FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2143991512> ?p ?o ?g. }

• W2143991512 abstract "Article19 November 2014Open Access Source Data Pathway connectivity and signaling coordination in the yeast stress-activated signaling network Deborah Chasman Deborah Chasman Department of Computer Sciences, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Yi-Hsuan Ho Yi-Hsuan Ho Laboratory of Genetics, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author David B Berry David B Berry Laboratory of Genetics, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Corey M Nemec Corey M Nemec Department of Biochemistry, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Matthew E MacGilvray Matthew E MacGilvray Laboratory of Genetics, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author James Hose James Hose Laboratory of Genetics, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Anna E Merrill Anna E Merrill Department of Chemistry, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author M Violet Lee M Violet Lee Department of Chemistry, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Jessica L Will Jessica L Will Laboratory of Genetics, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Joshua J Coon Joshua J Coon Department of Chemistry, University of Wisconsin-Madison, Madison, WI, USA Genome Center of Wisconsin, University of Wisconsin-Madison, Madison, WI, USA Department of Biological Chemistry, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Aseem Z Ansari Corresponding Author Aseem Z Ansari Department of Biochemistry, University of Wisconsin-Madison, Madison, WI, USA Genome Center of Wisconsin, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Mark Craven Corresponding Author Mark Craven Department of Computer Sciences, University of Wisconsin-Madison, Madison, WI, USA Genome Center of Wisconsin, University of Wisconsin-Madison, Madison, WI, USA Department of Biostatistics and Medical Informatics, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Audrey P Gasch Corresponding Author Audrey P Gasch Laboratory of Genetics, University of Wisconsin-Madison, Madison, WI, USA Genome Center of Wisconsin, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Deborah Chasman Deborah Chasman Department of Computer Sciences, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Yi-Hsuan Ho Yi-Hsuan Ho Laboratory of Genetics, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author David B Berry David B Berry Laboratory of Genetics, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Corey M Nemec Corey M Nemec Department of Biochemistry, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Matthew E MacGilvray Matthew E MacGilvray Laboratory of Genetics, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author James Hose James Hose Laboratory of Genetics, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Anna E Merrill Anna E Merrill Department of Chemistry, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author M Violet Lee M Violet Lee Department of Chemistry, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Jessica L Will Jessica L Will Laboratory of Genetics, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Joshua J Coon Joshua J Coon Department of Chemistry, University of Wisconsin-Madison, Madison, WI, USA Genome Center of Wisconsin, University of Wisconsin-Madison, Madison, WI, USA Department of Biological Chemistry, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Aseem Z Ansari Corresponding Author Aseem Z Ansari Department of Biochemistry, University of Wisconsin-Madison, Madison, WI, USA Genome Center of Wisconsin, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Mark Craven Corresponding Author Mark Craven Department of Computer Sciences, University of Wisconsin-Madison, Madison, WI, USA Genome Center of Wisconsin, University of Wisconsin-Madison, Madison, WI, USA Department of Biostatistics and Medical Informatics, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Audrey P Gasch Corresponding Author Audrey P Gasch Laboratory of Genetics, University of Wisconsin-Madison, Madison, WI, USA Genome Center of Wisconsin, University of Wisconsin-Madison, Madison, WI, USA Search for more papers by this author Author Information Deborah Chasman1,‡, Yi-Hsuan Ho2,‡, David B Berry2,8, Corey M Nemec3, Matthew E MacGilvray2, James Hose2, Anna E Merrill4, M Violet Lee4,9, Jessica L Will2,10, Joshua J Coon4,5,6, Aseem Z Ansari 3,5, Mark Craven 1,5,7 and Audrey P Gasch 2,5 1Department of Computer Sciences, University of Wisconsin-Madison, Madison, WI, USA 2Laboratory of Genetics, University of Wisconsin-Madison, Madison, WI, USA 3Department of Biochemistry, University of Wisconsin-Madison, Madison, WI, USA 4Department of Chemistry, University of Wisconsin-Madison, Madison, WI, USA 5Genome Center of Wisconsin, University of Wisconsin-Madison, Madison, WI, USA 6Department of Biological Chemistry, University of Wisconsin-Madison, Madison, WI, USA 7Department of Biostatistics and Medical Informatics, University of Wisconsin-Madison, Madison, WI, USA 8Present address: Institute for Neurodegenerative Disease, University of California-San Francisco, San Francisco, CA, USA 9Present address: Genentech, South San Francisco, CA, USA 10Present address: University of Georgia, Athens, GA, USA ‡These authors share first authorship *Corresponding author. Tel: +1 608 265 4690; E-mail: [email protected] *Corresponding author. Tel: +1 608 265 6181; E-mail: [email protected] *Corresponding author. Tel: +1 608 265 0859; E-mail: [email protected] Molecular Systems Biology (2014)10:759https://doi.org/10.15252/msb.20145120 PDFDownload PDF of article text and main figures. Peer ReviewDownload a summary of the editorial decision process including editorial decision letters, reviewer comments and author responses to feedback. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions Figures & Info Abstract Stressed cells coordinate a multi-faceted response spanning many levels of physiology. Yet knowledge of the complete stress-activated regulatory network as well as design principles for signal integration remains incomplete. We developed an experimental and computational approach to integrate available protein interaction data with gene fitness contributions, mutant transcriptome profiles, and phospho-proteome changes in cells responding to salt stress, to infer the salt-responsive signaling network in yeast. The inferred subnetwork presented many novel predictions by implicating new regulators, uncovering unrecognized crosstalk between known pathways, and pointing to previously unknown ‘hubs’ of signal integration. We exploited these predictions to show that Cdc14 phosphatase is a central hub in the network and that modification of RNA polymerase II coordinates induction of stress-defense genes with reduction of growth-related transcripts. We find that the orthologous human network is enriched for cancer-causing genes, underscoring the importance of the subnetwork's predictions in understanding stress biology. Synopsis An experimental and computational pipeline was developed to infer the yeast salt-activated signaling network. The resulting network provides new insights into how cells integrate upstream signals to produce a coordinated transcriptional response to stress. An integer linear programming method for integrating disparate high-throughput datasets was developed and used to infer the yeast signaling network activated by salt stress. The network shows high connectivity between what are typically considered distinct pathways. The phosphatase Cdc14 coordinates several aspects of the stress response, and RNA Pol II modification is a key regulatory point for the induction of stress-defense genes with repression of growth-related genes. The orthologous human network is enriched for cancer-related genes, underscoring the importance of stress-responsive signaling networks in human disease biology. Introduction All cells respond to stress by orchestrating complex responses customized for each situation. When grown in optimal conditions, Saccharomyces cerevisiae maintains high expression of growth-related genes and low transcription of stress-defense genes, in part via nutrient responsive TOR and RAS-regulated protein kinase A (PKA) signaling (Smets et al, 2010; Broach, 2012). Suboptimal conditions suppress these pathways in an unknown manner while activating stress-specific signaling networks that coordinate changes in transcription and translation, protein function, and metabolic fluxes with transient arrest of growth and cell cycle progression. How these disparate physiological processes are coordinated is poorly understood but likely critical for surviving and acclimating to stressful conditions. At the level of gene expression, stressed yeast activate condition-specific transcript changes that provide specialized stress defenses. These responses are typically regulated by condition-specific transcription factors (TFs) and upstream signaling pathways that are activated under limited circumstances (Hohmann & Mager, 2003). Concurrently, stressed yeast activate the common environmental stress response (ESR) (Gasch et al, 2000; Causton et al, 2001). The ESR includes ~300 induced ESR (iESR) genes that are broadly involved in stress defense and ~600 repressed ESR (rESR) genes that together encode ribosomal proteins (RPs) and proteins involved in ribosome biogenesis/protein synthesis (RiBi). While the complete set of ESR regulators remains elusive, it is clear that the program is regulated by different upstream signaling factors under different situations (Gasch et al, 2000, 2001; Gasch, 2002). Activation of the ESR, and of transcript changes more broadly, is in fact not required to survive the initial stressor, but rather is necessary for acquired resistance to subsequent stress (Berry & Gasch, 2008; Westfall et al, 2008; Mitchell et al, 2009; Berry et al, 2011). Therefore, screens for mutants sensitive to a single dose of stress have likely missed many signaling proteins, rendering stress-dependent signaling networks incomplete. Although several isolated ‘pathways’ are well characterized, how signaling is integrated through a single cellular system is poorly understood. Here, we present an experimental and computational pipeline to infer the complete sodium chloride (NaCl)-activated signaling network from a combination of data types. A key feature of our approach is that we generated several large-scale datasets (including mutant transcriptome profiles, phospho-proteome changes, and gene fitness contributions) under the same culture system in cells responding to acute NaCl stress. Because stress responses are highly context dependent (Van Wuytswinkel et al, 2000; O'Rourke & Herskowitz, 2004; Berry & Gasch, 2008), we restrict our analysis to datasets generated in our own laboratory, despite many insightful prior studies characterizing the salt response in yeast (e.g., Causton et al, 2001; Hirasawa et al, 2006; Capaldi et al, 2008; Melamed et al, 2008; Westfall et al, 2008; Halbeisen & Gerber, 2009; Soufi et al, 2009; Martinez-Montanes et al, 2010; Warringer et al, 2010; Miller et al, 2011). We wished to develop a computational method to integrate these datasets and infer the stress-activated signaling subnetwork, both to implicate missing regulators and to understand their connections. Prior approaches tackling the challenge of network inference have leveraged large-scale biological datasets, most commonly transcriptome data (see Friedman (2004) and Schadt et al (2005)). Extensions focusing on the osmotic response include the work of Gat-Viks et al, whose probabilistic method described regulatory relationships between known regulators of the Hog pathway, assuming a known network topology (Gat-Viks et al, 2006; Gat-Viks & Shamir, 2007). Several approaches leverage protein–protein and protein–nucleic acid interactions to infer relevant connections between regulators and their downstream gene targets (Liang et al, 1998; Ideker et al, 2000; Yeang et al, 2004; Yeung et al, 2004; Markowetz et al, 2005; Tu et al, 2006; Suthram et al, 2008; Huang & Fraenkel, 2009, 2012; Vaske et al, 2009; Yeger-Lotem et al, 2009; Novershtern et al, 2011). The method we present here is most closely related to methods that infer subnetworks by solving an integer linear program (IP) (Ourfali et al, 2007; Gitter et al, 2011; Silverbush et al, 2011). In particular, Gitter et al (2013) developed a combined probabilistic/IP method to discern signaling in the potassium chloride-responsive subnetwork from time series expression data (Gitter et al, 2013). However, their approach incorporated transcriptome data only, whereas we were interested in incorporating other data types. Methods that integrate disparate datasets are emerging, for example, the work of Huang et al (2013) that considered existing transcriptomic and proteomic data to study oncogene-induced signaling (Huang et al, 2013). In our case, we wanted to design a method that could also take mutant transcriptome profiles generated in our own laboratory. We therefore designed an integer linear programming (IP) approach to integrate and interpret our disparate datasets by inferring a signaling subnetwork. The novel facets of our computational approach include a means to integrate these varied data sources, using new types of input paths to the IP, and a multi-part objective function. The resulting subnetwork generated many new insights into stress signaling, by implicating new regulators, unveiling the connections between them, and presenting organization principles that shed light on stress biology. Results We previously identified 225 genes important for acquired stress resistance after NaCl pretreatment (Berry et al, 2011), including a subset of the known signaling proteins activated by NaCl (Supplementary Fig S1). Because only a fraction of NaCl-dependent transcript changes are important for acquired stress resistance, the selection misses many of the upstream transcriptome regulators. Therefore, to implicate the complete upstream signaling subnetwork, we began by profiling NaCl-dependent expression changes in 16 mutants implicated in NaCl-induced acquired stress tolerance (Fig 1, see 4). Together, this generated a matrix of regulator–gene target predictions that encompassed 3,300 genes (Supplementary Fig S2 and Table 1). A third of the affected genes were dependent on ≥ 2 regulators, and there was significant overlap in several target-gene sets (hypergeometric test, Fig 1). These results hint at the complex upstream signaling that controls the NaCl-responsive transcriptome. Table 1. Gene targets identified in regulator mutants. Mutanta Defectiveb Amplifiedb Source regulators hog1Δ (3) 1378 565 pde2Δ (3) 517 59 mck1Δ (3) 794 101 msn2Δ (3) 184 26 rim101Δ (3) 75 227 gpb2Δ (2) 202 37 rim15Δ (2) 438 106 npr2Δ (2) 75 69 npr3Δ (2) 184 89 swc3Δ (2) 108 257 swc5Δ (2) 84 55 whi2Δ (2) 118 201 pph3Δ (2) 235 21 sub1Δ (2) 431 97 tpk1Δ (2) 35 96 ygr122wΔ (2) 106 502 Validation mutants cdc14-3 (3)c 929 346 nnk1Δ (1) 94 278 bck1Δ (1) 107 169 yak1Δ (1) 226 248 kin2Δ (1) 52 266 pho85Δ (1) 614 342 cka2Δ (2) 155 63 cka1Δ (2) 58 133 ckb1Δ ckb12Δ (2) 129 176 arf3Δ (2) 466 331 scd6Δ (2) 0 0 a Mutant and number of replicates in parentheses. b Number of genes with smaller (defective) or larger (amplified) expression changes compared to the wild-type strain. Note, this table includes non-coding RNAs that were excluded from the inference. The table lists the number of targets identified from the originally interrogated ‘source’ regulators and validation mutants. c cdc14-3 was compared to its isogenic and identically treated wild-type. Figure 1. Overlapping targets of interrogated ‘source’ regulatorsThe number of genes whose osmotic response was defective in each of 16 mutants is represented by the size of each circle. Edge thickness represents the fraction of the smaller node's targets that overlap between two nodes. Edge color is proportional to significance of the overlap (hypergeometric test), where black represents a −log(P-value) of 5 or greater. Download figure Download PowerPoint Because much of signal transduction occurs post-translationally, we next measured changes to the phospho-proteome before and at 5 and 15 min after NaCl treatment, using chemical isobaric tags for phosphopeptide quantification (see 4). Nearly 600 of 1,937 identified phospho-sites (mapping to 973 proteins) showed a ≥ 2-fold change in phosphorylation, roughly split between sites with increased and decreased modification (Supplementary Fig S3). Over 10% of the altered phospho-proteins represented kinases and phosphatases (including regulators of cell cycle progression, actin organization, and signal transduction) as well as transcriptional regulators (such as activators Hot1, Sko1, and Sub1 and repressors Mot2, Dot6, and Dig1). Proteins affected at the later time point were involved in cytokinesis, bud-site selection, and actin reorganization (Bonferroni-corrected P < 0.01, hypergeometric test), implying downstream physiological effects on these processes. This analysis generated a rich source of datasets (outlined in Fig 2). To integrate and interpret these disparate datasets, we designed an integer linear programming-based (IP) approach (Fig 3 and 4). Using a background network of physical or chemical protein interactions, the method infers a subnetwork that predicts the paths by which each interrogated source regulator is connected to its downstream targets (identified as dysregulated genes in the source mutant responding to NaCl treatment). Each path is a directed, linear chain of interactions between yeast proteins, where the terminal protein node represents a sequence-specific transcription factor (TF) or RNA-binding protein (RBP) known to bind the downstream promoters or transcripts, respectively. The IP's objective function favors the inclusion of salt-responsive proteins, that is, those with differential phosphorylation or required for acquired stress fitness after NaCl treatment, and allows the sparing inclusion of additional proteins. Figure 2. Overview of the experimental data collection and analysis to generate IP inputSee text for details. Download figure Download PowerPoint Figure 3. Overview of the subnetwork inference method The input to the method includes a background network of yeast interactions combined with experimental data that describes the yeast salt stress response, including proteins with phospho-changes (yellow), fitness contribution (blue), or two known upstream regulators (pink), as described in the key. The three different types of paths that we enumerate using the background network and experimental data, where ‘hit’ refers to proteins identified in the original fitness screen or with significant changes in phosphorylation. The IP for subnetwork inference and the output ensemble of inferred subnetworks. Download figure Download PowerPoint Specifically, we start with a background network of directed and undirected intracellular interactions representing protein–protein, kinase–substrate, and gene regulatory interactions between proteins and genes/mRNAs (Guelzim et al, 2002; Ptacek et al, 2005; MacIsaac et al, 2006; Stark et al, 2006; Hogan et al, 2008; Everett et al, 2009; Pu et al, 2009; Breitkreutz et al, 2010; Scherrer et al, 2010; Tsvetanova et al, 2010; Abdulrehman et al, 2011; Fasolo et al, 2011; Sharifpoor et al, 2011; Venters et al, 2011; Heavner et al, 2012; Huebert et al, 2012). For each interrogated source regulator, we identify candidate TFs and RBPs whose known binding targets significantly overlap with the source's targets (Fig 3A). We then enumerate all possible directed candidate paths (using an iterative deepening search up to a given length) that connect each of the 16 interrogated source regulators to the majority of their targets, through candidate TFs or RBPs (Fig 3B). Other candidate paths connect proteins required for fitness (Fig 3B, blue nodes), proteins with NaCl-dependent phosphorylation changes (yellow nodes), and two known upstream sensors (pink nodes). The candidate paths serve as input to the IP, which encodes the relevance of each network element as a binary variable and characterizes possible subnetworks using a set of linear constraints over these variables (Fig 3C). Subnetwork inference is performed by choosing a union of relevant, directed paths that optimize a series of successively applied objective functions that aim to connect experimentally implicated proteins while minimally including proteins not currently supported by experimental evidence. Because many distinct subnetworks may score equally well, we use the IP to identify an ensemble of high-scoring subnetworks. In turn, each protein, interaction, and path is assigned a confidence value based on its frequency across the ensemble. Validation analysis provides strong support for the inferred subnetwork Using the datasets described above, the method identified a consensus subnetwork encompassing 380 nodes (predicted regulators) and 1,131 edges (relevant interactions) present at 75% confidence (Fig 4A). To assess the inferred subnetwork's predictive accuracy, we performed precision–recall analysis using an assembled list of known NaCl regulators and another list of unlikely regulators that included metabolic enzymes and exclusively subcellular proteins. We excluded from consideration proteins with phospho-changes or fitness contributions (since they are preferentially included by the inference) and plotted the precision and recall over varying node-confidence thresholds (Fig 4B). The inferred ensemble achieved substantially higher accuracy than the enumerated candidate paths provided as input to the IP method, highlighting the power of the inference step (Fig 4B, green line). To assess the effects of the topological properties in the background network, we ran the method on permuted source–target pairs (maintaining the degree distribution from the real data; see 4). This permuted baseline achieved high accuracy in the low-recall range, suggesting that some regulators are highly central in the background network. However, our inferred ensemble significantly outperformed the permuted baseline at higher levels of recall; thus, our method's accuracy is not simply due to properties of the background network's topology. To understand the contribution of each component of our method, we also performed additional enrichment analyses and other computational evaluations, with results available in Supplementary Information Section 2. Figure 4. Inferred NaCl-activated signaling network Inferred consensus subnetwork at 75% confidence, where node size indicates degree (number of connections) and color is according to the key. Nodes representing proteins with phospho-changes are outlined in bold. Precision–recall of the inferred consensus network was calculated using a list of true positives from the literature and a list of likely negatives, after excluding proteins with phospho-changes and those required for fitness (see 4). Precision is the fraction of predicted nodes known to be involved in the osmo response, and recall is the fraction of true positives that are above the threshold. The curves represent the performance of the IP method on the real data (blue), of the method on randomized permutations of the input network (yellow), and of the candidate enumerated pathways used as input (green, see 4). Download figure Download PowerPoint We found additional support for the inferred subnetwork in the non-random inclusion of specific protein functional groups. When compared to the background network, to the enumerated candidate pathways used as input to the IP, and to the permuted subnetworks, the inferred consensus subnetwork was enriched for proteins annotated as ‘stress’ proteins (background, P = 5e-21; candidates, P = 2e-6; permutations, P = 0.007) and for proteins encoded by genes with genetic interactions (background and candidates, P ≈ 0; permutations, P = 0.003) (Stark et al, 2006), which suggests functional dependencies. The consensus subnetwork was also slightly enriched for kinases (relative to the candidate paths and background network) and for essential genes (relative to the background network), but not relative to the permuted subnetworks (suggesting its bias toward kinases and essential genes). The inferred subnetwork included many regulators not previously linked to the NaCl response. To test some of the novel predictions, we analyzed osmo-dependent transcriptome changes in 14 mutants lacking predicted regulators, with preferences for kinases and phosphatases (see Table 1; Supplementary Fig S4 and Supplementary Information). The results provided strong support overall for the inferred subnetwork. All but one of the mutants (93%) displayed a defect in osmo-responsive expression. Furthermore, the predicted targets of 80% of these regulators overlapped significantly (P < 1e-3) with their measured targets, highlighting the accuracy of regulator–target predictions. To garner support for the subnetwork's structure, we investigated the overlap in targets of each interrogated mutant and the known or measured targets of proteins predicted to lie in the interrogated regulator's paths. Using stringent scoring, we found support for 30–100% of nodes in most paths (53% on average, Supplementary Table S1). Together, these results provide strong support for the validity of the inferred consensus subnetwork. Known and new players captured in the NaCl-responsive signaling subnetwork We therefore explored the consensus subnetwork for new insights into stress signaling. Many expected pathways were captured, including the canonical HOG, PKA, and TOR pathways. The inferred subnetwork included other stress-activated pathways not previously linked to the NaCl response, such as PKC, Pho85, Rim15 pathways, and GSK-3 kinase Mck1 (Fig 5A). We tested the involvement of these pathways by analyzing our phospho-proteomic data and mutant transcriptome profiles: We found that members of all of these pathways showed NaCl-dependent phospho-changes, and cells lacking specific pathway members (including BCK1, YAK1, PHO85, RIM15, and MCK1) had defects in NaCl-dependent expression changes (Supplementary Fig S4 and Supplementary Information). The subnetwork also included the ‘STE’ mating pathway, which shares upstream components with the Hog network and is known to be suppressed by Hog1 signaling (O'Rourke & Herskowitz, 1998; Marles et al, 2004; Zarrinpar et al, 2004; McClean et al, 2007; Shock et al, 2009; Patterson et al, 2010; Nagiec & Dohlman, 2012). The inclusion of the mating pathway indicates that some connections in the consensus subnetwork represent signaling suppression that prevents crosstalk to other pathways. We also validated several newly implicated regulators, including the CK2 kinase complex (see Supplementary Information) and the Cdc14 phosphatase (see below). Figure 5. Connectivity between known pathways and hubs of signal integration A subregion of the inferred subnetwork, highlighting proteins in known pathways according to the key. Hexagons represent interrogated ‘source’ regulators, nodes outlined in bold indicate validated players in the NaCl response, and asterisks represent proteins with phospho-changes upon NaCl treatment. Dashed edges represent physical interactions and solid arrows indicate kinase–substrate relationships. Edge directionality is as predicted by the inference, and edge color is according to the edge's source node. Inhibitory edges were taken from the literature. Connectivity between known pathways, where blue boxes represent the number of interactions between any members of two pathways. Pathway membership is indicated in parentheses. The top 15-ranked ‘integrator’ nodes with connections to the greatest number of different pathways, as shown in (B). A purified CTD peptide was incubated with Hrr25-TAP or Hog1-TAP purified from cells with and without NaCl treatment for 10 min, incubated with and without the reversible p38-specific inhibitor SB203580 (INH) added in vitro. Reactions with buffer or yeast whole-cell extract (WCE) served as negative and positive controls, respectively. CTD phosphorylated on serine 2 (Ser2) or Ser5 was detected by immunoblotting (see 4). TAP-tagged proteins were subsequently quantified on the same blot with the anti-TAP antibody. Quantification of Hog1 phosphorylation, shown to the right, was normalized to Hog1-TAP abundance and then to the corresponding unstressed sample. Download figure Download PowerPoint Interconnectivity in the inferred signaling subnetwork The structure of the subnetwork revealed surprising cross-connectivity" @default.
• W2143991512 created "2016-06-24" @default.
• W2143991512 creator A5010228563 @default.
• W2143991512 creator A5030700140 @default.
• W2143991512 creator A5033022411 @default.
• W2143991512 creator A5042325480 @default.
• W2143991512 creator A5048778946 @default.
• W2143991512 creator A5066902283 @default.
• W2143991512 creator A5068962367 @default.
• W2143991512 creator A5071722704 @default.
• W2143991512 creator A5073402912 @default.
• W2143991512 creator A5074984612 @default.
• W2143991512 creator A5076796719 @default.
• W2143991512 creator A5086840789 @default.
• W2143991512 creator A5088715947 @default.
• W2143991512 date "2014-11-01" @default.
• W2143991512 modified "2023-10-15" @default.
• W2143991512 title "Pathway connectivity and signaling coordination in the yeast stress‐activated signaling network" @default.
• W2143991512 cites W1481919380 @default.
• W2143991512 cites W1483463110 @default.
• W2143991512 cites W1514558704 @default.
• W2143991512 cites W1546324274 @default.
• W2143991512 cites W159325311 @default.
• W2143991512 cites W1644724366 @default.
• W2143991512 cites W1954289976 @default.
• W2143991512 cites W1964134459 @default.
• W2143991512 cites W1964327488 @default.
• W2143991512 cites W1966827022 @default.
• W2143991512 cites W1966918880 @default.
• W2143991512 cites W1967330678 @default.
• W2143991512 cites W1970140396 @default.
• W2143991512 cites W1971266246 @default.
• W2143991512 cites W1972129511 @default.
• W2143991512 cites W1972946410 @default.
• W2143991512 cites W1973394750 @default.
• W2143991512 cites W1975644211 @default.
• W2143991512 cites W1981930838 @default.
• W2143991512 cites W1989040724 @default.
• W2143991512 cites W1989773675 @default.
• W2143991512 cites W1998230749 @default.
• W2143991512 cites W1999084705 @default.
• W2143991512 cites W2001406822 @default.
• W2143991512 cites W2004408181 @default.
• W2143991512 cites W2006215395 @default.
• W2143991512 cites W2007757455 @default.
• W2143991512 cites W2008240036 @default.
• W2143991512 cites W2009678739 @default.
• W2143991512 cites W2012170303 @default.
• W2143991512 cites W2013947447 @default.
• W2143991512 cites W2019136808 @default.
• W2143991512 cites W2025836692 @default.
• W2143991512 cites W2034073307 @default.
• W2143991512 cites W2038372746 @default.
• W2143991512 cites W2038716712 @default.
• W2143991512 cites W2039979288 @default.
• W2143991512 cites W2041372179 @default.
• W2143991512 cites W2041889933 @default.
• W2143991512 cites W2044463066 @default.
• W2143991512 cites W2047275456 @default.
• W2143991512 cites W2050199432 @default.
• W2143991512 cites W2050357198 @default.
• W2143991512 cites W2051079338 @default.
• W2143991512 cites W2053184946 @default.
• W2143991512 cites W2053389510 @default.
• W2143991512 cites W2057264950 @default.
• W2143991512 cites W2058522624 @default.
• W2143991512 cites W2059139426 @default.
• W2143991512 cites W2060163314 @default.
• W2143991512 cites W2061624209 @default.
• W2143991512 cites W2064477073 @default.
• W2143991512 cites W2069898971 @default.
• W2143991512 cites W2079517684 @default.
• W2143991512 cites W2080232323 @default.
• W2143991512 cites W2081098333 @default.
• W2143991512 cites W2081954164 @default.
• W2143991512 cites W2085646340 @default.
• W2143991512 cites W2089691349 @default.
• W2143991512 cites W2090744112 @default.
• W2143991512 cites W2096217402 @default.
• W2143991512 cites W2096536093 @default.
• W2143991512 cites W2097127232 @default.
• W2143991512 cites W2097283846 @default.
• W2143991512 cites W2097497257 @default.
• W2143991512 cites W2102221598 @default.
• W2143991512 cites W2103218750 @default.
• W2143991512 cites W2103453943 @default.
• W2143991512 cites W2103696082 @default.
• W2143991512 cites W2104254337 @default.
• W2143991512 cites W2106357557 @default.
• W2143991512 cites W2106647330 @default.
• W2143991512 cites W2107067224 @default.
• W2143991512 cites W2111374113 @default.
• W2143991512 cites W2111862142 @default.
• W2143991512 cites W2113836312 @default.
• W2143991512 cites W2115530456 @default.
• W2143991512 cites W2115559271 @default.
• W2143991512 cites W2116052270 @default.
• W2143991512 cites W2116483876 @default.
• W2143991512 cites W2117007075 @default.
• W2143991512 cites W2117466701 @default.

• next