UBC Faculty Research and Publications

Modeling of cell signaling pathways in macrophages by semantic networks Hsing, Michael; Bellenson, Joel L; Shankey, Conor; Cherkasov, Artem Oct 19, 2004

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12859_2004_Article_272.pdf [ 1.44MB ]
JSON: 52383-1.0228472.json
JSON-LD: 52383-1.0228472-ld.json
RDF/XML (Pretty): 52383-1.0228472-rdf.xml
RDF/JSON: 52383-1.0228472-rdf.json
Turtle: 52383-1.0228472-turtle.txt
N-Triples: 52383-1.0228472-rdf-ntriples.txt
Original Record: 52383-1.0228472-source.json
Full Text

Full Text

ralssBioMed CentBMC BioinformaticsOpen AcceMethodology articleModeling of cell signaling pathways in macrophages by semantic networksMichael Hsing1, Joel L Bellenson2, Conor Shankey3 and Artem Cherkasov*4Address: 1CIHR/MSFHR Strategic Training Program in Bioinformatics, Genetics Graduate Program, Faculty of Graduate Studies, University of British Columbia, Vancouver, British Columbia, V5Z 3J5, Canada, 2Upstream Biosciences, Inc., Vancouver, British Columbia, V6H 1H2, Canada, 3Visual Knowledge, Inc., Vancouver, British Columbia, V6H 1H2, Canada and 4Division of Infectious Diseases, Department of Medicine, Faculty of Medicine, University of British Columbia, Vancouver, British Columbia, V5Z 3J5, CanadaEmail: Michael Hsing - mhsing@interchange.ubc.ca; Joel L Bellenson - jbellenson@biocad.com; Conor Shankey - cshankey@visualknowledge.com; Artem Cherkasov* - artc@interchange.ubc.ca* Corresponding author    AbstractBackground: Substantial amounts of data on cell signaling, metabolic, gene regulatory and otherbiological pathways have been accumulated in literature and electronic databases. Conventionally,this information is stored in the form of pathway diagrams and can be characterized as highly"compartmental" (i.e. individual pathways are not connected into more general networks). Currentapproaches for representing pathways are limited in their capacity to model molecular interactionsin their spatial and temporal context. Moreover, the critical knowledge of cause-effect relationshipsamong signaling events is not reflected by most conventional approaches for manipulatingpathways.Results: We have applied a semantic network (SN) approach to develop and implement a modelfor cell signaling pathways. The semantic model has mapped biological concepts to a set of semanticagents and relationships, and characterized cell signaling events and their participants in thehierarchical and spatial context. In particular, the available information on the behaviors andinteractions of the PI3K enzyme family has been integrated into the SN environment and a cellsignaling network in human macrophages has been constructed. A SN-application has beendeveloped to manipulate the locations and the states of molecules and to observe their actionsunder different biological scenarios. The approach allowed qualitative simulation of cell signalingevents involving PI3Ks and identified pathways of molecular interactions that led to known cellularresponses as well as other potential responses during bacterial invasions in macrophages.Conclusions: We concluded from our results that the semantic network is an effective methodto model cell signaling pathways. The semantic model allows proper representation and integrationof information on biological structures and their interactions at different levels. The reconstructionof the cell signaling network in the macrophage allowed detailed investigation of connectionsamong various essential molecules and reflected the cause-effect relationships among signalingevents. The simulation demonstrated the dynamics of the semantic network, where a change ofstates on a molecule can alter its function and potentially cause a chain-reaction effect in the system.Published: 19 October 2004BMC Bioinformatics 2004, 5:156 doi:10.1186/1471-2105-5-156Received: 13 April 2004Accepted: 19 October 2004This article is available from: http://www.biomedcentral.com/1471-2105/5/156© 2004 Hsing et al; licensee BioMed Central Ltd. This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Page 1 of 13(page number not for citation purposes)BMC Bioinformatics 2004, 5:156 http://www.biomedcentral.com/1471-2105/5/156BackgroundInteractions among genes, gene products and small mole-cules regulate all cellular processes involving cell survival,cell proliferation, and cell differentiation among others.Such interactions are organized into complex lattice struc-tures conventionally divided into cell signaling, metabolicand gene regulatory networks in a cell [1]. In recent years,large amounts of information and knowledge on cell sig-naling networks have been accumulated in the literatureand databases [2,3].Conventionally, this information is highly compartmen-tal: various individual signaling pathways are mostlystored in separated and non-linked diagrams. Traditionalpathway diagrams, where molecules are represented asnodes and their interactions are depicted as lines andarrows have significant limitations as they lack spatial andtemporal context [4]. Moreover, the critical knowledge ofcause-effect relationships among signaling events is notreflected by most conventional approaches for manipulat-ing pathways. Not surprisingly, the current state of path-way representation does not allow of complexinvestigation of qualitative or quantitative changes in cellsignaling networks in response to external perturbationssuch as bacterial infections. Thus, an adequate computa-tional environment for modeling cell signaling networksis needed for proper biological data integration as well asfor simulation and prediction of cellular behaviors [5].Recently, many models have been proposed for represent-ing, storing and retrieving interactions among various bio-logical entities. BIND [6] and IntAct [7] focus on protein-protein interactions and their resulting complexes. BioCyc[8] developed models for metabolic events and curatedmetabolic pathways from many organisms. The modeldeveloped by aMAZE [9] combines interactions in cell-sig-naling, metabolic and gene regulatory pathways. In addi-tion, the System Biology Markup Language (SBML) hasbeen developed for representing biochemical reactionnetworks and for communicating models used for varioussimulation programs [10]. Programs such as E-cell [11],Gepasi 3 [12] and Virtual Cell [13] use differential equa-tions to represent molecular interactions, and their simu-lation results are obtained by solving these questionsnumerically [14]. It should be noted, however, that manycellular processes are sensitive to the stochastic behaviorof a small number of molecules, and therefore, theassumptions in differential-equation methods can oftenbe compromised [15]. Several studies have attempted toaddress the stochastic property of a cell. Vasudeva andBhalla [16] proposed a hybrid simulation method thatcombined both deterministic and stochastic calculations.In addition, a stochastic simulator, StochSim [15] repre-that useful cell signaling simulators should be capable ofrepresenting each molecule individually and reflecting thestochastic behavior of molecular interactions in a cell.Semantic networksRecently an artificial intelligence approach known assemantic networks (SN) have gained the attention of thebiological community as a potentially powerful tool fororganizing and integrating large amounts of biologicalinformation [17]. For instance, the semantic network inthe Unified Medical Language System (UMLS) wasdesigned to retrieve and integrate biomedical informationfrom various resources [18]. The UMLS semantic networkhas also been applied and expanded to include informa-tion and knowledge from other domains such as genom-ics [19]. In addition, other studies have suggested asemantic approach where proteins are viewed as "adaptiveand logical agents", whose properties and behaviors areaffected by other agents in their spatial organizationincluding intracellular compartments and protein com-plexes [20,21]. Defining the semantics among agentscould characterize both local and global behaviors of asystem, and therefore, it is potentially useful to apply suchapproach to study cell signalling in biological systems[21].A semantic network is a method to represent informationor knowledge by nodes and edges in a graphic form,where a node represents a concept and an edge representsa relationship [22]. A semantic network, which can existabstractly in a human mind or be implemented by apply-ing computer technology, can model many real-worldproblems [22]. Figure 1 illustrates a semantic network,where a concept such as a protein, a chemical reaction ora subcellular location is modeled by a semantic agent, andits relationships with other agents are represented asarrows. A proper semantic network implementationallows the identity and properties of an agent to arise fromits relationships with other agents, not from descriptionsor labels [23]. Hence, within a semantic network "thingsare what they do".Previously an application development environmentknown as Visual Knowledge (VK) has been created, andVK is capable of different formalizations and implementa-tions of semantic networks for various knowledgedomains [23]. Visual Knowledge has been applied suc-cessfully to model and manipulate complex "interac-tomes", including corporate enterprise systems, flightscheduling networks, hardware maintenance simulators,and integrated currency exchange boards [23]. It has beenanticipated that Visual Knowledge can address many ofthe current limitations on modeling cell signaling path-Page 2 of 13(page number not for citation purposes)sented molecules as individual software objects that inter-act according to probabilities. Thus, it is feasible to suggestways. Using the latest VK-based environment, BioCAD[24], specifically designed for biological applications, weBMC Bioinformatics 2004, 5:156 http://www.biomedcentral.com/1471-2105/5/156have developed a semantic model for cell signaling path-ways occurring in human macrophages.Bacterial invasions in macrophagesIt is the current knowledge that many pathogenic bacteriaare capable of entering and surviving within mammalianmacrophages by modulating the host signaling pathways[25]. One well-studied example is the activation of the Fcγmacrophage receptor by the IgG antibody, which binds tothe surface of bacteria such as Mycobacterium tuberculo-sis [26]. Activation of the Fcγ receptor induces phagocyto-sis of M. tuberculosis and the formation of a phagosomewithin the macrophage. These processes are mediated bythe class I phosphoinositide 3-kinase (PI3K) – one of themost well-characterized enzymes to date [27]. The class IPI3K is a heterodimer composed a p110 catalytic subunitand a p85 regulatory subunit, which maintains a low-levelactivity of p110 [28]. The p110 subunit is activated whenp85 binds at a phosphotyrosine site on a receptor or anadaptor protein, or by direct binding to activated Ras [29].Activated PI3K-p110 phosphorylates phosphatidylinosi-tol-4,5-bisphosphate (PIP2) into phospatidylinositol-3,4,5-trisphosphate (PIP3), which is an essential signalingmolecule that stimulates many downstream proteins,including PDK1 and Akt [30]. The formation of a phago-some is normally followed by the phagosome maturationever, it has been hypothesized that phosphatidylinositolanalogs, such as ManLAM, produced by M. tuberculosiscan inhibit the activity of the class III PI3K, arrestingphagosome maturation process, and ensuring the survivalof M. tuberculosis inside the macrophage [27,32].In addition to their role in phagocytosis, PI3Ks are essen-tial proteins that regulate cell survival, cell growth, cellcycle and other cellular processes [33]. Although, it is clearthat PI3Ks play an important role in bacterial invasions,the knowledge of PI3K-mediated interactions is scatteredin a number of literature and pathway databases. A coher-ent picture of detailed molecular interactions that linkreceptors to PI3Ks and to various cellular responses hasyet to be constructed before bacterial invasions can befully understood. To address this goal, a cell signaling net-work of the human macrophage was reconstructed withthe semantic model, and qualitative changes in the net-work were investigated with a SN-simulator.ResultsA semantic model for cell signaling pathwaysIn the paper, the word "model" refers to a set of rules intwo different but related contexts. In the context of thesemantic network, the model refers to a set of rules thatspecify how a biological concept is mapped to one or mul-tiple semantic agents/relationships. In the context of cellsignaling pathways, the model is a set of rules that specifywhat, how, and when molecules interact with each other.The model has been formalized and implemented, usingBioCAD software system, and it is presented in the follow-ing sections.Overall classification of biological structures and their relationshipsWithin the semantic network, all biological structures aremodeled by semantic agents that are members in one ofthe 6 different prototypes. Table 1 shows the 6 types ofstructures in the order of their hierarchy. From the highestto the lowest level, they are "Cell", "Subcellular Compart-ment", "Macromolecule", "Domain/Site", "Small Mole-cule/Molecular Fragment", and "Atom". A structure agentcan be composed of multiple structures of the same pro-totype or a lower-level prototype, and the agent is con-nected to its components by the composition relationshipin the SN. Thus, a human macrophage has been modeledas a semantic agent of the "Cell" prototype, and it wascomposed of various "Subcellular Compartment" agents,including plasma membrane, cytosol, nucleus and others.In addition, each compartment such as nucleus containedvarious agents of the "Macromolecule" prototype includ-ing proteins, DNA and RNA. A macromolecule such as aprotein was further composed of "Domain/Site" agentslike catalytic domains and phosphorylation sites, and aAn example of a semantic networkFigure 1An example of a semantic network. Characteristics and behaviors of a semantic agent (SA) are defined by its relation-ships (RE) with other agents. Semantic agents are repre-sented as circles, and relationships are depicted as arrows. This SN-model represents that a protein A can be located at a nucleus, can interact with a protein B or catalyze a chemical reaction. For explanatory purpose, this figure illustrates an example of a semantic network. The implemented semantic network (as presented in the paper) is more complex and involves different types of relationships and agents.Page 3 of 13(page number not for citation purposes)process, which is responsible for intracellular killing ofbacteria and is regulated by the class III PI3K [31]. How-DNA was composed of sites such as promoters and generegulatory sites.BMC Bioinformatics 2004, 5:156 http://www.biomedcentral.com/1471-2105/5/156Modeling interactions among biological structuresTo create an adequate semantic model, we have postu-lated that structures of different levels in the cellular hier-archy can interact with one another. One example of suchinteractions is the movement of a molecule from one sub-plasma membrane). This is referred to as a translocationevent, and it is demonstrated on the left panel of Figure 2.Table 2 shows that translocations have been modeled asone the five major "event" prototypes in the SN. Everytranslocation event has been connected to three structureagents: a molecule to be moved (macromolecule or smallmolecule), an original location (subcellular compart-ment), and a destination (subcellular compartment).Hence, the construction of translocation events has ena-bled us to confine all possible movements of molecules ina cell.Interactions that occur by non-covalent or covalent forceshave also been modeled as two distinct "event" proto-types as shown in Table 2. The right panel of Figure 2 illus-trates a general case of a molecular interaction between aprotein A and a protein B occurring via non-covalentforces. Such interaction can cause changes of the formsand functions of the interacting molecules, and thesechanges have been accommodated within the developedSN model by specifying two distinct types of states: "con-formational states" and "binding states", also representedby semantic agents.All hypothetical spatial changes occurring in the three-dimensional structure of a given macromolecule havebeen modeled within the SN as switches in the corre-sponding conformational states, and the changes do notTable 1: Classification of biological structures in 6 prototypes in the semantic network.Semantic Agent – Structure Biological ExampleCell Human macrophage, Mycobacterium tuberculosisSubcellular Compartment Plasma membrane, cytosol, phagosome, nucleusMacromolecule Protein, nucleic acid, polysaccharide, fat/lipidDomain and Site Catalytic domain, SH2 domain, PH domain, binding site, phosphorylation site, promoter, gene regulatory site.Small Molecule and Molecular Fragment Amino acid, nucleotide, sugar, fatty acidAtom Hydrogen, carbon, oxygen, nitrogen, phosphorus, sulfurTable 2: Classification of biological events in 5 prototypes in the semantic network.Semantic Agent – Event Biological ExampleTranslocation A protein moves from cytosol to plasma membrane.Non-covalent Interaction A ligand binds to a receptor.Covalent Interaction An enzyme catalyzes a chemical reaction where substrates are converted to products.Allosteric Regulation A ligand binding on site A of a protein causes a conformational change on site B of the protein.Cellular Response Cell survival, cell death, phagosome formation, increase of intracellular glucose level.Interactions among biological structures of different levels in the SNFigure 2Interactions among biological structures of different levels in the SN. The left panel shows an example of a translocation event when a protein B is moved from the cytosol to the plasma membrane. The right panel shows an example of a non-covalent interaction between a protein A and a protein B via non-covalent forces.Page 4 of 13(page number not for citation purposes)cellular compartment (e.g. cytosol) to another (e.g. lead to the creation of new semantic agents. Domains orsites for every protein encoded into the SN model haveBMC Bioinformatics 2004, 5:156 http://www.biomedcentral.com/1471-2105/5/156been assigned to either "Functional" or "Non-functional"conformational states. The "Functional" state representsthat a domain/site is currently in a conformation that ena-bles a certain interaction. On the other hand, a "Non-functional" state implies a domain/site is in a conforma-tion that prevents an interaction. To illustrate thisconstruct we have graphed the semantic agents and theirrelationships created within the developed SN. It shouldbe noted that within the SN, all semantic agents are visu-alized as icons, and their relationships are depicted asconnecting arrows. In addition, all agents are related bypairs of reciprocal relationships, and for simplicity, onlyone direction of each pair of the relationships was visual-ized. The left panel of Figure 3 features a p110 subunit ofthe class I PI3K that has been modeled as a "macromole-cule" agent and contains a binging site for a Ras proteinand a catalytic domain. The Ras binding site has beenassigned a state of "Functional", depicted as a check sym-hand, the catalytic domain is "Non-functional', depictedas a cross symbol. Figure 8 shows the description of iconsused in this paper.In addition to the conformational states, a proteindomain or site has been assigned one of the two bindingstates: "Bound" or "Not-bound". A "Bound" state impliesthat this domain/site currently associates with a domain/site of another molecule through a non-covalent interac-tion. On the other hand, a "Not Bound" state indicatessuch an association does not exist. Since ligand bindingscan affect the conformation of a macromolecule throughallosteric regulations, two types of such regulations havebeen implemented within the SN. A positive allosteric reg-ulation event has been assigned to the scenario when a"Bound" binding state of a domain/site causes the confor-mational state of another domain/site to switch to "Func-tional". The right panel of Figure 3 shows that when thePI3K-p110 has bound to a Ras by a non-covalent interac-tion, the binding state of the Ras-binding site on p110 hasswitched to "Bound". As a result, the conformational stateof the catalytic domain has switched to "Functional" dueto a positive allosteric regulation. The "Functional" cata-lytic domain now enables the PI3K-p110 to phosphor-ylate its substrate. On the other hand, a negative allostericregulation event has been attributed to those cases whena "Bound" state of a domain/site causes the conforma-tional state of another domain/site to switch to "Non-functional". It should be noted that the semantic modelstores the information that specifies the non-covalentevent between the prototypic Ras and the prototypic PI3K-p110, and the condition for the event to occur. Figure 3illustrates an instance of the Ras-binding event occurredduring a simulation. The PI3K-p110 is an instance of thePI3K-p110 prototype, and it is the same agent before andafter it binds to the Ras.A more complex allosteric regulation event can be speci-fied for mapping the binding states of multiple domains/sites (the condition or the input) to the conformationalstates of multiple domains/sites (the response or the out-put). Hence, a domain is switched to "functional" only ifa certain combination of ligand bindings has occurred.The utilization of the states on domains/sites and allos-teric regulation events in the SN has enabled us to expressthe cause-effect relationships among various signalingevents explicitly.In the developed semantic model, any molecular complexformed due to non-covalent interaction has been treatedas a transient state of these two molecules, and a complexwas not represented by a new semantic agent. Instead, theexistence of a protein complex is inferred from the non-A model of a non-covalent interaction between a PI3K-p110 and a RasFigure 3A model of a non-covalent interaction between a PI3K-p110 and a Ras. The figure was graphed from the developed SN to illustrate the relationships among different agents. The figure visualizes the agents as icons and their relationship as arrows. The left panel illustrates that a PI3K-p110 contains a "Not Bound" Ras-binding site and a "Non-Functional" catalytic domain. The right panel shows that when the PI3K-p110 has bound to a Ras, its Ras-binding site has switched to "Bound", and the catalytic domain has become "Functional" due to a positive allosteric regulation event. State changes as a result of the interaction are shown in bold. Note that the model stores the information, which specifies the non-covalent event between the prototypic Ras and the prototypic PI3K-p110, and the condition for the event to occur. This figure illustrates an instance of the Ras-binding event occurred dur-ing a simulation. The PI3K-p110 is an instance of the PI3K-p110 prototype, and it is the same agent before and after it binds to the Ras. Figure 8 shows the description of each icon.Page 5 of 13(page number not for citation purposes)bol (square) on Figure 3. The "Functional" state enablesthe PI3K-p110 to bind to a Ras protein. On the othercovalent interaction event. Thus, Figure 3 illustrates a pro-tein complex of the PI3K-p110 and the Ras existedBMC Bioinformatics 2004, 5:156 http://www.biomedcentral.com/1471-2105/5/156because of the occurrence of the non-covalent interactionevent, which connected the two molecules.Conventionally, there is often some inconsistencybetween representing chemical modifications of smallmolecules in metabolic pathways and modifications ofproteins in signaling pathways. In the developed model,two molecules that interacted by covalent forces haveresulted the creation of distinct semantic agents within theSN. This rule has been implemented consistently for bothmacromolecules such as proteins and small moleculessuch as ATP. As one example, Figure 4 features the phos-phorylation of an Akt protein by an enzyme PDK1,yielding a distinct Akt-phosphate (Akt-P) agent and a freedepicted as arrows. In addition, the Akt-P and the ADP arerelated to the event by "Product" relationships, while thePDK1 is related by the "Enzyme" relationship. The PDK1enzyme in this example contains a catalytic domain (notshown on the figure), which must be "functional" for thereaction to occur. The state of this domain is under theregulation of the binding of a ligand and an allostericevent as previously defined. In addition, new propertiescan be assigned to the modified protein. In this case, thephosphorylation by PDK1 switched the catalytic domainin Akt-P to "functional", while this domain was "non-functional" in Akt, the dephosphorylated form. Figure 4illustrates that a covalent interaction event also applies tometabolites, and a metabolite such as glucose is phospho-rylated into a glucose-6-phosphate by an enzymeHexokinase. Other types of modifications includingmethylation, acetylation and glycosylation can also bemodeled in a similar manner but involve different sub-strate types.In the semantic model, a molecule can participate in dif-ferent sets of interactions in different locations. The trans-location events define all possible localizations ofmolecules, and therefore, an interaction can only occur ifthe participating molecules can be present in the samelocation at the same time. Alternatively, an interaction(non-covalent or covalent) can directly associate with asubcellular compartment, and this interaction is onlyavailable to molecules in that location.In addition, all qualitative cellular responses such as cellsurvival and phagosome formation have been imple-mented within the SN under a distinct "event" prototype.They have been implemented in a way that the formationor the activation of certain signaling molecules such asPIP3 can trigger the occurrence of these cellular responseevents in a simulation.As it has been mentioned previously, the behavior of anysemantic agent can be clearly defined by its relationshipsor connections to other agents. Thus, the formalization ofthe five types of events, which are translocations, non-cov-alent interactions, covalent interactions, allosteric regula-tions and cellular responses, has enabled us to model thebehaviors of molecules depicted in conventional path-ways and to reconstruct a cell signaling network of thehuman macrophage.Case study: a reconstruction of a cell signaling network in the macrophageData sourceThe molecular composition of human macrophages andinformation of known intracellular interactions haveA model for covalent interactionsFigure 4A model for covalent interactions. Figure 4a shows that an Akt protein can be phosphorylated to an Akt-phosphate by an enzyme, PDK1, and an ATP is converted to an ADP in the process. Figure 4b shows a similar covalent interaction event where substrate Glucose can be converted to Glucose-6-phosphate by an enzyme Hexokinase.Page 6 of 13(page number not for citation purposes)ADP. Within the SN, the Akt and the ATP are related to acovalent interaction event by "Substrate" relationships,been extracted from various research articlesBMC Bioinformatics 2004, 5:156 http://www.biomedcentral.com/1471-2105/5/156[26,27,32,34-47], review articles [25,28-31,33,48-52] andpathway databases [2,3].Translation and integration of pathway information into the semantic modelA pathway diagram in the literature or an electronic data-base, in principle, represents some scenario of what mayhappen in a call if every depicted molecule is expressed inthe correct location, at the correct time and with the cor-rect states. Hence, the aggregation of multiple pathwaydiagrams describes some, if not all, possible molecularevents that can potentially occur in a cell under the rightconditions. To utilize such information and build a cellsignaling network, we have decomposed conventionalpathways into individual pieces of information such assubcellular localization of a protein, a pairwise proteinbinding, a chemical reaction or a cellular response. Then,using the sets of semantic rules described in the model, wehave represented and integrated each piece of those infor-mation in the form of semantic agents and their relation-ships. Table 3 illustrates the overall SN model for the cellsignaling network contained a total of 93 prototypicalmacromolecules localized in several subcellular compart-ments. It included several cell receptors (such as Fcγ, CR3,CD 14, CD18, TLR2) relevant to the process of bacterialinternalization of macrophages. Two distinct classes ofPI3Ks have been modeled: the class I PI3K composed ofp85 regulatory and p110 catalytic subunits, and the classIII PI3K composed of p150 and Vps34p subunits [28]. Themodel also included various kinases such as Lyn, PDK1and Akt, small GTPases including Ras, Rac1 and Rab5,and adaptor proteins like Gab2. Events of various proto-types have also been extrapolated from the literature andpathway diagrams.Visualization and analysis of the cell signaling networkThe defined semantic agents have been connected in theand covalent interactions have been integrated into aunified cell signaling network. The longest path in the cellsignaling network we have created contained 24 consecu-tive molecular interaction events, linking Fcγ receptor tothe class I PI3K enzyme and further through class III PI3Kto various cellular responses.Such detailed semantic reconstruction of the cell signalingnetwork has allowed thorough investigation of biochem-ical relationships between essential proteins. One suchexample is presented on Figure 5 featuring the connec-tions among cell receptors Fcγ and CR3, and tyrosinekinase Lyn which they both activate. It has also beenreconstructed by the SN model that both of thesereceptors can activate the class I PI3K via an adaptor pro-tein, Gab2. The corresponding finding will now be sub-jected to testing in an experimental lab.Another example of successful SN reconstruction is therelationship between CD14 macrophage receptor and theclass I PI3K; such a relationship was previously suspectedbut not clear [39]. By incorporating the available literaturedata [35,45] into the semantic environment we were ableto reconstruct the scenario where CD14 activates the classI PI3K by the association of Toll-like receptor 2 (TLR2), asit is illustrated in Figure 5. Such model will also be testedexperimentally.Simulation of changes in the cell signaling network during bacterial invasionsIn the implemented semantic model, the "possible"behaviors of a molecule are defined through its relation-ships to other agents (for example a non-covalent event),and all instances of that prototypical molecule will inheritthe same behaviors. However, the action of a molecule atany given time is affected by factors including its currentstates and its current location with respect to otherTable 3: The number of structure and event prototypes modeled in the cell signaling network of the macrophage.Structure SumSubcellular Compartment 9Macromolecule 93Domains and Site 20Small Molecule 9Event SumTranslocation 24Non-covalent Interaction 57Covalent Interaction 28Cellular response 22Page 7 of 13(page number not for citation purposes)semantic network and can be visualized at different levels.Figure 5 shows one example of how various non-covalentmolecules in the system. Hence, we have built an applica-tion that enabled us to produce instances of molecules inBMC Bioinformatics 2004, 5:156 http://www.biomedcentral.com/1471-2105/5/156various locations and to observe the "current" action of amolecule qualitatively under different biological scenar-ios. We refer such scenario-play as simulation in thispaper.The application or the SN-simulator allows the moleculesto move among various locations, to interact with eachother and to create events when the conditions are met. Inaddition, every instance of a molecule has beenrepresented as an individual agent while every instance ofa molecular interaction has also been implemented as anindividual event agent. Thus, the simulator provides atraceable "trajectory" of all the events that have happenedon every molecule during a simulation.As illustrated in Figure 6, the macrophage cell has beengenerally divided into four subcellular compartments ortion in the beginning of a simulation, and the simulatorsynthesized molecules in each location accordingly. At thevery first simulation step, the simulator has created atranslocation event moving a molecule (the current tar-get) from one location to another. The initial transloca-tion has been specified as the movement of an IgGmolecule from the extracellular space to the plasma mem-brane as shown in the pathway-viewer on Figure 6. Theoccurrence of this initial event allowed the simulator totrigger a search and advanced to the next step. The searchlooked for other potential molecules (with the correctstates) that can interact with the target molecule in thesame location. If multiple instances of potentially inter-acting molecules were present in that location, a singlemolecule would be randomly selected to interact with thetarget.Because an Fcγ receptor was the only interacting molecule(for the IgG) present at plasma membrane in the simula-tion, it has bound to the IgG by a non-covalent interactionevent, as illustrated in Figure 7. This non-covalent interac-tion has switched the state of the Fcγ receptor's bindingsite for a Lyn kinase to "Functional", and thus it enabledthe Fcγ receptor to bind to a Lyn. However, the Lyn wasnot initially present in plasma membrane, but it waslocalized in cytosol in the beginning of the simulation, asshown in Figure 6. Thus, when the Lyn has been translo-cated from the cytosol to the plasma membrane, a non-covalent interaction between the Lyn and the "Func-tional" Fcγ receptor occurred in the following step asshown in Figure 7. The search was iterated and the simu-lation continued until all interacting molecules have beendepleted in the macrophage.Figure 7 demonstrates the consecutive events in this sim-ulation scenario where the Lyn protein phosphorylated aGab2, which then bound to a class I PI3K. When acti-vated, the PI3K phosphorylated a PIP2 into a PIP3, whichin turn caused a phagosome formation response. Differ-ent setups of the initial localization of molecules haveaffected the outcome of the simulation. For instance, aninitial presence of a Rab5 (a downstream protein of thePIP3) and a class III PI3K in the cytosol extended the pre-vious pathway from the PIP3. This localization setupstimulated a PIP3-mediated activation of the class IIIPI3K, which led to phagosome maturation response in thesimulation. However, if a phosphatidylinositol analog,ManLAM, of M. tuberculosis was initially present in theplasma membrane, it would inhibit the class III PI3K andthus arrest the phagosome maturation response in themacrophage. Table 4 shows that the activation of PI3Ks-mediated pathways by M. tuberculosis has caused severalknown cellular responses as well as additional responsesPhagocytosis of bacteria in macrophagesFigure 5Figure 5a- Phagocytosis of bacteria in macrophages. The pic-ture shows macrophages ingesting green fluorescent myco-bacteria (indicated by arrows). The host cell membrane was stained by red fluorochorme PKH to define the limit of the cell. (The picture was provided by Zakaria Hmama)Figure 5b- A SN-representation of the cell signaling network that regulates phagocytosis in the human macrophage. Both molecules and their interactions (non-covalent and covalent interactions) are represented as semantic agents and visual-ized as nodes (with distinct icons) on the diagram. Arrows represent the semantic relationships between different agents.Page 8 of 13(page number not for citation purposes)locations within the simulator. We have specified whatmolecules to be present initially in each subcellular loca-such as cell survival of the macrophage, cell cycle entry,increase of protein synthesis and increase of intracellularBMC Bioinformatics 2004, 5:156 http://www.biomedcentral.com/1471-2105/5/156A SN- simulator: at the beginning of the simulationFigure 6A SN- simulator: at the beginning of the simulation. The sim-ulation showed the actions of molecules under a biological scenario. 1. The initializing buttons synthesize molecules in each subcellular compartment. 2. The localization window shows molecules present in each subcellular compartment. In this simulation, an IgG molecule was present at the extracel-lular space (E.S.). There were 2 ATP molecules, an Fcγ recep-tor (FcγR), a Gab2 and a PIP2 (PI[4,5]P2) present at the plasma membrane (P.M.). The cytosol contained a Lyn kinase, a PI3K-p85 and a PI3K-p110 subunit. There was no molecule present at the nucleus in this simulation. 3. The "Start Simula-tion" button creates a previously specified translocation event. In this simulation, the translocation has already occurred and moved the IgG from the extracellular space to the plasma membrane. 4. The "Next" button triggers a search that determines a proper event to occur and advances to the next step. 5. The pathway-viewer shows a series of events occurred during the simulation.A SN- simulator: at the end of the simulationFigure 7A SN- simulator: at the end of the simulation. The pathway-viewer shows that the initial translocation of the IgG mole-cule has led to the occurrence of a series of events, which include several non-covalent interactions, covalent interac-tions, and translocations of various molecules: Event #1: the IgG was translocated from the extracellular space to the plasma membrane. Event #2: the IgG bound to the Fcγ recep-tor at the plasma membrane. Event #3: the Lyn was translo-cated from the cytosol to the plasma membrane. Event #4: the Lyn bound to the Fcγ receptor at the plasma membrane. Event #5: the Lyn phosphorylated the Gab2 to a Gab2-phos-phate (Gab2-P) at the plasma membrane. Event #6: the PI3K-p85 and p110 (already bound to each other) were translo-cated together from the cytosol to the plasma membrane. Event #7: the PI3K-p85 bound to the Gab2-P at the plasma membrane. Event #8: the PI3K-p110 phosphorylated the PIP2 to a PIP3 (PI[3,4,5]P3) at the plasma membrane. Event #9: The formation of the PIP3 caused phagosome formation.Table 4: Simulation results from M. tuberculosis invasion in the human macrophage.Known response in the macrophage Other potential response in the macrophageActin Polymerization Cell survivalPseudopod Extension Cell cycle entry – S phasePhagosome Formation Increase of protein synthesisPhagosome Maturation Arrest Increase of intracellular glucose levelPage 9 of 13(page number not for citation purposes)BMC Bioinformatics 2004, 5:156 http://www.biomedcentral.com/1471-2105/5/156DiscussionFeatures of the semantic modelIn the present work we have developed a semantic modelto represent the properties and the behaviors of moleculesand their interactions in the context of cell signaling path-ways. The proposed model offers some additional fea-tures, compared to other existing pathway models. Thosefeatures are essential for characterizing the complexbehaviors of biological entities, and they include:Specify the spatial organization of moleculesThe semantic model has specified the hierarchical rela-tionships among the different biological structures, fromcells to compartments, molecules and domains/sites. Thehierarchy between subcellular compartments andmolecules has allowed us to specify the spatial organiza-represent the effects of locations on the different interac-tions among molecules.Model proteins as integrating and logical devicesThe hierarchy between molecules and their domains/siteshas enabled us to explicitly model the relationshipbetween forms and functions for proteins. Through theallosteric regulation events, proteins have been modeledand implemented as integrating and logical devices in thesemantic network, and their conformational states (out-puts) are switched by the combination of non-covalentligand bindings or covalent modifications (inputs).Provide a direct communication from models to simulationsThrough the prototyping system in the semantic network,any rule or interaction specified on a prototypical mole-cule automatically define the properties and behaviors ofall its instances. As demonstrated by the simulator, thesemantic network provided a direct communication fromthe interaction model to an application where the actionsof molecules can be observed under different scenarios.Therefore, the semantic network is dynamic as a change ofstates on a molecule can alter its function and potentiallycause a chain-reaction effect in the system.Reduce the need for labelsIn addition, the current semantic model is different fromthe previous models in BioCAD. An essential difference isthe representation of functional labels or roles on pro-teins. The meanings of functional descriptions or associa-tion words such as "enzyme", "activator/activates" or"inhibitor/inhibits", which are often used to characterizethe behaviors of proteins in most pathway models, havebeen represented explicitly through events and relation-ships in the developed semantic network. For example, aprotein acts as an "enzyme" if 1) the protein participatesin a "covalent interaction event", 2) the presence of a"functional" catalytic domain on the protein is requiredfor the occurrence of the event, and 3) the protein itself isnot modified after the event. Similarly, a protein A "acti-vates" a protein B if a non-covalent binding event fromprotein A turns on the "functional" state of a domain/siteon protein B. Hence, the model has reduced the need forlabels, which are often confusing or misleading on con-ventional pathway representation.Future directionsThe use of non-covalent and covalent events has enabledus to model protein-protein interactions and chemicalmodifications on molecules including proteins andmetabolites. The next challenge is to model the complexinteractions that govern gene regulations. The current con-struction of non-covalent interaction events can modelDescription of icons used in other figuresFigu e 8Description of icons used in other figures.Page 10 of 13(page number not for citation purposes)tion of molecules, model the translocation events and the binding of an individual transcription factor to a par-ticular site of a gene, and the covalent interaction eventBMC Bioinformatics 2004, 5:156 http://www.biomedcentral.com/1471-2105/5/156can represent the transcription process that leads to theproduction of an mRNA, and the translation process thatproduces a protein. However, a successful transcription ina eukaryotic cell requires the formation of a protein com-plex that involves more than one hundred subunits, andthe complex may be assembled in various orders [53]. Weanticipate the improvement of the current allostericregulation model to characterize the more complex logicin gene regulation.The semantic network representation can be exploited forperforming analysis of cell signaling pathways. The exam-ples of Fcγ receptor, CR and the class I PI3K demonstratedthat connections can be queried and analyzed among dif-ferent biological entities. The semantic model is also com-patible with other pathway models. Therefore, thenumber of biological entities and interactions in thesemantic network can be greatly increased as pathwaydata from existing databases is integrated. Previous studyhas shown the value of combining gene expression pro-files with protein-protein interaction networks for identi-fying active subnetworks [54]. Similarly, data from geneand protein expression experiments could be integratedwith the semantic network for "pathway filtering". Forinstance, within a particular cell, there could be multiplepaths that connect two proteins, while each path consistsof different number of nodes. When the cell receives a sig-nal, the shortest path, the one with the least number ofnodes that require activation, is more likely to be"walked" than a longer path. Hence, the gene/proteinexpression data will provide some estimation of an overallprotein expression and activation states to identify"active" pathways in a cell under a given conditionIn this study, the proposed semantic model has beenapplied to cell signaling pathways in the macrophage as acase study. The model is not limited to those pathways.The hierarchical classification of the biological structuresand the events can model other cell signaling pathways fordifferent cells and organisms. An interactive website iscurrently under development. We anticipate that throughthe web, researchers can utilize the semantic networkapproach for creating pathways in cells of their interestand for analyzing any existing pathways including thePI3K pathways of the human macrophage presented inthe paper.The current capability and applicability of the SN simulatorIn this study, we have developed a simple simulator todemonstrate the dynamics of the semantic network and toobserve the actions of molecules qualitatively. In order toperform a realistic cellular simulation in the future, threebinding affinity, directly associated with non-covalentevents, will affect the probability and the duration of thebinding of molecules. Reaction kinetics, associated withcovalent events, will determine the rate of production.Second, the two parameters, the population of moleculesand their localization, which influence the simulationoutcome, could be initialized and supported by experi-mental data. For instance, gene expression data frommicroarrays supports the relative abundance of tran-scripts, and protein expression data supports the relativeabundance of proteins. Computer algorithms such asPSORT [55] can assist in predicting the localization ofproteins.Third, the proximity of molecules has been represented bysubcellular compartments in the simulation. Thisapproximation can be improved in two different ways.First, a compartment can be further divided into smallersub-locations. Increasing the number of locations andreducing the size of each location will improve the accu-racy of the simulation. Second, the occurrence of non-cov-alent events in the simulation has allowed us to identifymolecular complexes and their members effectively.Hence, the proximity can be approximated throughmolecular complexes, such that molecules in a complexhave higher probability to interact with members of thesame complex.The simulator has demonstrated that a biological pathwaycan emerge from the creation of semantic agents and theirrelationships in the SN, and such a pathway represents aseries of consecutive events resulting from the activationof a single molecule. It is anticipated that further develop-ment will improve our ability to track and visualize differ-ent instances of molecules participated in multiplepathways. Hence, the occurrence of a cellular responseevent can be triggered by the accumulation of certainmolecular species with particular states.ConclusionsWe concluded from our results that the semantic networkis an effective method to model cell signaling pathways.Utilizing the semantic agents and the relationships in themodel, information on biological structures and theirinteractions at different levels has been properly repre-sented and integrated in the hierarchical and spatialcontext. The reconstruction of the cell signaling networkin the macrophage has allowed qualitative investigationof connections among various essential molecules andreflected the cause-effect relationships among the events.The simulation demonstrated the dynamics of the seman-tic network, where actions of molecules are affected byPage 11 of 13(page number not for citation purposes)components need to be improved. First, quantitative fac-tors should be integrated into the model. For exampletheir current states and locations, and the history of eventscan be traced and analyzed. In addition, changes causedBMC Bioinformatics 2004, 5:156 http://www.biomedcentral.com/1471-2105/5/156by the invading M. tuberculosis in the macrophage wereinvestigated by the simulator. As a result, the simulationidentified pathways of molecular interactions that led toknown cellular responses as well as other potentialresponses during bacterial invasions.MethodsThe Visual Knowledge environmentVisual Knowledge (VK) is an application developmentenvironment, and its implementation has been influ-enced by the theory of semantic networks as well as otherapproaches including set theory, frame system, object-ori-ented modeling theory and systems based on networks ofactive software agents [23]. Different from other passiveknowledge representation technology, VK is dynamic andscalable, and it is capable of active representation andintegration of different domain knowledge. Bymanipulating a number of fundamental classes of seman-tic agents like "physical thing", "event" and "trigger",models of various complexity can be constructed with VK.In addition, VK allows the creation of "prototypes" withineach basic class of agents, and therefore it enables anyclassification of agents based on their common character-istics and behaviors.The BioCAD softwareBioCAD, a Visual Knowledge-based development envi-ronment, is developed by Upstream Biosciences, Inc. andcustomized to model biological systems [24]. TheBioCAD software provides tools for managing large-scalebiological data and for visualizing and editing biologicalpathways and networks. BioCAD currently contains mil-lions of biological concepts and hundreds of pathwaysthat have been integrated and curated from publicly avail-able data sources. A locally installed client program allowssemantic agents to be created, stored and queried from aremote central server. The BioCAD software is availablecommercially, and a collaborative modeling server will bepublicly accessible soon.Authors' contributionsThe semantic model was developed jointly by all authorsand implemented by MH, JLB and CS. MH implementedthe simulation, collected and analyzed data, constructedpathways in the macrophage and drafted the manuscript.JLB, CS, AC developed general concepts, provided scien-tific support, participated in the manuscript writing andcoordinated the study. All authors read and approved thefinal version of the manuscript.AcknowledgmentsAuthors acknowledge Zakaria Hmama, Neil E. Reiner and Jimmy Lee (Divi-sion of Infectious Diseases, Department of Medicine, Faculty of Medicine, University of British Columbia) for their knowledge and advice on the bac-opment and implementation, and Ian Upright and Jason Ng (Visual Knowl-edge, Inc.) for the graphical interface and technical support.This research was funded by the CIHR/MSFHR Strategic Training Program in Bioinformatics, sponsored by Canadian Institutes of Health Research and Michael Smith Foundation for Health Research.References1. Ideker T, Lauffenburger D: Building with a scaffold: emergingstrategies for high- to low-level cellular modeling. TrendsBiotechnol 2003, 21:255-262.2. BioCarta  [http://www.biocarta.com]3. Signal Transduction Knowledge Environment (STKE)  [http://stke.sciencemag.org/]4. Kitano H: The standard graphical notation for biochemicalnetworks. ICSB-2002 workshop on SBML/SBW (Stockholm) 2002.5. Demir E, Babur O, Dogrusoz U, Gursoy A, Nisanci G., Cetin-AtalayR, Ozturk M: Patika: an integrated visual environment for col-laborative construction and analysis of cellular pathways. Bio-informatics 2002, 18:996-1003.6. BIND  [http://www.blueprint.org/bind/bind.php]7. IntAct  [http://www.ebi.ac.uk/intact/index.html]8. BioCyc  [http://www.biocyc.org/]9. aMAZE  [http://www.amaze.ulb.ac.be/]10. Hucka M, Finney A, Sauro HM, Bolouri H, Doyle JC, Kitano H, ArkinAP, Bornstein BJ, Bray D, Cornish-Bowden A, Cuellar AA, Dronov S,Gilles ED, Ginkel M, Gor V, Goryanin II, Hedley WJ, Hodgman TC,Hofmeyr JH, Hunter PJ, Juty NS, Kasberger JL, Kremling A, KummerU, Le Novere N, Loew LM, Lucio D, Mendes P, Minch E, MjolsnessED, Nakayama Y, Nelson MR, Nielsen PF, Sakurada T, Schaff JC, Sha-piro BE, Shimizu TS, Spence HD, Stelling J, Takahashi K, Tomita M,Wagner J, Wang J: The systems biology markup language(SBML): a medium for representation and exchange of bio-chemical network models. Bioinformatics 2003, 19:524-31.11. Tomita M, Hashimoto K, Takahashi K, Shimizu TS, Matsuzaki Y, Miy-oshi F, Saito K, Tanida S, Yugi K, Venter JC, Hutchison CA 3rd: E-CELL: software environment for whole-cell simulation. Bioin-formatics 1999, 15:72-84.12. Mendes P: Biochemistry by numbers: simulation of biochemi-cal pathways with Gepasi 3. Trends Biochem Sci 1997, 22:361-363.13. Loew LM, Schaff JC: The Virtual Cell: a software environmentfor computational cell biology. Trends Biotechnol 2001,19:401-406.14. Neves SR, Lyengar R: Modeling of signaling networks. BioEssays2002, 24:1110-1117.15. Le Novere N, Shimizu TS: StochSim: modeling of stochastic bio-molecular processes. Bioinformatics 2001, 17:575-576.16. Vasudeva K, Bhalla US: Adaptive stochastic-deterministicchemical kinetic simulations. Bioinformatics 2004, 20:78-84.17. McCray AT, Nelson SJ: The representation of meaning in theUMLS. Methods Inf Med 1995, 34:193-201.18. Lindberg DA, Humphrey BL, McCray AT: The Unified MedicalLanguage System. Methods Inf Med 1993, 32:281-291.19. Yu H, Friedman C, Rhzetsky A, Kra P: Representing genomicknowledge in the UMLS semantic network. Proc AMIA Symp1999:181-5.20. Fisher MJ, Paton RC, Matsuno K: Intracellular signalling proteinsas smart' agents in parallel distributed processes. Biosystems1999, 50:159-71.21. Fisher MJ, Malcolm G, Paton RC: Spatio-logical processes inintracellular signalling. Biosystems 2000, 55:83-92.22. Griffith RL: Three principles of representation for semanticnetworks. ACM Transactions on Database Systems 1982, 7:417-442.23. Visual Knowledge  [http://www.visualknowledge.com]24. BioCAD  [http://www.biocad.com]25. Tjelle TE, Lovdal T, Berg T: Phagosome dynamics and function.BioEssays 2000, 22:255-263.26. Gu H, Botelho RJ, Yu M, Grinstein S, Neel BG: Critical role forscaffolding adapter Gab2 in Fcγ R-mediated phagocytosis. JCell Biol 2003, 161:1151-1161.27. Vieira OV, Botelho RJ, Rameh L, Brachmann SM, Matsuo T, DavidsonHW, Schreiber A, Backer JM, Cantley LC, Grinstein S: Distinct rolesPage 12 of 13(page number not for citation purposes)terial invasion process. We thank Kyle Recsky and Shawn Anderson (Upstream Biosciences, Inc) for their advice and help on the model devel-of class I and class III phosphatidylinositol 3-kinase in phago-some formation and maturation. J Cell Biol 2001, 155:19-25.Publish with BioMed Central   and  every scientist can read your work free of charge"BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime."Sir Paul Nurse, Cancer Research UKYour research papers will be:available free of charge to the entire biomedical communitypeer reviewed and published immediately upon acceptancecited in PubMed and archived on PubMed Central BMC Bioinformatics 2004, 5:156 http://www.biomedcentral.com/1471-2105/5/15628. Vanhaesebroeck B, Waterfield MD: Signaling by distinct classesof phosphoinositide 3-kinases. Exp Cell Res 1999, 253:239-254.29. Cantley LC: The phosphoinositide 3-kinse pathway. Science2002, 296:1655-1657.30. Vanhaesebroeck B, Leevers SJ, Ahmadi K, Timms J, Katso R, DriscollPC, Woscholski R, Parker PJ, Waterfield MD: Synthesis and func-tion of 3-phosphorylated inositol lipids. Annu Rev Biochem 2001,70:535-602.31. Stephens L, Ellson C, Hawkins P: Roles of PI3Ks in leukocytechemotaxis and phagocytosis. Curr Opin Cell Biol 2002,14:203-213.32. Fratti RA, Backer JM, Grunerberg J, Corvera S, Deretic V: Role ofphosphatidylinositol 3-kinase and Rab5 effectors in phago-somal biogenesis and mycobacterial phagosome maturationarrest. J Cell Biol 2001, 154:631-644.33. Wymann MP, Zvelebil M, Laffargue M: Phosphoinositide 3-kinasesignalling – which way to target? Trends Pharmacol Sci 2003,24:366-376.34. Araki N, Johnson MT, Swanson JA: A role for phosphoinositide 3-kinase in the completion of macropinocytosis and phagocy-tosis by macrophages. J Cell Biol 1996, 135:1249-1260.35. Arbibe L, Mira JP, Teusch N, Kline L, Guha M, Mackman N, GodowskiPJ, Ulevitch RJ, Knaus UG: Toll-like receptor 2-mediated NF-κBactivation requires a Rac1-dependent pathway. Nat Immunol2000, 1:533-540.36. Bokoch G.M, Vlahos CJ, Wang Y, Knaus UG, Traynor-Kaplan AE: RacGTPase interacts specifically with phosphatidylinositol 3-kinase. Biochem J 1996, 315:775-779.37. Celli J, Oliver M, Finlay BB: Enteropathogenic Escherichia colimediates antiphagocytosis through the inhibition of PI 3-kinase-dependent pathways. EMBO J 2001, 20:1245-1258.38. Cox D, Tseng CC, Bjekic G., Greenberg S: A requirement forphosphatidylinositol 3-kinase in pseudopod extension. J BiolChem 1999, 274:1240-1247.39. Hmama Z, Knutson KL, Herrera-Velit P, Nandan D, Reiner NE:Monocyte adherence induced by lipopolysaccharide involvesCD14, LFA-1, and cytohesin-1. J Biol Chem 1999, 274:1050-1057.40. Hmama Z, Nandan D, Sly L, Knutson KL, Herrera-Velit P, Reiner NE:1α, 25-Dihydroxyvitamin D3-induced myeloid cell differenti-ation is regulated by a vitamin D receptor-phosphatidylinosi-tol 3-kinase signaling complex. J Exp Med 1999, 190:1583-1594.41. Ireton K, Payrastre B, Chap H, Sakaue H, Kasuga M, Cossart P: A rolefor phosphoinositide 3-kinase in bacterial invasion. Science1996, 274:780-2.42. Ireton K, Payrastre B, Cossart P: The Listeria monocytogenes pro-tein InlB is an agonist of mammalian phosphoinositide 3-kinase. J Biol Chem 1999, 274:17025-17032.43. Leverrier Y, Okkenhaug K, Sawyer C, Bilancio A, Vanhaesebroeck B,Ridley AJ: Class I phosphoinositide 3-kinase p110β is requiredfor apoptotic cell and Fcγ receptor-mediated phagocytosisby macrophage. J Biol Chem 2003, 278:38437-38442.44. Martinez JJ, Mulvey MA, Schilling JD, Pinkner JS, Hultgren SJ: Type 1pilus-mediated bacterial invasion of bladder epithelial cells.EMBO J 2000, 19:2803-2812.45. Muta T, Takeshige K: Essential roles of CD14 and lipopolysac-charide-binding protein for activation of toll-like receptor(TLR)2 as well as TLR4 Reconstitution of TLR2- and TLR4-activation by distinguishable ligands in LPS preparations. EurJ Biochem 2001, 268:4580-4589.46. Odorizzi G., Babst M, Emr SD: Phosphoinositide signaling andthe regulation of membrane trafficking in yeast. Trends Bio-chem Sci 2000, 25:229-235.47. Sly LM, Lopez M, Nauseef WM, Reiner NE: 1α, 25-dihydroxyvita-min D3-induced monocyte antimycobacterial activity is reg-ulated by phosphatidylinositol 3-kinase and mediated by theNADPH-dependent phagocyte oxidase. J Biol Chem 2001,276:35482-35493.48. Finlay BB, Falkow S: Common themes in microbial pathogenic-ity revisited. Microbiol Mol Biol Rev 1997, 61:136-169.49. Fruman DA, Cantley LC: Phosphoinositide 3-kinase in immuno-logical systems. Semin Immunol 2002, 14:7-18.50. Martin TFL: Phosphoinositide lipids as signaling molecules:Common themes for signal transduction, cytoskeletal regu-lation, and membrane trafficking. Annu Rev Cell Dev Biol 1998,51. Russell DG: Mycobacterium tuberculosis : here today, and heretomorrow. Nat Rev Mol Cell Biol 2001, 2:569-577.52. Velasco-Velazquez MA, Barrera D, Gonzalez-Arenas A, Rosales C,Agramonte-Hevia J: Macrophage-Mycobacterium tuberculosisinteractions: role of complement receptor 3. Microb Pathog2003, 35:125-131.53. Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter P: Molecu-lar biology of the cell. 4th edition. New York: Garland Science;2002:299-335. 54. Ideker T, Ozier O, Schwikowski B, Siegel AF: Discovering regula-tory and signalling circuits in molecular interactionnetworks. Bioinformatics 2002, 18(Suppl 1):S233-40.55. PSORT  [http://psort.nibb.ac.jp/]yours — you keep the copyrightSubmit your manuscript here:http://www.biomedcentral.com/info/publishing_adv.aspBioMedcentralPage 13 of 13(page number not for citation purposes)14:231-264.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items