wzy gene outside the O86:H2 O-antigen gene cluster. Wzz proteins determine the degree of polymerization of the O antigen. When separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, the lipopolysaccharide (LPS) of O86:B7 exhibited a modal distribution of LPS bands with relatively short O units attached to lipid A-core, which differs from the LPS pattern of O86:H2. We proved that the wzz genes are responsible for the different LPS patterns found in the two O86 subtypes, and we also showed that the very short type of LPS is responsible for the serum sensitivity of the O86:B7 strain.     

Drug Repositioning through Systematic Mining of Gene Coexpression Networks in Cancer

Gene coexpression network analysis is a powerful “data-driven” approach essential for understanding cancer biology and mechanisms of tumor development. Yet, despite the completion of thousands of studies on cancer gene expression, there have been few attempts to normalize and integrate co-expression data from scattered sources in a concise “meta-analysis” framework. We generated such a resource by exploring gene coexpression networks in 82 microarray datasets from 9 major human cancer types. The analysis was conducted using an elaborate weighted gene coexpression network (WGCNA) methodology and identified over 3,000 robust gene coexpression modules. The modules covered a range of known tumor features, such as proliferation, extracellular matrix remodeling, hypoxia, inflammation, angiogenesis, tumor differentiation programs, specific signaling pathways, genomic alterations, and biomarkers of individual tumor subtypes. To prioritize genes with respect to those tumor features, we ranked genes within each module by connectivity, leading to identification of module-specific functionally prominent hub genes. To showcase the utility of this network information, we positioned known cancer drug targets within the coexpression networks and predicted that Anakinra, an anti-rheumatoid therapeutic agent, may be promising for development in colorectal cancer. We offer a comprehensive, normalized and well documented collection of >3000 gene coexpression modules in a variety of cancers as a rich data resource to facilitate further progress in cancer research.     

Evidence for the Presence of a Ganglioside Transfer Protein in Brain

An extract from rat brain has been shown to catalyze the transfer of ganglioside GM1 from sonicated vesicles to erythrocyte ghosts. It also enhanced the transfer of GM1 to a crude neuronal membrane preparation, whereas myelin took up only a very limited amount. The transfer activity was heat-labile. Similar transfer activities were found in extracts from bovine gray and white matter, that of the former being comparable to rat brain whereas the latter was greater per milligram protein.     

Evidence for a Structural Role of Protein in Algal Cell Walls

Load-extension measurements were made on three filamentous algaebefore and after either digestion with proteolytic enzymes ortreatment with dithiothreitol. Large differences in tensileproperties of the walls were observed, particularly after pronasedigestion, in two algae, Cladophora and Chaetomorpha, whichcontain hydroxyproline in the wall. Pronase had little or noeffect on a third alga, Nitella, lacking hydroxyproline. A smallerdifference was found on treatment with dithiothreitol, a specificreducing agent for disulphide bonds. These results suggest thata hydroxyproline containing protein is a structural componentof these algal walls, and that hydroxyproline itself is involvedin the carbohydratepeptide linkage.