A dynamic network of transcription in LPS-treated human subjects
© Seok et al; licensee BioMed Central Ltd. 2009
Received: 27 October 2008
Accepted: 28 July 2009
Published: 28 July 2009
Understanding the transcriptional regulatory networks that map out the coordinated dynamic responses of signaling proteins, transcription factors and target genes over time would represent a significant advance in the application of genome wide expression analysis. The primary challenge is monitoring transcription factor activities over time, which is not yet available at the large scale. Instead, there have been several developments to estimate activities computationally. For example, Network Component Analysis (NCA) is an approach that can predict transcription factor activities over time as well as the relative regulatory influence of factors on each target gene.
In this study, we analyzed a gene expression data set in blood leukocytes from human subjects administered with lipopolysaccharide (LPS), a prototypical inflammatory challenge, in the context of a reconstructed regulatory network including 10 transcription factors, 99 target genes and 149 regulatory interactions. We found that the computationally estimated activities were well correlated to their coordinated action. Furthermore, we found that clustering the genes in the context of regulatory influences greatly facilitated interpretation of the expression data, as clusters of gene expression corresponded to the activity of specific factors or more interestingly, factor combinations which suggest coordinated regulation of gene expression. The resulting clusters were therefore more biologically meaningful, and also led to identification of additional genes under the same regulation.
Using NCA, we were able to build a network that accounted for between 8–11% genes in the known transcriptional response to LPS in humans. The dynamic network illustrated changes of transcription factor activities and gene expressions as well as interactions of signaling proteins, transcription factors and target genes.
An achievement that would have a major impact on our understanding of transcriptional regulatory networks would be to map out the coordinated dynamic responses of signaling proteins, transcription factors and target genes over time. The primary challenges to such an effort are development of high-throughput technologies to measure transcription factor activities at the genome-scale, and computational tools to interpret the data and predict the structure and dynamics of the underlying networks.
Recent development of high-throughput technologies has enabled large-scale measurements of biological signals related to transcription, such as the expression of target genes and the activities of transcription factors. For target gene expression, microarrays measure the expression levels of thousands of genes simultaneously [1–3]. However, efforts to broadly assess transcription factor activities on a genome wide scale are much more limited. Technologies such as chromatin immunoprecipitation-on-a-chip can identify all of the DNA binding sites occupied by a single transcription factor for a given condition [4, 5]. Flow cytometry can also be used to determine transcription factor activities by labeling active factors with fluorescently labeled antibodies , but throughput is limited by the number of available antibodies and colors. As yet, there is no transcription factor-focused equivalent of the gene expression array, which would enable monitoring of all transcription factor activities at a time. Such technology would be critical to generating a complete dynamic network of transcription empirically.
To compensate for this inability to assay transcription factor activity at the large scale, there have been several efforts to infer regulatory networks computationally . One of these approaches, called Network Component Analysis (NCA), is a method for determining both activities and regulatory influence for a set of transcription factors with known target genes . NCA has been successfully applied in several areas. It was used to identify previously unnoticed oscillatory activity patterns in the yeast cell cycle , as well as to generate a predicted activation time course of catabolite repressor protein in Escherichia coli, which was verified experimentally . More recently, NCA was used to predict activities of important transcription factors like sterol regulatory element-binding proteins and peroxisome proliferative-activated receptors in a mouse knockout model of human glycerol kinase deficiency [10, 11]. In parallel, several studies have expanded and strengthened NCA as a computational tool [12–14].
In eukaryotic systems, inflammation and activation of innate immunity are fundamental host responses to microbial invasion and endogenous danger signals. Blood leukocytes contribute to this inflammatory response, and exposure to a prototypical stimulus such as LPS leads first to changes in gene expression, then production of cytokines which are secreted and cause secondary transcriptional and other responses . In previous work, we and others generated a set of gene expression profiles from human subjects over 24 hours following the intravenous administration of bacterial endotoxin LPS . Experimental endotoxicosis produces in the previously healthy individual a transient but significant systemic inflammatory response, characterized by fever, tachycardia, malaise, and a hepatic acute phase response. Administration of endotoxin is presumed to model the early inflammatory changes associated with a microbial invasion, sepsis and the systemic inflammatory response syndrome . We used this data to determine important clusters of genes involved in the early inflammatory response, as well as to depict the temporal changes in gene expression as inflammation resolved over the first twenty-four hours. In this study, we calculated transcription factor activities and regulatory influences in the above dataset using NCA, and interpreted the results to develop a dynamic network of transcription events following experimental endotoxicosis in humans.
Results and Discussion
To define a regulatory network which could account for a significant percentage of the gene expression response, we identified a set of key transcription factors previously known to be involved in the LPS response, together with a set of known target genes for these factors. Ten transcription factors were chosen for our study (listed here by gene name for continuity). NFKB1 (encoding p50/p105), RELA (encoding p65) and IRF3 were chosen as factors involved in the primary response to endotoxin. Endotoxin binding to Toll-like receptors (TLR) leads to activation of NF-κB dimers, among which p65:p50 is common . LPS stimulation also induces IRF3 activation through TLR4 . These transcription factors induce expression of several cytokines which can further activate a secondary transcription response through factors such as STAT1, 3 and 6 [15, 20]. CREB1 is activated by LPS through the p38 kinase-SAPK2 pathway . It is known that LPS activates AP-1 complexes consisting of FOS, JUN, JUNB and JUND . Among these factors, JUN and FOS were chosen for our model. The role of MYC in inflammation response is poorly understood ; however, many genes connected to MYC showed significant changes in their expression levels in the original study and so we included it as well .
To identify established regulatory interactions between these transcription factors and target genes, we relied largely on the primary literature [15, 16, 20, 22, 24–28] (Figure 1A). However, two knowledge-bases were also used: Ingenuity Systems http://www.ingenuity.com and Pathway Studio . Both the Ingenuity and Pathway Studio knowledge-bases consist of regulatory relationships parsed from MEDLINE abstracts; the Ingenuity knowledge-base also includes information from manually-curated peer-reviewed publications. For our ten transcription factors, this strategy resulted in a list of 1,287 target genes, with 2,183 interactions between transcription factors and target genes. To reconcile differences in these different sources of regulatory network information (literature, Ingenuity, Pathway Studio), we only included an interaction in our network if it could be identified in two out of the three resources. This filtering process reduced our list to 219 target genes regulated by 306 interactions with the ten transcription factors. To focus on the most useful expression information, we only considered target genes for which expression changed significantly over time (p-value < 0.01).
The network for the inflammatory response finally included 10 transcription factors, 99 target genes and 149 regulatory relations. This network can be represented in matrix form, with a density of ~15%, or 149 relations/(10 factors × 99 targets). In contrast, the expected density of a genome-wide regulatory relationship matrix, given our current state of knowledge about human transcriptional regulation would be about 0.1% (~20,000 relations in Ingenuity Systems and Pathway Studio databases, ~1,000 transcription factors and ~20,000 target genes). Our network density is therefore relatively high, reflecting the comparatively high level of research interest in this system.
We estimated the activation of the transcription factors in our network over time using NCA (Figure 1B). NCA decomposes a matrix containing gene expression values (E) into a matrix which represents the influence of a transcription factor on a target gene (strength matrix S) and a matrix which contains the transcription factor activities (activity matrix A) . We found that both outputs of NCA – predicted factor activities A and regulatory influences S – have added additional insights to gene expression data where the underlying regulatory network structure is partially known.
Transcription factor activities
Transcription factor activities are sometimes, but not always, correlated with the gene expression of the factor. We compared the calculated transcription factor activities with the gene expression data for each factor (Figure 2B). NFKB1, RELA, STAT1, STAT3 and MYC showed strong positive correlation between activities and expression (correlation coefficient c > 0.56), possibly due to auto- or cross-regulation. For example, NFKB1 activity and expression are tightly correlated (c = 0.6022), possibly because the NF-κB p65:p50 complex can regulate NFKB1 [38, 39]. STAT1 activity and expression are also strongly correlated (c = 0.8362), which might relate to the transcriptional effect of STAT3 on STAT1 expression , particularly given that STAT1 and STAT3 have highly correlated activities (Figure 2C, c = 0.9329). On the other hand, the activities and expression show lower or no correlation for IRF3, JUN, FOS and CREB (-0.15 <c < 0.37).
The linear model of gene expression upon which NCA rests does not account for the interactions between transcription factors. However, we wondered if the NCA-predicted correlation in transcription factor activities could be due to the combined action of two transcription factors, either as a complex or otherwise. We therefore checked transcription factor pairs with significant activity correlation to published protein-protein interactions catalogued in the Biomolecular Interaction Network Database (BIND) . Interestingly, transcription factors known to act together showed high correlation in their activity profiles (Figure 2D). For example, highly correlated transcription factors NFKB1(p50/p105) and RELA(p65) regulate their target genes as a p65:50 heterodimer form , and STAT1 and STAT3 are also known to act as a dimer , as are JUN and FOS . Additionally, some transcription factors (STAT1 and CREB1, STAT6 and MYC) showed a positive correlation in their activity even though they are not known to form a complex with other transcription factors. Transcription factors can have similar – and even coordinated – activities without direct interaction, so it may be that these latter predictions reflect an indirect interaction.
NCA simulation with random noisy connections
Regulatory influence matrix and gene expression clustering
Major clusters formed from the adjusted strength matrix.
Num. of genes
Dominant transcription factors
Avg. of pair-wise correlations
Correlation test and prediction over extended regulatory sets
We first wanted to see if a gene in the extended gene group had similar expression to a cluster (Figure 4B). First, Pearson's correlations were calculated between each gene in the extended gene group and the average normalized gene expression of each cluster. We then also randomly selected one thousand genes from the no-evidence gene group, and calculated correlations between expression of these genes and the clusters. To obtain standard deviations, we performed this step one hundred times. The fraction of genes with a Pearson's correlation > 0.5 was then compared between both groups using Fisher's exact test (Figure 4C). We found that average gene expression in each cluster was more highly correlated with genes in the corresponding extended gene set than in the no-evidence gene set, particularly for Clusters A, B and G.
Predicted genes for Cluster A and B from the extended gene sets.
CASP8 and FADD-like apoptosis regulator
chemokine (C-X-C motif) ligand 1
interleukin 1 receptor, type I
interleukin 1 receptor antagonist
oxidized low density lipoprotein (lectin-like) receptor 1
polo-like kinase 3
pentraxin-related gene, rapidly induced by IL-1 beta
runt-related transcription factor 1
tumor necrosis factor, alpha-induced protein 6
TNFAIP3 interacting protein 1
caspase 4, apoptosis-related cysteine peptidase
cytokine inducible SH2-containing protein
guanylate binding protein 1, interferon-inducible, 67 kDa
guanylate binding protein 2, interferon-inducible
GTP cyclohydrolase 1
heat shock 70 kDa protein 1A
interferon, gamma-inducible protein 16
interferon-induced protein with tetratricopeptide repeats 3
interferon induced transmembrane protein 1
insulin-like growth factor 1 receptor
interleukin 10 receptor, beta
interferon regulatory factor 2
ISG15 ubiquitin-like modifier
interferon stimulated exonuclease gene 20 kDa
Janus kinase 2
myxovirus (influenza virus) resistance 1, interferon-inducible protein p78
phospholipid scramblase 1
receptor (TNFRSF)-interacting serine-threonine kinase 1
suppressor of cytokine signaling 1
signal transducer and activator of transcription 2
TIMP metallopeptidase inhibitor 1
ubiquitin specific peptidase 18
Predicted genes for Cluster A and B form the "no evidence" group
v-ets erythroblastosis virus E26 oncogene homolog 2
metal-regulatory transcription factor 1
SAM domain, SH3 domain and nuclear localization signals 1
influenza virus NS1A binding protein
interferon gamma receptor 2
plasminogen activator, urokinase receptor
interleukin 1 receptor, type II
antizyme inhibitor 1
EH-domain containing 1
Related to innate immunity
interferon gamma receptor 2
interleukin 1 receptor, type II
mitogen-activated protein kinase kinase kinase kinase 4
neutrophil cytosolic factor 1
receptor-interacting serine-threonine kinase 2
Supported by other evidence
interferon gamma receptor 2
G protein-coupled receptor 84
Fc fragment of IgA, receptor for
growth arrest and DNA-damage-inducible, beta
tripartite motif-containing 5
SP110 nuclear body protein
transmembrane protein 140
chromatin modifying protein 5
rhomboid 5 homolog 2
sterile alpha motif domain containing 9
DEAD (Asp-Glu-Ala-Asp) box polypeptide 58
transducin-like enhancer of split 3
tripartite motif-containing 21
Related to innate immunity
colony stimulating factor 2 receptor, beta, low-affinity
hemopoietic cell kinase
interleukin 13 receptor, alpha 1
toll-like receptor 1
tumor necrosis factor receptor superfamily, member 1A
Supported by other evidence
DEAD (Asp-Glu-Ala-Asp) box polypeptide 5
hect domain and RLD 5
interferon-induced protein with tetratricopeptide repeats 2
sorting nexin 10
acyl-CoA synthetase long-chain family member 4
Overall regulatory dynamics in response to LPS
The overall goal of this work was to build a dynamic network of transcription events following endotoxin administration from the time course response by global gene expression in peripheral blood leukocytes. From the expression profiles, we were able to predict the activities of ten transcription factors over time, as well as the regulatory strength a given transcription factor exerted on its target genes using NCA. Taken together, the activities often exhibited a high degree of correlation, both between factors and also between a factor's activity and its gene expression profile.
We also found that the regulatory strength matrix can be clustered to determine groups of genes which are not only co-expressed, but also co-regulated. Importantly, new and biologically relevant clusters were determined, suggesting that clustering by this approach is potentially more meaningful than methods which do not incorporate regulatory network information. Identification of these clusters also led us to identify many additional putative interactions between transcription factors and target genes not included in the known network, and most importantly, enabled us to describe and visualize the activation of regulatory proteins and target genes over time.
Certain limitations in both the available expression data as well as NCA itself could be addressed to make this approach more powerful. Gene expression analyses obtained from whole blood leukocyte samples provide an integrated signal from different leukocyte populations which are difficult to deconvolute, and so using a single cell population would be advantageous, such as could be obtained using cell sorting or other methods. Additionally, the number of transcription factors which can be used in NCA is approximately the number of expression profiles in the data set, and so a greater number of expression profiles – obtained at best shortly after the endotoxin administration – would also have been useful. Finally, NCA's scaling property, which makes it difficult to predict the direction of transcription factor activity, as well as NCA's current inability to incorporate time course information from the data set are important limitations to the method. Some approaches that may overcome these challenges include recent studies in which transcription factor activities were estimated using ordinary differential equation  or probabilistic models [51, 52] of time course data. Future work might therefore focus on combining NCA with such efforts.
Notwithstanding these limitations, we were able to reconstruct the dynamics of endotoxin-dependent transcription in human peripheral blood leukocytes using the above results. This included identifying the activity of ten transcription factors regulating expression of ninety-nine genes. We also were able to identify additional genes that could be included in our model, notably 36 which had less initial evidence, but were substantiated by our predictions. Given that there were 1,215 genes with significant changes in gene expression for which regulatory relations were known, we were therefore able to capture between 8% (= 99 initial model genes/1,215 genes with significant expression changes and known regulatory relations) and 11% (99 + 36 additional genes = 135/1,215) of the explainable response. Furthermore, we were also able to identify new target genes based on the average gene expression profile of significant clusters, which could expand the scope of our temporal network still further. With a larger network reconstruction and data set specifically designed for use with NCA, it might be possible to move toward a near-complete characterization of dynamic transcription responses.
Data preprocessing and statistical analysis
To process our gene expression dataset prior to NCA, the log2 ratio of post-injection time points to the pre-injection time point was calculated. The significance of expression changes was then tested using one-way ANOVA, where the null hypothesis was that average gene expression levels were the same for each time point. We selected genes for our model if the ANOVA p-value was less than 0.01. Among 18,398 genes in the dataset, 5,518 genes were determined to be induced or repressed significantly. 1,215 of the genes that experienced a significant change in expression also had information about their regulation in the knowledge-bases we used.
Network component analysis
where S•jand Aj•represent the j th column of S or row of A, respectively. This normalization stabilizes the calculation by preventing too large or too small values of A, but has no effect on the overall results due to NCA's scaling property .
Hierarchical clustering on adjusted strength matrix
where <,> represents the inner product of two vectors. We used hierarchical clustering to divide the adjusted strength matrix into meaningful clusters using angle cosine as the distance metric.
Transcription factors for which the contribution factor was larger than 0.2 were chosen as dominant transcription factors of the cluster.
This work was funded by the NIH through the Inflammation and the Host Response to Injury research project (U54-GM62119), as well as a K99/R00 award to MWC (CA125994-01A1). We thank Jonathan Karr and Robert Winfield for critical reading of the manuscript. The following is a list of participating investigators of the Inflammation and the Host Response to Injury research project: Henry V. Baker, Ulysses GJ. Balis, Paul E. Bankey, Timothy R. Billiar, Bernard H. Brownstein, Steven E. Calvano, David G. Camp II, Irshad H. Chaudry, J. Perren Cobb, Joseph Cuschieri, Ronald W. Davis, Asit K. De, Celeste C. Finnerty, Bradley Freeman, Richard L. Gamelli, Nicole S. Gibran, Brian G. Harbrecht, Douglas L. Hayden, Laura Hennessy, David N. Herndon, Marc G. Jeschke, Jeffrey L. Johnson, Matthew B. Klein, James A. Lederer, Stephen F. Lowry, Ronald V. Maier, John A. Mannick, Philip H. Mason, Grace P. McDonald-Smith, Carol L. Miller-Graziano, Michael N. Mindrinos, Joseph P. Minei, Lyle L. Moldawer, Ernest E. Moore, Avery B. Nathens, Grant E. O'Keefe, Laurence G. Rahme, Daniel G. Remick, David A. Schoenfeld, Michael B. Shapiro, Geoffrey M. Silver, Richard D. Smith, John D. Storey, Robert Tibshirani, Ronald G. Tompkins, Mehmet Toner, H. Shaw Warren, Michael A. West, Rebbecca P. Wilson, Wenzhong Xiao.
- Schena M, Shalon D, Davis RW, Brown PO: Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science. 1995, 270: 467-70. 10.1126/science.270.5235.467View ArticlePubMedGoogle Scholar
- Chee M, Yang R, Hubbell E, Berno A, Huang XC, Stern D, Winkler J, Lockhart DJ, Morris MS, Fodor SP: Accessing genetic information with high-density DNA arrays. Science. 1996, 274: 610-4. 10.1126/science.274.5287.610View ArticlePubMedGoogle Scholar
- Heller RA, Schena M, Chai A, Shalon D, Bedilion T, Gilmore J, Woolley DE, Davis RW: Discovery and analysis of inflammatory disease-related genes using cDNA microarrays. Proc Natl Acad Sci USA. 1997, 94: 2150-5. 10.1073/pnas.94.6.2150PubMed CentralView ArticlePubMedGoogle Scholar
- Ren B, Robert F, Wyrick JJ, Aparicio O, Jennings EG, Simon I, Zeitlinger J, Schreiber J, Hannett N, Kanin E, Volkert TL, Wilson CJ, Bell SP, Young RA: Genome-wide location and function of DNA binding proteins. Science. 2000, 290: 2306-9. 10.1126/science.290.5500.2306View ArticlePubMedGoogle Scholar
- Kim TH, Ren B: Genome-wide analysis of protein-DNA interactions. Annu Rev Genomics Hum Genet. 2006, 7: 81-102. 10.1146/annurev.genom.7.080505.115634View ArticlePubMedGoogle Scholar
- Irish JM, Kotecha N, Nolan GP: Mapping normal and cancer cell signalling networks: towards single-cell proteomics. Nat Rev Cancer. 2006, 6: 146-55. 10.1038/nrc1804View ArticlePubMedGoogle Scholar
- Herrgard MJ, Covert MW, Palsson BØ: Reconstruction of microbial transcriptional regulatory networks. Curr Opin Biotechnol. 2004, 15: 70-7. 10.1016/j.copbio.2003.11.002View ArticlePubMedGoogle Scholar
- Liao JC, Boscolo R, Yang YL, Tran LM, Sabatti C, Roychowdhury VP: Network component analysis: reconstruction of regulatory signals in biological systems. Proc Natl Acad Sci USA. 2003, 100: 15522-7. 10.1073/pnas.2136632100PubMed CentralView ArticlePubMedGoogle Scholar
- Kao KC, Yang YL, Boscolo R, Sabatti C, Roychowdhury V, Liao JC: Transcriptome-based determination of multiple transcription regulator activities in Escherichia coli by using network component analysis. Proc Natl Acad Sci USA. 2004, 101: 641-6. 10.1073/pnas.0305287101PubMed CentralView ArticlePubMedGoogle Scholar
- MacLennan NK, Rahib L, Shin C, Fang Z, Horvath S, Dean J, Liao JC, McCabe ER, Dipple KM: Targeted disruption of glycerol kinase gene in mice: expression analysis in liver shows alterations in network partners related to glycerol kinase activity. Hum Mol Genet. 2005, 15: 405-15. 10.1093/hmg/ddi457View ArticlePubMedGoogle Scholar
- Rahib L, MacLennan NK, Horvath S, Liao JC, Dipple KM: Glycerol kinase deficiency alters expression of genes involved in lipid metabolism, carbohydrate metabolism, and insulin signaling. Eur J Hum Genet. 2007, 15: 646-57. 10.1038/sj.ejhg.5201801View ArticlePubMedGoogle Scholar
- Galbraith SJ, Tran LM, Liao JC: Transcriptome network component analysis with limited microarray data. Bioinformatics. 2006, 22: 1886-94. 10.1093/bioinformatics/btl279View ArticlePubMedGoogle Scholar
- Tran LM, Brynildsen MP, Kao KC, Suen JK, Liao JC: gNCA: a framework for determining transcription factor activity based on transcriptome: identifiability and numerical implementation. Metab Eng. 2005, 7: 128-41. 10.1016/j.ymben.2004.12.001View ArticlePubMedGoogle Scholar
- Boscolo R, Sabatti C, Liao JC, Roychowdhury VP: A generalized framework for network component analysis. IEEE/ACM Trans Comput Biol Bioinform. 2005, 2: 289-301. 10.1109/TCBB.2005.47View ArticlePubMedGoogle Scholar
- Doyle S, Vaidya S, O'Connell R, Dadgostar H, Dempsey P, Wu T, Rao G, Sun R, Haberland M, Modlin R, Cheng G: IRF3 mediates a TLR3/TLR4-specific antiviral gene program. Immunity. 2002, 17: 251-63. 10.1016/S1074-7613(02)00390-4View ArticlePubMedGoogle Scholar
- Calvano SE, Xiao W, Richards DR, Felciano RM, Baker HV, Cho RJ, Chen RO, Brownstein BH, Cobb JP, Tschoeke SK, Miller-Graziano C, Moldawer LL, Mindrinos MN, Davis RW, Tompkins RG, Lowry SF, : A network-based analysis of systemic inflammation in humans. Nature. 2005, 437: 1032-7. 10.1038/nature03985View ArticlePubMedGoogle Scholar
- Fong YM, Marano MA, Moldawer LL, Wei H, Calvano SE, Kenney JS, Allison AC, Cerami A, Shires GT, Lowry SF: The acute splanchnic and peripheral tissue metabolic response to endotoxin in humans. J Clin Invest. 1990, 85: 1896-904. 10.1172/JCI114651PubMed CentralView ArticlePubMedGoogle Scholar
- Ghosh S, May MJ, Kopp EB: NF-kappa B and Rel proteins: evolutionarily conserved mediators of immune responses. Annu Rev Immunol. 1998, 16: 225-60. 10.1146/annurev.immunol.16.1.225View ArticlePubMedGoogle Scholar
- Aksoy E, Albarani V, Nguyen M, Laes JF, Ruelle JL, De Wit D, Willems F, Goldman M, Goriely S: Interferon regulatory factor 3-dependent responses to lipopolysaccharide are selectively blunted in cord blood cells. Blood. 2007, 109: 2887-93.PubMedGoogle Scholar
- Shuai K, Liu B: Regulation of JAK-STAT signalling in the immune system. Nat Rev Immunol. 2003, 3: 900-11. 10.1038/nri1226View ArticlePubMedGoogle Scholar
- Rolli M, Kotlyarov A, Sakamoto KM, Gaestel M, Neininger A: Stress-induced stimulation of early growth response gene-1 by p38/stress-activated protein kinase 2 is mediated by a cAMP-responsive promoter element in a MAPKAP kinase 2-independent manner. J Biol Chem. 1999, 274: 19559-64. 10.1074/jbc.274.28.19559View ArticlePubMedGoogle Scholar
- Medvedev AE, Blanco JC, Qureshi N, Vogel SN: Limited role of ceramide in lipopolysaccharide-mediated mitogen-activated protein kinase activation, transcription factor induction, and cytokine release. J Biol Chem. 1999, 274: 9342-50. 10.1074/jbc.274.14.9342View ArticlePubMedGoogle Scholar
- Soucek L, Lawlor ER, Soto D, Shchors K, Swigart LB, Evan GI: Mast cells are required for angiogenesis and macroscopic expansion of Myc-induced pancreatic islet tumors. Nat Med. 2007, 13: 1211-8. 10.1038/nm1649View ArticlePubMedGoogle Scholar
- Uematsu S, Akira S: Toll-like receptors and innate immunity. J Mol Med. 2006, 84: 712-25. 10.1007/s00109-006-0084-yView ArticlePubMedGoogle Scholar
- Covert MW, Leung TH, Gaston JE, Baltimore D: Achieving stability of lipopolysaccharide-induced NF-kappaB activation. Science. 2005, 309: 1854-7. 10.1126/science.1112304View ArticlePubMedGoogle Scholar
- Werner SL, Barken D, Hoffmann A: Stimulus specificity of gene expression programs determined by temporal control of IKK activity. Science. 2005, 309: 1857-61. 10.1126/science.1113319View ArticlePubMedGoogle Scholar
- Oda K, Kitano H: A comprehensive map of the toll-like receptor signaling network. Mol Syst Biol. 2006, 2: 2006.0015- 10.1038/msb4100057PubMed CentralView ArticlePubMedGoogle Scholar
- Eliopoulos AG, Dumitru CD, Wang CC, Cho J, Tsichlis PN: Induction of COX-2 by LPS in macrophages is regulated by Tpl2-dependent CREB activation signals. EMBO J. 2002, 21: 4831-40. 10.1093/emboj/cdf478View ArticlePubMedGoogle Scholar
- Nikitin A, Egorov S, Daraselia N, Mazo I: Pathway studio–the analysis and navigation of molecular networks. Bioinformatics. 2003, 19: 2155-7. 10.1093/bioinformatics/btg290View ArticlePubMedGoogle Scholar
- Guha M, Mackman N: LPS induction of gene expression in human monocytes. Cell Signal. 2001, 13: 85-94. 10.1016/S0898-6568(00)00149-2View ArticlePubMedGoogle Scholar
- Banerjee A, Gerondakis S: Coordinating TLR-activated signaling pathways in cells of the immune system. Immunol Cell Biol. 2007, 85: 420-4. 10.1038/sj.icb.7100098View ArticlePubMedGoogle Scholar
- Ardeshna KM, Pizzey AR, Devereux S, Khwaja A: The PI3 kinase, p38 SAP kinase, and NF-kappaB signal transduction pathways are involved in the survival and maturation of lipopolysaccharide-stimulated human monocyte-derived dendritic cells. Blood. 2000, 96: 1039-46.PubMedGoogle Scholar
- Giono LE, Varone CL, Canepa ET: 5-Aminolaevulinate synthase gene promoter contains two cAMP-response element (CRE)-like sites that confer positive and negative responsiveness to CRE-binding protein (CREB). Biochem J. 2001, 353: 307-16. 10.1042/0264-6021:3530307PubMed CentralView ArticlePubMedGoogle Scholar
- Belmonte N, Phillips BW, Massiera F, Villageois P, Wdziekonski B, Saint-Marc P, Nichols J, Aubert J, Saeki K, Yuo A, Narumiya S, Ailhaud G, Dani C: Activation of extracellular signal-regulated kinases and CREB/ATF-1 mediate the expression of CCAAT/enhancer binding proteins beta and – delta in preadipocytes. Mol Endocrinol. 2001, 15: 2037-49. 10.1210/me.15.11.2037PubMedGoogle Scholar
- Takeda K, Tanaka T, Shi W, Matsumoto M, Minami M, Kashiwamura S, Nakanishi K, Yoshida N, Kishimoto T, Akira S: Essential role of Stat6 in IL-4 signalling. Nature. 1996, 380: 627-30. 10.1038/380627a0View ArticlePubMedGoogle Scholar
- Dickensheets HL, Venkataraman C, Schindler U, Donnelly RP: Interferons inhibit activation of STAT6 by interleukin 4 in human monocytes by inducing SOCS-1 gene expression. Proc Natl Acad Sci USA. 1999, 96: 10800-5. 10.1073/pnas.96.19.10800PubMed CentralView ArticlePubMedGoogle Scholar
- Ramana CV, Grammatikakis N, Chernov M, Nguyen H, Goh KC, Williams BR, Stark G: Regulation of c-myc expression by IFN-gamma through Stat1-dependent and -independent pathways. EMBO J. 2000, 19: 263-72. 10.1093/emboj/19.2.263PubMed CentralView ArticlePubMedGoogle Scholar
- Ten RM, Paya CV, Israël N, Le Bail O, Mattei MG, Virelizier JL, Kourilsky P, Israël A: The characterization of the promoter of the gene encoding the p50 subunit of NF-kappa B indicates that it participates in its own regulation. EMBO J. 1992, 11: 195-203.PubMed CentralPubMedGoogle Scholar
- Cogswell PC, Scheinman RI, Baldwin AS: Promoter of the human NF-kappa B p50/p105 gene. Regulation by NF-kappa B subunits and by c-REL. J Immunol. 1993, 150: 2794-804.PubMedGoogle Scholar
- He F, Ge W, Martinowich K, Becker-Catania S, Coskun V, Zhu W, Wu H, Castro D, Guillemot F, Fan G, de Vellis J, Sun YE: A positive autoregulatory loop of Jak-STAT signaling controls the onset of astrogliogenesis. Nat Neurosci. 2005, 8: 616-25. 10.1038/nn1440PubMed CentralView ArticlePubMedGoogle Scholar
- Bader GD, Donaldson I, Wolting C, Ouellette BF, Pawson T, Hogue CW: BIND–The Biomolecular Interaction Network Database. Nucleic Acids Res. 29: 242-5.
- Gilmore TD: Introduction to NF-kappaB: players, pathways, perspectives. Oncogene. 2006, 25: 6680-4. 10.1038/sj.onc.1209954View ArticlePubMedGoogle Scholar
- Hoffmann A, Leung TH, Baltimore D: Genetic analysis of NF-kappaB/Rel transcription factors defines functional specificities. EMBO J. 2003, 22: 5530-9. 10.1093/emboj/cdg534PubMed CentralView ArticlePubMedGoogle Scholar
- Hayden MS, West AP, Ghosh S: NF-kappaB and the immune response. Oncogene. 2006, 25: 6758-80. 10.1038/sj.onc.1209943View ArticlePubMedGoogle Scholar
- Barenco M, Tomescu D, Brewer D, Callard R, Stark J, Hubank M: Ranked prediction of p53 targets using hidden variable dynamic modeling. Genome Biol. 2006, 7: R25- 10.1186/gb-2006-7-3-r25PubMed CentralView ArticlePubMedGoogle Scholar
- Schreiber J, Jenner RG, Murray HL, Gerber GK, Gifford DK, Young RA: Coordinated binding of NF-kappaB family members in the response of human cells to lipopolysaccharide. Proc Natl Acad Sci USA. 2006, 103: 5899-904. 10.1073/pnas.0510996103PubMed CentralView ArticlePubMedGoogle Scholar
- Tabach Y, Brosh R, Buganim Y, Reiner A, Zuk O, Yitzhaky A, Koudritsky M, Rotter V, Domany E: Wide-scale analysis of human functional transcription factor binding reveals a strong bias towards the transcription start site. PLoS ONE. 2007, 2: e807- 10.1371/journal.pone.0000807PubMed CentralView ArticlePubMedGoogle Scholar
- Li J, Yu B, Song L, Eschrich S, Haura EB: Effects of IFN-gamma and Stat1 on gene expression, growth, and survival in non-small cell lung cancer cells. J Interferon Cytokine Res. 2007, 27: 209-20. 10.1089/jir.2006.0111View ArticlePubMedGoogle Scholar
- Chedid M, Yoza BK, Brooks JW, Mizel SB: Activation of AP-1 by IL-1 and phorbol esters in T cells. Role of protein kinase A and protein phosphatases. J Immunol. 1991, 147: 867-73.PubMedGoogle Scholar
- Hurme M, Henttinen T, Karppelin M, Varkila K, Matikainen S: Effect of interleukin-10 on NF-kB and AP-1 activities in interleukin-2 dependent CD8 T lymphoblasts. Immunol Lett. 1994, 42: 129-33. 10.1016/0165-2478(94)90075-2View ArticlePubMedGoogle Scholar
- Partridge JD, Sanguinetti G, Dibden DP, Roberts RE, Poole RK, Green J: Transition of Escherichia coli from aerobic to micro-aerobic conditions involves fast and slow reacting regulatory components. J Biol Chem. 2007, 282: 11230-7. 10.1074/jbc.M700728200View ArticlePubMedGoogle Scholar
- Sanguinetti G, Lawrence ND, Rattray M: Probabilistic inference of transcription factor concentrations and gene-specific regulatory activities. Bioinformatics. 2006, 22: 2775-81. 10.1093/bioinformatics/btl473View ArticlePubMedGoogle Scholar