Conserved host–pathogen PPIs

Integrative Biology PAPER Cite this: DOI: 10.1039/c2ib20206a Conserved host–pathogen PPIs† Globally conserved inter-species bacterial PPIs based conserved host-pathogen interactome derived novel target in C. pseudotuberculosis, C. diphtheriae, M. tuberculosis, C. ulcerans, Y. pestis, and E. coli targeted by Piper betel compounds Debmalya Barh,z*ab Krishnakant Gupta,ac Neha Jain,a Gourav Khatri,ac Nidia León-Sicairos,d Adrian Canizalez-Roman,d Sandeep Tiwari,a Ankit Verma,ac Sachin Rahangdale,ac Syed Shah Hassan,e Anderson Rodrigues dos Santos,e Amjad Ali,e Luis Carlos Guimarães,e Rommel Thiago Jucá Ramos,f Pratap Devarapalli,g Neha Barve,ac Marriam Bakhtiar,e Ranjith Kumavath,g Preetam Ghosh,ah Anderson Miyoshi,e Artur Silva,f Anil Kumar,c Amarendra Narayan Misra,bi Kenneth Blum,ajkl Jan Baumbachm and Vasco Azevedoze Received 27th August 2012, Accepted 5th November 2012 DOI: 10.1039/c2ib20206a www.rsc.org/ibiology Although attempts have been made to unveil protein–protein and host–pathogen interactions based on molecular insights of important biological events and pathogenesis in various organisms, these efforts have not yet been reported in Corynebacterium pseudotuberculosis (Cp), the causative agent of Caseous Lymphadenitis (CLA). In this study, we used computational approaches to develop common conserved intra-species protein–protein interaction (PPI) networks first time for four Cp strains (Cp FRC41, Cp 316, Cp 3/99-5, and Cp P54B96) followed by development of a common conserved inter-species bacterial PPI using conserved proteins in multiple pathogens (Y. pestis, M. tuberculosis, C. diphtheriae, C. ulcerans, E. coli, and all four Cp strains) and E. Coli based experimentally validated PPI data. Furthermore, the interacting proteins in the common conserved inter-species bacterial PPI were used to generate a conserved host–pathogen interaction (HP-PPI) network considering human, goat, sheep, bovine, and horse as hosts. The HP-PPI network was validated, and acetate kinase (Ack) was identified as a novel broad spectrum target. Ceftiofur, penicillin, and two natural compounds derived from Piper betel were predicted to inhibit Ack activity. One of these Piper betel compounds found to inhibit E. coli O157:H7 growth similar to penicillin. The target specificity of these betel compounds, their effects on other studied pathogens, and other in silico results are currently being validated and the results are promising. a Centre for Genomics and Applied Gene Technology, Institute of Integrative Omics and Applied Biotechnology (IIOAB), Nonakuri, Purba Medinipur, West Bengal-721172, India. E-mail: dr.barh@gmail.com; Fax: +91-944 955 0032; Tel: +91-944 955 0032 b Department of Biosciences and Biotechnology, School of Biotechnology, Fakir Mohan University, Jnan Bigyan Vihar, Balasore, Orissa, India c School of Biotechnology, Devi Ahilya University, Khandwa Road Campus, Indore, MP, India d Unidad de investigacion, Facultad de Medicina, Universidad Autónoma de Sinaloa. Cedros y Sauces, Fraccionamiento Fresnos, Culiacán Sinaloa 80246, México e Instituto de Ciências Biológicas, Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brazil f Instituto de Ciências Biológicas, Universidade Federal do Pará, Belém, PA, Brazil g Department of Genomic Science, School of Biological Sciences, Riverside Transit Campus, Central University of Kerala, Kasaragod, India h Department of Computer Science and Center for the Study of Biological Complexity, Virginia Commonwealth University, 401 West Main Street, Room E4234, P.O. Box 843019, Richmond, Virginia 23284-3019, USA i Center for Life Sciences, School of Natural Sciences, Central University of Jharkhand, Ranchi, Jharkhand State, India j University of Florida, College of Medicine, Gainesville, Florida, USA k Global Integrated Services Unit University of Vermont Center for Clinical & Translational Science, College of Medicine, Burlington, VT, USA l Dominion Diagnostics LLC, North Kingstown, Rhode Island, USA m Computational Biology Group, Department of Mathematics and Computer Science, University of Southern Denmark, Campusvej 55, DK-5230 Odense, Denmark † Electronic supplementary information (ESI) available: Supplementary Tables 1–8. See DOI: 10.1039/c2ib20206a ‡ DB and VA conceived the idea; DB designed the study, collected and analyzed primary data to finalize the protocol, coordinated and leaded the entire project, and wrote the manuscript. DB, KG, NJ, GK, ST, AV, and SR performed all in silico analyses; SSH, ARS, AA, LCG, and ATJR performed Cp genome annotation and cross checked all other analyses; PD, RK, MB, NB, and PG cross checked all analyses; NLS and ACR conducted microbial experiments with betel compounds; AK, KB, ANM, AM, PG, JB, and VA provided technical consultations and reviewed the manuscript. All authors have read and approved the final manuscript. This journal is c The Royal Society of Chemistry 2013 Integr. Biol. Paper Integrative Biology Insight, innovation, integration Here, for the first time we represent the intra-species PPIs in C. pseudotuberculosis (Cp). Further, a novel method was used to develop common conserved interspecies bacterial PPIs for C. pseudotuberculosis, Y. pestis, M. tuberculosis, C. diphtheriae, C. ulcerans, C. glutamicum, and E. coli (pathogenic, nonpathogenic, closed and distant taxa) to identify the conserved common essential PPIs in these bacteria. This inter-species bacterial PPI was then used to make conserved common host–pathogen interactions. Using network analysis strategies and subtraction genomics approaches, from this conserved common host–pathogen interactions; Ack was identified as a key target for all these bacteria. Virtual screening shows Penicillin and Ceftiofur can inhibit Ack. However, Piper betel derived Piperdardine and Dehydropipernonaline are predicted to have similar or superior effects compared to Penicillin and Ceftiofur on Ack. Piperdardine inhibits E. coli O157:H7 growth similar to penicillin and can also work on other pathogens in a similar way. Introduction Protein–protein interactions (PPIs) are crucial events in several biological processes. PPI-based decoding of the functionality of uncharacterized proteins can reveal unknown molecular mechanisms behind important biological events within a cell or at the system level.1,2 Therefore, PPIs of an entire proteome or between a set of proteins in a pathogen and its corresponding host can be useful in identifying precise molecular mechanisms of host–pathogen interactions, thereby leading to the development of effective drug targets against the pathogen.3–5 Initial computational approaches for the prediction of PPIs were based on the structural context of proteins. However, in the post-genomic era, the focus has shifted, and sequence information is now used.6,7 The availability of genomic and proteomic data and the advent of yeast two-hybrid, affinity purification, mass spectrometry, and other high-throughput techniques have tremendously enriched the field. Recently, a number of computational approaches have also been developed to facilitate the prediction and study of these ubiquitous interactions. A number of in silico approaches were recently reviewed that highlight the use of genomic, structural, and biological contexts of proteins and genes in complete genomes for PPI predictions and determination of the functional relationship among them.8 Using these approaches, the development of highly reliable PPIs in several organisms including yeast9 and human10 are close to completion. However, false-positive interactions are a concern.11,12 Similarly, sequence-based computational methods including gene neighborhood,13 phylogenetic profiles,14 gene fusion,15 co-evolution,16 and domain interactions,17 along with several newly developed methods, have been used to generate genome-/proteome-wide interactions in a number of organisms including, M. tuberculosis18 and E. coli.19 Genomic sequences are used as the primary data sources in these prediction techniques, which assume that evolutionary co-inherited gene pairs have a functional association.20,21 Similarly, amino acid (AA) sequencebased PPIs identify interacting protein pairs that have specific AA residues due to their co-evolution or binding to one another.22 Yeats et al. have catalogued the commonly occurring domains for PPIs.23–25 However, in general, a PPI denotes the binding of proteins to other proteins. Concurrently, in silico host–pathogen interactions have been reported in many organisms, including Plasmodium,26,27 M. tuberculosis,28 and Streptococcus.29 Combined computational and yeast two-hybrid based approaches have been recently Integr. Biol. published for B. anthracis, F. tularensis, and Y. pestis PPIs.5 Although, it gives only 20% positive interactions and therefore produces a high degree of false-negative interaction,30 the yeast two-hybrid method and related high-throughput and computational interaction data have been analyzed to identify targets in many pathogens. Although extensive studies have been conducted for host–pathogen interactions and target identification in M. tuberculosis31–34 and Corynebacterium diphtheriae,35–38 another member of the Corynebacterium, Mycobacterium, Nocardia, and Rhodococcus (CMNR) group of pathogens, C. pseudotuberculosis, remains uninvestigated with respect to both its PPI and host–pathogen interactions. C. pseudotuberculosis causes Caseous Lymphadenitis (CLA) or ‘‘cheesy gland’’ in small ruminants worldwide, which can result in a significant economic loss. CLA is characterized by the formation of external or internal abscesses, chronic limb infections (lymphangitis) and lymphadenitis.39,40 It also infects visceral organs such as the liver, spleen, kidneys and lungs.41 Although the bacterium rarely infects humans, there are reports of human lymphadenitis, and clinical strains have been isolated.42 Other important pathogens in the CMNR group, M. tuberculosis and C. diphtheriae, cause tuberculosis and diphtheria, respectively. According to the WHO, approximately 1.7 million people died from tuberculosis in 2009 and 50,000 died from diphtheria in 2004. Yersinia pestis causes plague and poses a threat for use in bioterrorism.43 Most of its isolates are derived from Y. pseudotuberculosis,44 and lymphadenitis or lymphadenopathy caused by Cp is one of the symptoms of a Y. pestis45–47 and M. tuberculosis48,49 infection. Here, for the first time, using a combination of comparative, functional, and phylogenomics approaches, supported by published, experimentally validated data we report (a) a probable conserved PPIs in the Cp proteome. (b) Further, we created proteome-wide common conserved PPIs for a number of pathogenic and non-pathogenic bacteria (C. pseudotuberculosis, C. diphtheriae, C. ulcerans, M. tuberculosis, Y. pestis, and E. coli). (c) Thereafter, the proteins involved in this common conserved intra-species bacterial PPIs were used to generate host–pathogen interactions considering human, goat, sheep, and horse as hosts. This host–pathogen PPI was based on experimentally validated published host–pathogen interactions data. (d) By analyzing the host–pathogen interaction networks, we identified common conserved targets in these pathogens. This journal is c The Royal Society of Chemistry 2013 Integrative Biology (e) Finally, we use the identified targets to develop broad spectrum drugs from an existing antibiotic regime and phytochemicals derived from Piper betel. Materials and methods Selection of highly identical conserved proteins in Cp, other bacteria, and hosts Selection of conserved genes for intra-species Cp PPI. In this work, we aimed to develop PPIs based on sequences. Therefore, highly identical common conserved proteins of Cp were selected using comparative genomics/proteomics approaches using the BLAST tool.50 As there is no report on Cp PPIs so far, first we approached to develop intra-species common conserved PPIs for four Cp strains (strain FRC41, 316, 3/99-5, and P54B96) that were isolated from four different hosts and recently sequenced. The strain FRC41 (biovar ovis) was isolated from a human; strain 316 (biovar equi) was isolated from a horse; strain 3/99-5 (biovar ovis) was isolated from a sheep; and strain P54B96 (biovar ovis) was isolated from an antelope. Highly conserved and common proteins of these four strains were selected using BLASTp cut off values: E = 0.0001 and Z 80% identity. Such BLAST parameters were used to select identical sequences from different strains of a species.51 Selection of conserved genes for inter-species bacterial PPI. Next, the highly identical common conserved genes across a wide range of pathogenic and non-pathogenic bacteria from the same and distant taxa (E. coli, Y. pestis, M. tuberculosis, C. diphtheriae, C. ulcerans, C. glutamicum, and all four Cp strains) were selected using the BLAST option available in the Prokaryotic Sequence homology Analysis (PSAT) Tool.52 The PSAT tool was selected because it compares gene neighborhoods, gene clusters, homologs, and orthologs among multiple bacterial genomes in a single run. It also accounts information of gene context including weak alignment scores therefore provides better sensitivity compared to other available comparative analysis methods. To get the homolog list we used Y. pestis genome as reference and compared with M. tuberculosis, C. diphtheriae, C. glutamicum, and E. coli. The BLAST score thresholds were set to: E = 0.01, bit score Z 100, identity Z 35% that was used in our previous report to identify homologs essential genes.51 The common homolog genes in these bacteria were selected and further tested for their presence in C. ulcerans and pool of conserved common genes of four Cp strains, and other selected bacterial strains (Table S1, ESI†) using NCBI BLASTp with same parameters. Finally, the common conserved genes that are present in all these selected bacteria were collected and the common conserved E. coli K12 genes were used in further analysis as most of the required experimentally validated data are available for this species. The list of bacteria used in this analysis is represented in Table S1 (ESI†). Selection of conserved genes in hosts. A range of hosts (human, goat, sheep, bovine, and horse) were selected based on the commonality of the pathogenesis from the selected pathogenic organisms. The conserved genes in these hosts were identified using the general NCBI BLASTp program (cut off values: E = 0.01, bit score Z 100, identity Z 35%). This journal is c The Royal Society of Chemistry 2013 Paper In all cases, the name of the protein or the functionality was matched during the selection. Classification and functional annotations of common conserved bacterial proteins The common conserved inter-species bacterial proteins were functionally classified as per the Clusters of Orthologous Groups classifications (COGs).53 E. coli genes were subjected to the COGNITOR BLAST (using default parameters) to group the proteins under each COG functional classifications. Each class of COG consists of evolutionary conserved (at least 3 distant lineages) individual protein or groups of paralogs having similar cellular function under 18 classes. Therefore, the COG database and its classification are very useful in comparative, evolutionary, and phylogenetic analysis of new genome or gene to assign their biological functions.54 Additionally, the proteins were annotated for their functionality using the NCBI and UniProt55 databases. Pathogenicity islands (PAIs) encode various virulence factors including type III secretion system proteins of a bacterium that are required for infection. Hence, to check the virulence of the common bacterial proteins, each protein was tested with the help of the BLASTp option at the Pathogenicity Island Database (PAIDB) server.56 The PIDB contains all reported PAIs from 497 pathogenic bacterial strains. The database also contains more than 310 predicted PAIs from 118 prokaryots. To map the pathway involvements of these conserved proteins, we used the KEGG pathway database.57 Generation of intra- and inter-species bacterial PPI, validation, and analysis The bacterial PPIs were developed and analyzed using VisANT 3.0.58,59 VisANT is an integrative platform for developing PPIs and network prediction, construction, editing, analysis, and visualization. It develops biological interactions based on data derived from 102 methods (computational and both high- and low- throughput experimental methods). The tool can integrate and mine KEGG57 pathways in biological interactions and multi-scale analysis and visualization of multiple pathways can also be done. Intra-species PPI of four Cp. The Cp genome is not available in VisANT. Therefore, a combination of genomic context-based methods including comparative and phylogenetic profiling,14 gene or domain fusion,15 and gene neighborhood methods13 were used to develop the intra-species PPIs for the conserved C. pseudotuberculosis proteins of the selected four Cp strains. The resultant PPIs along with KEGG pathways were incorporated in the VisANT for network analysis and in silico validation of the intra-species Cp PPIs. Inter-species bacterial PPI. The common conserved interspecies bacterial PPIs for Y. pestis, E. coli, M. tuberculosis, C. glutamicum, C. diphtheriae, and C. ulcerans and all four Cp strains were developed using VisANT. We used common conserved E. coli K12 proteins to develop this PPI as multiple experimentally validated data for E. coli PPIs are available in VisANT. Additionally, the VisANT generated E. coli based conserved PPIs were evaluated using anti-tag co-immunoprecipitation-based Integr. Biol. Paper binding PPIs from E. coli.60 Next, the COG-based classification was applied to construct interacting protein hubs (a group of proteins under a common COG). Further, KEGG pathways were incorporated into the PPI network and analyzed in VisANT for identification of correlations among the interacting individual proteins, hubs, connecting nodes, and pathways to determine if the selected common conserved proteins and their PPIs are involved in bacterial essential metabolic process as well as in pathogenesis. This is with the agreement that as we have taken common conserved proteins of multiple pathogenic and non-pathogenic bacteria from same and different taxa; the proteins and their resultant interspecies PPIs must be involved in bacterial essential metabolic as well as pathogenic pathways. Host–pathogen protein–protein interactions (HP-PPIs) Cp infects a broad range of hosts, commonly goat, sheep, and horse,51 and in rare cases, human.42 However, the other pathogens investigated in this analysis do affect humans. With the exception of the human host, the genomes of the other hosts (goat, sheep and horse) have not been fully characterized. It is presumed that the goat, sheep and horse genomes have protein products similar to those of human, as they are higher mammals.51 Several symptoms are shared between a Cp, Y. pestis,45–47 and Mycobacterium48,49 infection. Mycobacterium also falls under the same bacterial group of Cp (the CMNR group of pathogens). Therefore, we used our identified common conserved proteins (that interact with each other and forms common conserved PPIs) in our previous analysis step (inter-species PPIs) to generate a common conserved host–pathogen interaction that will be common to all the selected pathogens and hosts. Although several computational approaches based HP-PPIs have been reported over time for a number of pathogens,26,28,61,62 instead of using computational methods, we made our HP-PPIs based on published experimentally validated host–pathogen protein–protein binding data. To achieve the HP-PPIs; yeast two-hybrid assay based Y. pestis-human PPIs,5,63 liquid chromatography-tandem mass spectrometry based surface-affinity profiling data for S. gallolyticus-human PPIs,64 and protein microarray based streptococcu-human PPIs29 were extracted from corresponding published literatures. Although the yeast two-hybrid screens generate significant degree of false negatives interactions,65 we had no other option to generate the host pathogen PPIs because of unavailability of any other high throughput experimental data. In addition to these literature based data, 7180 experimentally validated host–pathogen protein binding interactions for 21 pathogens with the human proteins from the PathoSystems Resource Integration Center (Patric) database66 and 24 253 PPIs between 58 hosts and 416 pathogen species from HPIDB database67 were downloaded to enrich our interaction data. While the Patric contains interactions of bacterial proteins with only human; the HPIDB provides PPIs data for multiple hosts (including human, mouse, rat, and bovine, chicken etc.). Next, the identified common conserved bacterial proteins those interact with each other in intra-species bacterial PPIs were manually correlated with human interacting counterparts Integr. Biol. Integrative Biology based on the collected experimentally validated host–pathogen interaction data. In some cases, the correlation was difficult as the interacting partner protein from the bacteria was from species that is not considered in our analysis. Therefore, we used comparative genomics BLAST to identify if the interacting bacterial partner is a homologue to any of our selected common conserved bacterial proteins and if there is a 435% identity, we considered the interaction for our purpose. Towards validating and determining the significance of the HP-PPIs To identify and evaluate the significance of the host–pathogen interactions involved in the host response to the pathogenesis and the key bacterial proteins involved in the pathogenesis, we performed two analyses of the HP-PPIs. First, we performed gene set enrichment and enriched functional clustering based on Gene Ontology using the well known tool: Database for Annotation Visualization and Integrated Discovery (DAVID Vs6.7)68 for the host proteins in the HP-PPIs. Further, we used ToppGene69 for candidate gene prioritization, identification of network key nodes, and centrality analysis of the interacting host proteins by mapping their involvement in host pathways affected due to infection. ToppGene is a platform for gene set enrichment, functional annotations, and protein interactions network based candidate gene prioritization. It also provides information about relative importance of a candidate gene in a PPI network. For ToppGene analysis, the training sets for the respective biological processes were collected from data available at the Molecular signature Database (MsigDB).70 The key biological processes were selected that are modulated within the host such as TLR signaling and inflammatory pathways, immunity, cytoskeleton reorganization, phagocytosis, and apoptosis in response to infection of Y. pestis, E. coli, M. tuberculosis and several other pathogenic bacteria as described in manually curated PHIDIAS host–pathogen interactions database.71 Finally, the interacting pathway-specific key host proteins were selected based on the ToppGene analysis. The key bacterial proteins in the HP-PPIs that are involved in the pathogenesis were identified based on the functionality analysis. The functional annotation was done using the NCBI, UniProt,55 and KEGG databases.57 Additionally, the sub-cellular localization of the proteins were determined using CELLO72 and ‘‘Effective’’73 tools. While CELLO identifies extracellular, outer membrane, inner membrane, periplasmic, and cytoplasmic proteins; the ‘‘Effective’’ specifically predicts bacterial secreted proteins. The virulence was checked using PAIDB database.56 Identification of targets from the host–pathogen PPIs and virtual screening From the host–pathogen interaction network, the interacting essential non-host homolog bacterial proteins were identified as probable targets based on the method and criteria as described by Barh et al., 2011.51 Briefly, the interacting essential bacterial proteins were selected based on Database of Essential Genes (DEG)74 BLASTp (cut off values: E = 0.01, bit score Z 100, This journal is c The Royal Society of Chemistry 2013 Integrative Biology Paper Fig. 1 Simple flow diagram of the overall strategy used to develop intra-species Cp PPI, inter-species bacterial PPI, host–pathogen interaction PPI, and identification of targets from the host–pathogen interactions. identity Z 35%). Further, the non-host essential bacterial homologs were identified by subjecting essential proteins in NCBI BLASTp program against human, mouse, sheep, horse, and bovine proteomes. Finally, the bacterial essential non-host homolog core and PAI associated proteins having r100 KDa molecular weight and are involved in bacteria’s multiple unique essential metabolic pathways were selected as putative targets. The bacterial targets were modeled using the Phyre 275 and Swiss model servers76 and validated using the SAVS server Vs.4. (http://services.mbi.ucla.edu/SAVES/). A ligand library was developed with 30 well known antibiotics used against the selected pathogens and effective drugs for Cp.77 In India, a Cp infection is rare in areas where the cattle feed on betel vine leaves and stalks. Therefore, 120 compounds derived from betel vine were also used to enrich the ligand library and for testing these betel compounds on the identified targets. The catalytic pockets within the target proteins were determined using Molegro Virtual Docker.78 The docking was performed using GOLD software79 and the five best ligands based on their GOLD score. The overall strategy is represented in Fig. 1. Growth inhibitory effect of Piper betel compounds: preliminary validation The best lead compounds from Piper betel were tested for their individual growth inhibition efficacy against the pathogenic E. coli O157:H7. The bacteria were cultured in Mueller Hinton (MH) broth (Sigma-Aldrich Co. LLC) at 37 1C for 6 hours to reach the log phase. Then, cells were harvested by centrifugation and 107 CFU mL 1 cells were resuspended in tubes This journal is c The Royal Society of Chemistry 2013 containing MH broth and 10, 100 mM or 1, 10, and 100 mM concentrations of the Piper betel compounds. Treatment with 100 mg ml 1 of ampicillin was used as control. Cultures were then incubated at 37 1C for 2 hours in a shaker. The number of colony-forming units (CFUs) was counted each 30 min interval by obtaining the CFU/ml from serial 10-fold dilutions prepared in MH agar (Sigma-Aldrich Co. LLC). Results Bacterial protein–protein interactions Common conserved intra-strain PPI in Cp. We identified 1783 genes common to our 4 Cp selected strains. Using the computational approaches, we found 4186 conserved interactions common to these Cp strains. We found total 874 proteins are involved in these interactions. The number of predicted PPIs based on phylogenetic profile, domain fusion, and gene neighborhood methods are 2392, 2388, and 245, respectively. To analyze the pathways falling in these conserved interactions, we fed the PPIs and Cp FRC41 metabolic pathways (obtained from KEGG) into VisANT. Upon analysis, we found that 68 pathways can be mapped in this intra-strain PPI of the Cp. These pathways include various metabolisms, two component systems, ABC transporters, and bacterial secretion systems among others that are important for bacterial survival and pathogenesis. Therefore, our selected conserved common proteins and the developed intra-strain PPI of Cp will be useful to explain the biology and pathogenesis of the bacteria if further analyzed. Integr. Biol. Paper Although this PPI of Cp is very preliminary of its kind, we are reporting it because there is no report so far on Cp PPI. As our main aims are to develop conserved common inter-species bacterial PPIs and use the same to develop conserved common host–pathogen interactions to finally identify conserved common broad spectrum target; we did not analyze the intrastrain PPI of Cp in detail. Inter-species common conserved bacterial PPIs. To generate the common conserved inter-species bacterial PPIs, first we identified common conserved proteins in Y. pestis CO92, E. coli K-12 DH10B, E. coli O157:H7, M. tuberculosi H37Rvs, C. diphtheriae, and C. glutamicum R using PAST server. Seventy eight proteins were found to be conserved in all these species. Further, we checked if all these proteins are conserved in other virulent and non-virulent strains of various strains of these bacteria and Cp strains i.e. from closed and distance taxa. To achieve this we used amino acid sequences of these 78 Y. pestis CO92 proteins and performed comparative BLASTp in NCBI server against proteomes of E. coli str. K-12 substr. MG1655, C. glutamicum ATCC 13032 Kitasato, C. urealyticum DSM 7109, M. tuberculosis CDC1551, M. ulcerans Agy99, and four of our Cp strains (FRC41, 316, 3/99-5, and P54B96). We found all these 75 proteins are conserved in all these selected species and strains (Table S2, ESI†). As various experimental PPI data are available for E. coli str. K-12, we selected conserved 75 proteins of this species to make the common conserved inter-species PPIs using VisANT. Integrative Biology In VisANT, these 75 proteins form a PPI network with 1674 interactions involving 666 interacting nodes where 1210, 755, and 281 interactions are based on the tandem affinity purification, inferred by authors, and anti tag co-immunoprecipitation methods, respectively. There are interactions based on computational and other experimental methods such as cross-linking studies among others. Twenty seven total pathways were mapped in this PPI (Table S3a, ESI†). However, while we did internal interactions among these 75 proteins, we found only 142 interactions involving 23 pathways (Table S3b, ESI†). These 75 interacting proteins fall under 14 COGs (Fig. 2) and with the exception of 3 proteins, all other proteins were found to be virulent as per the PAIDB – BLASTp analysis (Table S2, ESI†). We selected pathogenic and non-pathogenic organisms from the same and distant taxa and their conserved genes to make the inter-species PPIs. Therefore, the resultant PPIs are common and conserved in all the bacterial species considered and the PPIs should involve pathways that are essential for bacterial survival as well as for pathogenesis. To check this, KEGG pathways were incorporated in the PPIs using VisANT’s ‘‘expand pathways’’ option and the interactions along with the pathways were analyzed. The analysis showed that the interacting networks were well linked and fit with various pathways that are well known for their involvements in bacterial survival and virulence such as various metabolism, two-component Fig. 2 Clusters of Orthologous Groups (COG) classifications of common conserved proteins of four C. pseudotuberculosis strains, Y. pestis, M. tuberculosis, C. glutamicum, C. diphtheriae, C. ulcerans, and E. coli. Integr. Biol. This journal is c The Royal Society of Chemistry 2013 Integrative Biology Paper Fig. 3 The conserved common PPIs with COG classifications of Cp FRC41, Cp 316, Cp 3/99-5, Cp P54B96, Y. pestis, M. tuberculosis, C. gluticum, C. diptherae, C. ulcerans, and E. coli. Important bacterial pathways involving these proteins and the relationship of these proteins and pathways are also shown. The relationships (edgs) between hubs and individual proteins are determined using VisANT. system,80 ABC transporter,81,82 redox signaling,83 and sphingolipid metabolism84,85-like pathways (Fig. 3), supporting the accuracy and significance of our PPIs. Host–pathogen protein–protein interactions (HP-PPIs) To make the HP-PPIs, we used the conserved bacterial proteins that interact with at least another protein of the bacteria in the inter-species bacterial PPI. Using the procedure described in the methods and such conserved interacting proteins, we This journal is c The Royal Society of Chemistry 2013 identified 14 bacterial proteins that interact with 122 host proteins. Functional annotations of these bacterial proteins revealed that eight are cytoplasmic enzymes and five are membrane localized. All these 14 proteins were predicted to be involved in virulence as per the PAIDB and the DEG based analysis showed; all these proteins are encoded by essential genes. Further, the functional annotation of these 14 proteins revealed that, they are involved in bacterial various essential metabolic pathways as well as pathogenicity-related pathways Integr. Biol. Paper Integrative Biology Fig. 4 Common conserved host–pathogen interaction network of multiple pathogens (four C. pseudotuberculosis strains, Y. pestis, M. tuberculosis, C. glutamicum, C. diphtheriae, C. ulcerans, and E. coli) and their usual hosts. such as two-component systems (dnaA) and ABC transporters (gluA) (Table S4, ESI†). All 122 interacting host proteins were found in well-known bacterial infection associated host pathways such as integrin-mediated signaling, endocytosis, TLR signaling, immunity, apoptosis, inflammation, and redox signaling71 (Table S5, ESI†). The ToppGene-based gene set enrichment analysis ranked CTNB1 and PIK3R1 at positions one and four, respectively. Both proteins interact with rpoB and are involved in immunity, apoptosis, and cell matrix adhesion (Table S6, ESI†). The bacterial proteins rpoB, carA, carB, leuD, groEL and their host interacting partners IGHV4-31, NFKB1, CHD8, and C12orf35, respectively were the key nodes in the host–pathogen protein–protein interaction network based on the degree of interactions and centrality analysis (Fig. 4). Drug target and lead selection From the host–pathogen protein–protein interaction network, we identified common conserved bacterial targets using Integr. Biol. subtractive genomics as described by Barh et al., 2011.51 The 14 identified genes were essential for the selected group of pathogens, and the cytoplasmic Acetate kinase (Ack) [EC = 2.7.2.1, Mass = 43.3 KDa] involved in the metabolism of taurine, hypotaurine, pyruvate, propanoate, and methane metabolism is the only non-host homolog satisfying most of the criteria of an ideal target for being (a) an essential non-host homolog enzyme for multiple organisms, (b) core gene of the organisms, (c) involvement in organisms’ multiple unique and essential pathways, (d) PAI-related enzyme, and (e) less than 100 KDa molecular weight51 (Table S4, ESI†). This common conserved target binds to host PRDX3 in yeast two hybrid assay (Fig. 4). PRDX3 is involved in the immune system, apoptosis, cell proliferation, and redox signaling-like pathways. Therefore, interaction of Ack-PRDX3 affects all these biological processes in the host, supporting a mechanism of bacterial infection. Four active sites were found in the modeled Ack using the Molegro Virtual Docker (Table S7, ESI†). The GOLD fitness This journal is c The Royal Society of Chemistry 2013 Integrative Biology score and MVD analysis of the docking showed that of the group of 30 selected antibiotics, ceftiofur and penicillin, commonly used to treat Cp, Diphtheria, Tuberculosis, and Y. pestis infections, were probably effective against Ack (Fig. 5 and Table S8, ESI†). Additionally, piperdardine and dehydropipernonaline derived from Piper betel were also predicted to be effective and possibly had a similar or superior inhibitory activity against the target as compared to penicillin and ceftiofur (Table S8, ESI†). Piperdardine inhibits E. coli O157:H7 growth Viable cells were counted during the culture in MH media containing the compounds in order to investigate their growthinhibiting effect on E. coli O157:H7. We observed that addition of 100.0 mM of piperdardine or their higher concentration dramatically decrease in the CFU counts, similar to bacteria treated with ampicillin (Fig. 6). Paper Discussion PPIs derived information along with a molecular basis for host– pathogen interactions are important in finding effective targets against a pathogen. Computational or high-throughput approaches based on the development of genome- or proteome-wide PPI networks have been applied to various organisms,9,10,18,19,26–29 allowing for the extraction of important information for specific biological processes. Predicted host–pathogen PPIs have been reported for HIV,86,87 Dengue virus,88 Mycobacterium, apicomplexa, kinetoplastida,28 and P. falciparum.26,27 Experimentally validated interactions and their implementations in drug or vaccine development against the various pathogens have also been reported for group-B streptococcus,29 Corynebacterium diphtheriae,36,89 M. tuberculosis,31–33 Yersinia pestis,90 and Yersinia pseudotuberculosis.91 However, these experiments were conducted for a small fraction of pathogenic proteins. Recently, yeast-two hybrid-based proteome-wide Fig. 5 Docking of Ack with ceftiofur (A–B), penicillin (C–D), piperdardine (E–F), and dehydropipernonaline (G–H). Fig. 6 Inhibitory effects of Piperdardine on the growth of Escherichia coli 0157:H7 as compared to ampicillin. This journal is c The Royal Society of Chemistry 2013 Integr. Biol. Paper host–pathogen protein–protein binding interactions were reported for B. anthracis, F. tularensis, and Y. pestis,5 and a number of novel interactions were documented for these pathogens. In this report, for the first time we represent 4186 common conserved intra-species PPIs for four Cp strains (Cp FRC41, Cp 316, Cp 3/99-5, and Cp P54B96) using phylogenetic profile, domain fusion, and gene neighborhood methods. In Cp, we found 874 proteins are involved in these interactions. The recently reported experimental PPI data on M. tuberculosis H37Rv, another CMNR group of pathogens, revealed B8000 novel interactions.92 When we compared our intra-species PPIs of Cp with these M. tuberculosis data, we found half of the number of M. tuberculosis interactions in Cp. This difference may be due to the larger genome size of M. tuberculosis (4062 genes, almost double that of the Cp genome), the methods applied, and the phylogenetically conserved proteins in Cp. The sixty eight pathways mapped in the Cp PPI belong to both the bacterial essential metabolic and virulence pathways. Therefore, our developed Cp PPI will be significant in explaining biology and pathology of Cp upon further analysis. While we developed the inter-species PPIs using common phylogenetically conserved proteins of different groups of organisms (pathogenic, non-pathogenic, and same and distance taxa), including four Cp strains, Y. pestis, M. tuberculosis, C. glutamicum, C. diphtheriae, C. ulcerans, and E. coli, we only observed 75 common interacting proteins that constituted a network of 142 interactions among each other and 1674 interactions involving 666 proteins to form the PPI network; however, important essential metabolic pathways and virulence related pathway can be mapped in these networks supporting the usefulness of the PPI in describing the common physiological process and virulence of these selected pathogens. It is also profound from these results that, the species-specific global PPIs exhibit a large number of interactions. However, number of interactions in conserved PPIs across a distantly related species of similar pathogenesis is reduced drastically, although essential and important pathogenesis-related proteins and pathways were found in the network. Human-based host–pathogen interactions have been reported for a number of individual pathogens.5,29,63,64 However, a common conserved HP-PPI for a number of pathogens and hosts have not been reported. Here, for the first time we used common conserved proteins from a broad spectrum of hosts (human, goat, sheep, and horse) to study the interactions. Additionally, for the first time, we have extended the strategy to generate conserved and common host–pathogen interactions for a group of pathogens using inter-species common conserved interacting proteins of Cp, Y. pestis, M. tuberculosis, C. diphtheriae, C. ulcerans, and E. coli with a mode of pathogenesis common to these selected hosts. This strategy helped to gain insight into common conserved host–pathogen interactions across a wide range of organisms and to identify broad spectrum targets in a single analysis. The PAI-related proteins are thought to be involved in pathogenesis.93 Our results support this finding, and we found that the 14 identified conserved pathogen proteins involved in host–pathogen Integr. Biol. Integrative Biology interactions were located in PAIs. These proteins are also involved in essential metabolic and virulence pathways. Similarly, GSEA, candidate gene prioritization, key nodes, and centrality analysis of the interacting host proteins revealed that they are involved in most of the infection-related signaling pathways,71 supporting the rationality of the developed host– pathogen interaction networks. Based on the strategy of target identification,51 Ack was selected as a broad spectrum target from the host–pathogen interaction network. Ack is essential to E. coli,94 M. genitalium,95 and M. pulmonis96 and is predicted to be a target in S. aureus.97 The HP-PPI showed that Ack interacted with Peroxiredoxin 3 (PRDX3) from the host. PRDX3 is a peroxidase and is involved in the NF-kappaB cascade, cell proliferation, apoptosis, and redox signaling. Redox-sensitive proteins in pathogens make them resistant to oxidative stress and antibiotics,98 and manipulation of the redox state can be an important strategy for the management of Tuberculosis.99 Ack, our identified target, is a kinase that interacts with the redox protein PRDX3 of the host. We hypothesized that the binding of Ack to PRDX3 modulates PRDX3 activity, thereby disrupting the redox signaling and immune system of the host. This interaction may help in SOD-mediated fibrocyte activation and scar or abscess formation100 in lymphadenitis, the common symptom of Cp and Y. Pestis infections. It may also be a vital mechanism for drug resistance in these pathogens, disrupting the host redox system. However, to interfere mitochondrial functions during pathogenesis, a bacterial protein needs to reach and bind to mitochondrial protein of the host.101 Bacteria that possess type III and type IV secretion system like injection machinery can directly inject bacterial proteins into the host cell cytoplasm during infection process.102,103 As per the ‘‘Effective’’73 prediction, Ack of M. tuberculosis H37Rv is a type III secreted protein and according to Couto et al. (2012), Ack is probably secreted or localizes to bacterial surface during M. mycoides infection in cattle and plays a role in immunogenic responses in the host.104 Therefore, it might be possible that Ack is injected into host cell through bacterial secretion system during infection and upon resealed into the host cytoplasm it interacts with mitochomdrial PRDX3. However, it should be proved experimentally and this is one of the future scopes of this research. Virtual screening showed that ceftiofur and penicillin could be effective antibiotics against the selected pathogens considering the target Ack. The natural products piperdardine and dehydropipernonaline from Piper betel had shown a similar or superior effect on Ack as per our in silico analysis. Until now, no experimental data were available that tested the efficacy of compounds targeted to Ack, and validation is thereby necessary using conventional antibiotics and our identified Piper betel compounds. The leaf extract of Piper betel has proven to be useful as an antimicrobial,105,106 antioxidant,107 anti-inflammatory,108 and immunomodulator.109 However, the specific compounds in the plant that produce these properties are yet to be determined. In our preliminary validation, we observed that, 100.0 mM of piperdardine inhibits E. coli O157:H7 growth This journal is c The Royal Society of Chemistry 2013 Integrative Biology similar to penicillin. Therefore, it is presumed that these compounds may also be effective against other pathogens considered in this work. We are currently testing the bactericidal effects of these betel compounds against C. pseudotuberculosis, C. diphtheriae, M. tuberculosis, C. ulcerans, and Y. pestis and their target specificity to Ack. The results are highly promising. Paper 6 7 Conclusion This study demonstrates intra-species PPI for Cp and illustrates the potential and importance of inter-species bacterial protein– protein and host–pathogen interactions in broad spectrum target identification. We report the conserved intra-species PPIs of Cp and a common conserved host pathogen-interaction network for Y. pestis, M. tuberculosis, C. diphtheriae, C. ulcerans, E. coli, and four Cp strains. Ack was identified as a broad spectrum target for all these pathogens considering human, goat, sheep, and horse as hosts. Ceftiofur, penicillin and two natural compounds derived from Piper betel, piperdardine and dehydropipernonaline, were predicted to be effective against Ack activity. Validation shows piperdardine is a highly effective antibacterial agent. The in silico approaches used in this work were supposed to be effective in developing and analyzing interspecies global bacterial PPIs as well as host–pathogen interactions to identify drug targets. 8 9 10 Competing interests The authors declare that they have no competing interests. Financial disclosure 11 This work was carried out without any financial support or grant. Note added after first publication 12 This article replaces the version published on 3rd January 2013, which contained an error in that the title was incomplete. The subtitle has been added to clarify this. 13 References 14 1 R. Sharan, I. Ulitsky and R. Shamir, Network-based prediction of protein function, Mol. Syst. Biol., 2007, 3, 88. 2 E. D. Levy and J. B. Pereira-Leal, Evolution and dynamics of protein interactions and networks, Curr. Opin. Struct. Biol., 2008, 18, 349–357. 3 F. Hormozdiari, R. Salari, V. Bafna and S. C. Sahinalp, Protein–protein interaction network evaluation for identifying potential drug targets, J. Comput. Biol., 2010, 17, 669–684. 4 Y. Y. Wang, J. C. Nacher and X. M. Zhao, Predicting drug targets based on protein domains, Mol. BioSyst., 2012, 8, 1528–1534. 5 M. D. Dyer, C. Neff, M. Dufford, C. G. Rivera, D. Shattuck, J. Bassaganya-Riera, T. M. Murali and B. W. Sobral, The This journal is c The Royal Society of Chemistry 2013 15 16 17 18 human-bacterial pathogen protein interaction networks of Bacillus anthracis, Francisella tularensis, and Yersinia pestis, PLoS One, 2010, 5, e12089. J. Shen, J. Zhang, X. Luo, W. Zhu, K. Yu, K. Chen, Y. Li and H. Jiang, Predicting protein–protein interactions based only on sequences information, Proc. Natl. Acad. Sci. U. S. A., 2007, 104(11), 4337–41. X. W. Zhao, Z. Q. Ma and M. H. Yin, Predicting protein– protein interactions by combing various sequence-derived features into the general form of Chou’s Pseudo amino acid composition, Protein Pept. Lett., 2011, 19(5), 492–500. L. Skrabanek, H. K. Saini, G. D. Bader and A. J. Enright, Computational prediction of protein–protein interactions, Mol. Biotechnol., 2008, 38, 1–17. H. Yu, P. Braun, M. A. Yildirim, I. Lemmens, K. Venkatesan, J. Sahalie, T. Hirozane-Kishikawa, F. Gebreab, N. Li, N. Simonis, T. Hao, J. F. Rual, A. Dricot, A. Vazquez, R. R. Murray, C. Simon, L. Tardivo, S. Tam, N. Svrzikapa, C. Fan, A. S. de Smet, A. Motyl, M. E. Hudson, J. Park, X. Xin, M. E. Cusick, T. Moore, C. Boone, M. Snyder, F. P. Roth, A. L. Barabasi, J. Tavernier, D. E. Hill and M. Vidal, Highquality binary protein interaction map of the yeast interactome network, Science, 2008, 322, 104–110. U. Stelzl, U. Worm, M. Lalowski, C. Haenig, F. H. Brembeck, H. Goehler, M. Stroedicke, M. Zenkner, A. Schoenherr, S. Koeppen, J. Timm, S. Mintzlaff, C. Abraham, N. Bock, S. Kietzmann, A. Goedde, E. Toksoz, A. Droege, S. Krobitsch, B. Korn, W. Birchmeier, H. Lehrach and E. E. Wanker, A human protein–protein interaction network: a resource for annotating the proteome, Cell, 2005, 122, 957–968. I. Ispolatov, A. Yuryev, I. Mazo and S. Maslov, Binding properties and evolution of homodimers in protein– protein interaction networks, Nucleic Acids Res., 2005, 33, 3629–3635. M. P. Stumpf, T. Thorne, S. E. de, R. Stewart, H. J. An, M. Lappe and C. Wiuf, Estimating the size of the human interactome, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 6959–6964. T. Dandekar, B. Snel, M. Huynen and P. Bork, Conservation of gene order: a fingerprint of proteins thatphysically interact, Trends Biochem. Sci., 1998, 23, 324–328. M. Pellegrini, E. M. Marcotte, M. J. Thompson, D. Eisenberg and T. O. Yeates, Assigning protein functions by comparative genome analysis: protein phylogenetic profiles, Proc. Natl. Acad. Sci. U. S. A., 1999, 96, 4285–4288. A. J. Enright, I. Iliopoulos, N. C. Kyrpides and C. A. Ouzounis, Protein interaction maps for complete genomes based on gene fusion events, Nature, 1999, 402, 86–90. C. S. Goh, A. A. Bogan, M. Joachimiak, D. Walther and F. E. Cohen, Co-evolution of proteins with their interaction partners, J. Mol. Biol., 2000, 299, 283–293. M. Singhal and H. Resat, A domain-based approach to predict protein–protein interactions, BMC Bioinf., 2007, 8, 199. K. Raman and N. Chandra, Mycobacterium tuberculosis interactome analysis unravels potential pathways to drug resistance, BMC Microbiol., 2008, 8, 234. Integr. Biol. Paper 19 M. Rashid, S. Ramasamy and G. P. Raghava, A simple approach for predicting protein–protein interactions, Curr. Protein Pept. Sci., 2010, 11(7), 589–600. 20 P. M. Bowers, S. J. Cokus, D. Eisenberg and T. O. Yeates, Use of logic relationships to decipher protein network organization, Science, 2004, 306, 2246–2249. 21 D. Barker and M. Pagel, Predicting functional gene links from phylogenetic-statistical analyses of whole genomes, PLoS Comput. Biol., 2005, 1, e3. 22 N. Tuncbag, G. Kar, O. Keskin, A. Gursoy and R. Nussinov, A survey of available tools and web servers for analysis of protein–protein interactions and interfaces, Briefings Bioinf., 2009, 10, 217–232. 23 C. Yeats, J. Lees, P. Carter, I. Sillitoe and C. Orengo, The Gene3D Web Services: a platform for identifying, annotating and comparing structural domains in protein sequences, Nucleic Acids Res., 2011, 39, W546–W550. 24 S. Hunter, P. Jones, A. Mitchell, R. Apweiler, T. K. Attwood, A. Bateman, T. Bernard, D. Binns, P. Bork, S. Burge, C. E. de, P. Coggill, M. Corbett, U. Das, L. Daugherty, L. Duquenne, R. D. Finn, M. Fraser, J. Gough, D. Haft, N. Hulo, D. Kahn, E. Kelly, I. Letunic, D. Lonsdale, R. Lopez, M. Madera, J. Maslen, C. McAnulla, J. McDowall, C. McMenamin, H. Mi, P. Mutowo-Muellenet, N. Mulder, D. Natale, C. Orengo, S. Pesseat, M. Punta, A. F. Quinn, C. Rivoire, A. Sangrador-Vegas, J. D. Selengut, C. J. Sigrist, M. Scheremetjew, J. Tate, M. Thimmajanarthanan, P. D. Thomas, C. H. Wu, C. Yeats and S. Y. Yong, InterPro in 2011: new developments in the family and domain prediction database, Nucleic Acids Res., 2012, 40, D306–D312. 25 J. Lees, C. Yeats, J. Perkins, I. Sillitoe, R. Rentzsch, B. H. Dessailly and C. Orengo, Gene3D: a domain-based resource for comparative genomics, functional annotation and protein network analysis, Nucleic Acids Res., 2012, 40, D465–D471. 26 M. D. Dyer, T. M. Murali and B. W. Sobral, Computational prediction of host–pathogen protein–protein interactions, Bioinformatics, 2007, 23, i159–i166. 27 S. Wuchty, Computational prediction of host-parasite protein interactions between P. falciparum and H. sapiens, PLoS One, 2011, 6, e26960. 28 F. P. Davis, D. T. Barkan, N. Eswar, J. H. McKerrow and A. Sali, Host pathogen protein interactions predicted by comparative modeling, Protein Sci., 2007, 16, 2585–2596. 29 I. Margarit, S. Bonacci, G. Pietrocola, S. Rindi, C. Ghezzo, M. Bombaci, V. Nardi-Dei, R. Grifantini, P. Speziale and G. Grandi, Capturing host–pathogen interactions by protein microarrays: identification of novel streptococcal proteins binding to human fibronectin, fibrinogen, and C4BP, FASEB J., 2009, 23, 3100–3112. 30 H. Huang, B. M. Jedynak and J. S. Bader, Where have all the interactions gone? Estimating the coverage of two-hybrid protein interaction maps, PLoS Comput. Biol., 2007, 3(11), e214. Integr. Biol. Integrative Biology 31 S. Gagneux, K. DeRiemer, T. Van, M. Kato-Maeda, B. C. de Jong, S. Narayanan, M. Nicol, S. Niemann, K. Kremer, M. C. Gutierrez, M. Hilty, P. C. Hopewell and P. M. Small, Variable host–pathogen compatibility in Mycobacterium tuberculosis, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 2869–2873. 32 D. P. Cifuentes, M. Ocampo, H. Curtidor, M. Vanegas, M. Forero, M. E. Patarroyo and M. A. Patarroyo, Mycobacterium tuberculosis Rv0679c protein sequences involved in host-cell infection: potential TB vaccine candidate antigen, BMC Microbiol., 2010, 10, 109. 33 K. Raman, A. G. Bhat and N. Chandra, A systems perspective of host–pathogen interactions: predicting disease outcome in tuberculosis, Mol. BioSyst., 2010, 6, 516–530. 34 Y. Wang, T. Cui, C. Zhang, M. Yang, Y. Huang, W. Li, L. Zhang, C. Gao, Y. He, Y. Li, F. Huang, J. Zeng, C. Huang, Q. Yang, Y. Tian, C. Zhao, H. Chen, H. Zhang and Z. G. He, Global protein–protein interaction network in the human pathogen Mycobacterium tuberculosis H37Rv, J. Proteome Res., 2010, 9, 6665–6677. 35 V. Kolodkina, T. Denisevich and L. Titov, Identification of Corynebacterium diphtheriae gene involved in adherence to epithelial cells, Infect., Genet. Evol., 2011, 11, 518–521. 36 L. Ott, M. Holler, R. G. Gerlach, M. Hensel, J. Rheinlaender, T. E. Schaffer and A. Burkovski, Corynebacterium diphtheriae invasion-associated protein (DIP1281) is involved in cell surface organization, adhesion and internalization in epithelial cells, BMC Microbiol., 2010, 10, 2. 37 L. Ott, M. Holler, J. Rheinlaender, T. E. Schaffer, M. Hensel and A. Burkovski, Strain-specific differences in pili formation and the interaction of Corynebacterium diphtheriae with host cells, BMC Microbiol., 2010, 10, 257. 38 E. Trost, J. Blom, S. S. de Castro, I. H. Huang, A. Al-Dilaimi, J. Schroder, S. Jaenicke, F. A. Dorella, F. S. Rocha, A. Miyoshi, V. Azevedo, M. P. Schneider, A. Silva, T. C. Camello, P. S. Sabbadini, C. S. Santos, L. S. Santos, R. Hirata, Jr., A. L. Mattos-Guaraldi, A. Efstratiou, M. P. Schmitt, H. Ton-That and A. Tauch, Pan-genomics of Corynebacterium diphtheriae: Insights into the genomic diversity of pathogenic isolates from cases of classical diphtheria, endocarditis and pneumonia, J. Bacteriol., 2012, 194(12), 3199–3215. 39 L. H. Williamson, Caseous lymphadenitis in small ruminants, Vet. Clin. North Am.: Food Anim Pract., 2001, 17, 359–371, vii. 40 M. Aleman, S. J. Spier, W. D. Wilson and M. Doherr, Corynebacterium pseudotuberculosis infection in horses: 538 cases (1982–1993), J. Am. Vet. Med. Assoc., 1996, 209, 804–809. 41 R. G. Batey, Pathogenesis of caseous lymphadenitis in sheep and goats, Aust. Vet. J., 1986, 63, 269–272. 42 E. Trost, L. Ott, J. Schneider, J. Schroder, S. Jaenicke, A. Goesmann, P. Husemann, J. Stoye, F. A. Dorella, F. S. Rocha, S. C. Soares, V. D’Afonseca, A. Miyoshi, J. Ruiz, A. Silva, V. Azevedo, A. Burkovski, N. Guiso, O. F. Join-Lambert, S. Kayal and A. Tauch, The complete This journal is c The Royal Society of Chemistry 2013 Integrative Biology 43 44 45 46 47 48 49 50 51 52 53 54 55 56 genome sequence of Corynebacterium pseudotuberculosis FRC41 isolated from a 12-year-old girl with necrotizing lymphadenitis reveals insights into gene-regulatory networks contributing to virulence, BMC Genomics, 2010, 11, 728. G. J. Annas, Bioterror and ‘‘bioart’’ – a plague o’ both your houses, N. Engl. J. Med., 2006, 354, 2715–2720. M. Drancourt, Plague in the genomic area, Clin. Microbiol. Infect., 2012, 18, 224–230. T. Karttunen, K. Nevasaari, O. Rasanen, P. J. Taskinen and M. Alavaikko, Immunoblastic lymphadenopathy with a high serum Yersinia enterocolitica titer. A case report, Cancer, 1983, 52, 2281–2284. S. J. Nesbitt, L. O. Neville, F. R. Scott and D. M. Flynn, Yersinia pseudotuberculosis in a 3 year old and rapid response to cefotaxime, J. R. Soc. Med., 1994, 87, 418–419. J. E. Comer, D. E. Sturdevant, A. B. Carmody, K. Virtaneva, D. Gardner, D. Long, R. Rosenke, S. F. Porcella and B. J. Hinnebusch, Transcriptomic and innate immune responses to Yersinia pestis in the lymph node during bubonic plague, Infect. Immun., 2010, 78, 5086–5098. P. R. Mohapatra and A. K. Janmeja, Tuberculous lymphadenitis, J. Assoc. Physicians India, 2009, 57, 585–590. J. Knox, G. Lane, J. S. Wong, P. G. Trevan and H. Karunajeewa, Diagnosis of Tuberculous Lymphadenitis Using Fine Needle Aspiration Biopsy, Int. Med. J., 2012, DOI: 10.1111/j.1445-5994.2012.02748.x. S. F. Altschul, T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, W. Miller and D. J. Lipman, Gapped BLAST and PSI-BLAST: a new generation of protein database search programs, Nucleic Acids Res., 1997, 25, 3389–3402. D. Barh, N. Jain, S. Tiwari, B. P. Parida, V. D’Afonseca, L. Li, A. Ali, A. R. Santos, L. C. Guimaraes, S. S. de Castro, A. Miyoshi, A. Bhattacharjee, A. N. Misra, A. Silva, A. Kumar and V. Azevedo, A novel comparative genomics analysis for common drug and vaccine targets in Corynebacterium pseudotuberculosis and other CMN group of human pathogens, Chem. Biol. Drug Des., 2011, 78, 73–84. C. Fong, L. Rohmer, M. Radey, M. Wasnick and M. J. Brittnacher, PSAT: a web tool to compare genomic neighborhoods of multiple prokaryotic genomes, BMC Bioinf., 2008, 9, 170. R. L. Tatusov, D. A. Natale, I. V. Garkavtsev, T. A. Tatusova, U. T. Shankavaram, B. S. Rao, B. Kiryutin, M. Y. Galperin, N. D. Fedorova and E. V. Koonin, The COG database: new developments in phylogenetic classification of proteins from complete genomes, Nucleic Acids Res., 2001, 29, 22–28. M. Kaufmann, The Role of the COG Database in Comparative and Functional Genomics, Curr. Bioinf., 2006, 1, 291–300. M. Magrane and U. Consortium, UniProt Knowledgebase: a hub of integrated protein data. Database, Oxford, 2011, bar009. S. H. Yoon, Y. K. Park, S. Lee, D. Choi, T. K. Oh, C. G. Hur and J. F. Kim, Towards pathogenomics: a web-based This journal is c The Royal Society of Chemistry 2013 Paper 57 58 59 60 61 62 63 64 65 66 67 68 69 resource for pathogenicity islands, Nucleic Acids Res., 2007, 35, D395–D400. M. Kanehisa and S. Goto, KEGG: kyoto encyclopedia of genes and genomes, Nucleic Acids Res., 2000, 28, 27–30. Z. Hu, J. Mellor, J. Wu and C. DeLisi, VisANT: an online visualization and analysis tool for biological interaction data, BMC Bioinf., 2004, 5, 17. Z. Hu, D. M. Ng, T. Yamada, C. Chen, S. Kawashima, J. Mellor, B. Linghu, M. Kanehisa, J. M. Stuart and C. DeLisi, VisANT 3.0: new modules for pathway visualization, editing, prediction and construction, Nucleic Acids Res., 2007, W625–W632. G. Butland, J. M. Peregrin-Alvarez, J. Li, W. Yang, X. Yang, V. Canadien, A. Starostine, D. Richards, B. Beattie, N. Krogan, M. Davey, J. Parkinson, J. Greenblatt and A. Emili, Interaction network containing conserved and essential protein complexes in Escherichia coli, Nature, 2005, 433, 531–537. N. Tyagi, O. Krishnadev and N. Srinivasan, Prediction of protein–protein interactions between Helicobacter pylori and a human host, Mol BioSyst., 2009, 5(12), 1630–1635. O. Krishnadev and N. Srinivasan, Prediction of protein– protein interactions between human host and a pathogen and its application to three pathogenic bacteria, Int. J. Biol. Macromol., 2011, 48(4), 613–619. H. Yang, Y. Ke, J. Wang, Y. Tan, S. K. Myeni, D. Li, Q. Shi, Y. Yan, H. Chen, Z. Guo, Y. Yuan, X. Yang, R. Yang and Z. Du, Insight into bacterial virulence mechanisms against host immune response via the Yersinia pestis-human protein–protein interaction network, Infect. Immun., 2011, 79(11), 4413–4424. A. Boleij, C. M. Laarakkers, J. Gloerich, D. W. Swinkels and H. Tjalsma, Surface-affinity profiling to identify host–pathogen interactions, Infect. Immun., 2011, 79(12), 4777–4783. T. Stellberger, R. Häuser, A. Baiker, V. R. Pothineni, J. Haas and P. Uetz, Improving the yeast two-hybrid system with permutated fusions proteins: the Varicella Zoster Virus interactome, Proteome Sci., 2010, 8, 8. J. J. Gillespie, A. R. Wattam, S. A. Cammer, J. L. Gabbard, M. P. Shukla, O. Dalay, T. Driscoll, D. Hix, S. P. Mane, C. Mao, E. K. Nordberg, M. Scott, J. R. Schulman, E. E. Snyder, D. E. Sullivan, C. Wang, A. Warren, K. P. Williams, T. Xue, H. S. Yoo, C. Zhang, Y. Zhang, R. Will, R. W. Kenyon and B. W. Sobral, PATRIC: the comprehensive bacterial bioinformatics resource with a focus on human pathogenic species, Infect. Immun., 2011, 79, 4286–4298. R. Kumar and B. Nanduri, HPIDB – a unified resource for host–pathogen interactions, BMC Bioinf., 2010, 11(Suppl 6), S16. D. W. Huang, B. T. Sherman and R. A. Lempicki, Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources, Nat. Protoc., 2009, 4(1), 44–57. B. E. A. B. J. A. Chen, ToppGene Suite for gene list enrichment analysis and candidate gene prioritization, Nucleic Acids Res., 2009, 37. Integr. Biol. Paper 70 A. Subramanian, P. Tamayo, V. K. Mootha, S. Mukherjee, B. L. Ebert, M. A. Gillette, A. Paulovich, S. L. Pomeroy, T. R. Golub, E. S. Lander and J. P. Mesirov, Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles, Proc. Natl. Acad. Sci. U. S. A., 2005, 102, 15545–15550. 71 Z. Xiang, Y. Tian and Y. He, PHIDIAS: a pathogen-host interaction data integration and analysis system, Genome Biol., 2007, 8(7), R150. 72 C. S. Yu, C. J. Lin and J. K. Hwang, Predicting subcellular localization of proteins for Gram-negative bacteria by support vector machines based on n-peptide compositions, Protein Sci., 2004, 13, 1402–1406. 73 M. A. Jehl, R. Arnold and T. Rattei, Effective–a database of predicted secreted bacterial proteins, Nucleic Acids Res., 2011, D591–D595. 74 R. Zhang, H. Y. Ou and C. T. Zhang, DEG: a database of essential genes, Nucleic Acids Res., 2004, 1, D271–D272. 75 L. A. Kelley and M. J. Sternberg, Protein structure prediction on the Web: a case study using the Phyre server, Nat. Protoc., 2009, 4, 363–371. 76 K. Arnold, L. Bordoli, J. Kopp and T. Schwede, The SWISSMODEL workspace: a web-based environment for protein structure homology modelling, Bioinformatics, 2006, 22, 195–201. 77 S. J. Spier, Corynebacterium pseudotuberculosis infection in horses: An emerging disease associated with climate change?, Equine vet. Educ, 2008, 20, 37–39. 78 R. Thomsen and M. H. Christensen, MolDock: a new technique for high-accuracy molecular docking, J. Med. Chem., 2006, 49, 3315–3321. 79 M. L. Verdonk, J. C. Cole, M. J. Hartshorn, C. W. Murray and R. D. Taylor, Improved protein-ligand docking using GOLD, Proteins, 2003, 52, 609–623. 80 T. Tobe, The roles of two-component systems in virulence of pathogenic Escherichia coli and Shigella spp, Adv. Exp. Med. Biol., 2008, 631, 189–99. 81 J. S. Klein and O. Lewinson, Bacterial ATP-driven transporters of transition metals: physiological roles, mechanisms of action and roles in bacterial virulence, Metallomics, 2011, 3(11), 1098–1108. 82 V. G. Lewis, M. P. Ween and C. A. McDevitt, The role of ATP-binding cassette transporters in bacterial pathogenicity, Protoplasma, 2012, 249(4), 919–942. 83 A. Trivedi, N. Singh, S. A. Bhat, P. Gupta and A. Kumar, Redox biology of tuberculosis pathogenesis, Adv. Microbiol. Physiol., 2012, 60, 263–324. 84 L. J. Heung, C. Luberto and M. Del Poeta, Role of sphingolipids in microbial pathogenesis, Infect. Immun., 2006, 74(1), 28–39. 85 D. An, C. Na, J. Bielawski, Y. A. Hannun and D. L. Kasper, Membrane sphingolipids as essential molecular signals for Bacteroides survival in the intestine, Proc. Natl. Acad. Sci. U. S. A., 2011, 108(Suppl 1), 4666–4671. 86 P. Evans, W. Dampier, L. Ungar and A. Tozeren, Prediction of HIV-1 virus-host protein interactions using virus and host sequence motifs, BMC Med. Genomics, 2009, 2, 27. Integr. Biol. Integrative Biology 87 O. Tastan, Y. Qi, J. G. Carbonell and J. Klein-Seetharaman, Prediction of interactions between HIV-1 and human proteins by information integration, Pac. Symp. Biocomput., 2009, 516–527. 88 J. M. Doolittle and S. M. Gomez, Mapping protein interactions between Dengue virus and its human and insect hosts, PLoS Neglected Trop. Dis., 2011, 5, e954. 89 V. Kolodkina, T. Denisevich and L. Titov, Identification of Corynebacterium diphtheriae gene involved in adherence to epithelial cells, Infect. Genet. Evol., 2011, 11, 518–521. 90 B. Li and R. Yang, Interaction between Yersinia pestis and the host immune system, Infect. Immun., 2008, 76, 1804–1811. 91 C. G. Zhang, A. D. Gonzales, M. W. Choi, B. A. Chromy, J. P. Fitch and S. L. McCutchen-Maloney, Subcellular proteomic analysis of host–pathogen interactions using human monocytes exposed to Yersinia pestis and Yersinia pseudotuberculosis, Proteomics, 2005, 5, 1877–1888. 92 Y. Wang, T. Cui, C. Zhang, M. Yang, Y. Huang, W. Li, L. Zhang, C. Gao, Y. He, Y. Li, F. Huang, J. Zeng, C. Huang, Q. Yang, Y. Tian, C. Zhao, H. Chen, H. Zhang and Z. G. He, Global protein–protein interaction network in the human pathogen Mycobacterium tuberculosis H37Rv, J. Proteome Res., 2010, 9, 6665–6677. 93 H. Schmidt and M. Hensel, Pathogenicity islands in bacterial pathogenesis, Clin. Microbiol. Rev., 2004, 17, 14–56. 94 S. Y. Gerdes, M. D. Scholle, J. W. Campbell, G. Balazsi, E. Ravasz, M. D. Daugherty, A. L. Somera, N. C. Kyrpides, I. Anderson, M. S. Gelfand, A. Bhattacharya, V. Kapatral, M. D’Souza, M. V. Baev, Y. Grechkin, F. Mseeh, M. Y. Fonstein, R. Overbeek, A. L. Barabasi, Z. N. Oltvai and A. L. Osterman, Experimental determination and system level analysis of essential genes in Escherichia coli MG1655, J. Bacteriol., 2003, 185, 5673–5684. 95 J. I. Glass, N. Assad-Garcia, N. Alperovich, S. Yooseph, M. R. Lewis, M. Maruf, C. A. Hutchison, III, H. O. Smith and J. C. Venter, Essential genes of a minimal bacterium, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 425–430. 96 C. T. French, P. Lao, A. E. Loraine, B. T. Matthews, H. Yu and K. Dybvig, Large-scale transposon mutagenesis of Mycoplasma pulmonis, Mol. Microbiol., 2008, 69, 67–76. 97 V. K. Morya, V. Dewaker, S. D. Mecarty and R. Singh, In silico Analysis Metabolic Pathways for Identification of Putative Drug Targets for Staphylococcus aureus, J. Comput. Sci. Syst. Biol., 2010, 3, 062–069. 98 P. R. Chen, P. Brugarolas and C. He, Redox signaling in human pathogens, Antioxid. Redox Signaling, 2011, 14, 1107–1118. 99 A. Kumar, A. Farhana, L. Guidry, V. Saini, M. Hondalus and A. J. Steyn, Redox homeostasis in mycobacteria: the key to tuberculosis control?, Expert Rev. Mol. Med., 2011, 13, e39. 100 M. C. Vozenin-Brotons, V. Sivan, N. Gault, C. Renard, C. Geffrotin, S. Delanian, J. L. Lefaix and M. Martin, Antifibrotic action of Cu/Zn SOD is mediated by TGF-beta1 repression and phenotypic reversion of myofibroblasts, Free Radical Biol. Med., 2001, 30, 30–42. This journal is c The Royal Society of Chemistry 2013 Integrative Biology 101 V. Kozjak-Pavlovic, K. Ross and T. Rudel, Import of bacterial pathogenicity factors into mitochondria, Curr. Opin. Microbiol., 2008, 11(1), 9–14. 102 G. R. Cornelis, The type III secretion injectisome, Nat. Rev. Microbiol., 2006, 4(11), 811–25. 103 S. Backert and T. F. Meyer, Type IV secretion systems and their effectors in bacterial pathogenesis, Curr. Opin. Microbiol., 2006, 9(2), 207–217. 104 M. S. R. Couto, C. S. Klein, D. Voss-Rech and H. Terenzi, Extracellular Proteins of Mycoplasma synoviae, ISRN Vet. Sci., 2012, 2012, 6. 105 R. Nair and S. Chanda, Antimicrobial Activity of Terminalia catappa, Manilkara zapota and Piper betel Leaf Extract, Indian J. Pharm. Sci., 2008, 70, 390–393. This journal is c The Royal Society of Chemistry 2013 Paper 106 I. Ali, F. G. Khan, K. A. Suri, B. D. Gupta, N. K. Satti, P. Dutt, F. Afrin, G. N. Qazi and I. A. Khan, In vitro antifungal activity of hydroxychavicol isolated from Piper betle L, Ann. Clin. Microbiol. Antimicrob., 2010, 9, 7. 107 N. Dasgupta and B. De, Antioxidantactivity of PiperbetleL. leafextract in vitro, Food Chem., 2004, 88, 219–224. 108 S. Ganguly, S. Mula, S. Chattopadhyay and M. Chatterjee, An ethanol extract of Piper betle Linn. mediates its anti-inflammatory activity via down-regulation of nitric oxide, J. Pharm. Pharmacol., 2007, 59, 711–718. 109 D. G. Kanjwani, T. P. Marathe, S. V. Chiplunkar and S. S. Sathaye, Evaluation of immunomodulatory activity of methanolic extract of Piper betel, Scand. J. Immunol., 2008, 67, 589–593. Integr. Biol.