Topology Of Signaling Networks Research Articles

EMT, Stemness, and Drug Resistance in Biological Context: A 3D Tumor Tissue/In Silico Platform for Analysis of Combinatorial Treatment in NSCLC with Aggressive KRAS-Biomarker Signatures.

Simple SummaryThe phenotypic transition of tumor cells from epithelial to mesenchymal characteristics is called EMT and is widely discussed in the scientific community as a game changer in drug resistance and metastasis formation. However, clinical studies could not prove the efficacy of EMT-interfering treatments, and in clinical routine, EMT is not investigated to assess invasion. To fill this gap between bench and bedside, we use in this study a lung tumor tissue model with a preserved basement membrane for investigation of EMT functions with respect to invasion across this membrane and drug resistance. Our results suggest EMT is more a marker of drug resistance than a maker. Invasion is enhanced by EMT but more dependent on intrinsic factors, and EMT is not detected in the center of invasive tumor nodules. An in silico signaling network model is used to integrate these in vitro results and to reveal determinants for drug response.Epithelial-to-mesenchymal transition (EMT) is discussed to be centrally involved in invasion, stemness, and drug resistance. Experimental models to evaluate this process in its biological complexity are limited. To shed light on EMT impact and test drug response more reliably, we use a lung tumor test system based on a decellularized intestinal matrix showing more in vivo-like proliferation levels and enhanced expression of clinical markers and carcinogenesis-related genes. In our models, we found evidence for a correlation of EMT with drug resistance in primary and secondary resistant cells harboring KRASG12C or EGFR mutations, which was simulated in silico based on an optimized signaling network topology. Notably, drug resistance did not correlate with EMT status in KRAS-mutated patient-derived xenograft (PDX) cell lines, and drug efficacy was not affected by EMT induction via TGF-β. To investigate further determinants of drug response, we tested several drugs in combination with a KRASG12C inhibitor in KRASG12C mutant HCC44 models, which, besides EMT, display mutations in P53, LKB1, KEAP1, and high c-MYC expression. We identified an aurora-kinase A (AURKA) inhibitor as the most promising candidate. In our network, AURKA is a centrally linked hub to EMT, proliferation, apoptosis, LKB1, and c-MYC. This exemplifies our systemic analysis approach for clinical translation of biomarker signatures.

Oncogenic β-catenin and PIK3CA instruct network states and cancer phenotypes in intestinal organoids.

Colorectal cancer is driven by cooperating oncogenic mutations. In this study, we use organotypic cultures derived from transgenic mice inducibly expressing oncogenic β-catenin and/or PIK3CAH1047R to follow sequential changes in cancer-related signaling networks, intestinal cell metabolism, and physiology in a three-dimensional environment mimicking tissue architecture. Activation of β-catenin alone results in the formation of highly clonogenic cells that are nonmotile and prone to undergo apoptosis. In contrast, coexpression of stabilized β-catenin and PIK3CAH1047R gives rise to intestinal cells that are apoptosis-resistant, proliferative, stem cell-like, and motile. Systematic inhibitor treatments of organoids followed by quantitative phenotyping and phosphoprotein analyses uncover key changes in the signaling network topology of intestinal cells after induction of stabilized β-catenin and PIK3CAH1047R We find that survival and motility of organoid cells are associated with 4EBP1 and AKT phosphorylation, respectively. Our work defines phenotypes, signaling network states, and vulnerabilities of transgenic intestinal organoids as a novel approach to understanding oncogene activities and guiding the development of targeted therapies.

Identifying the topology of signaling networks from partial RNAi data.

BackgroundMethods for inferring signaling networks using single gene knockdown RNAi experiments and reference networks have been proposed in recent years. These methods assume that RNAi information is available for all the genes in the signal transduction pathway, i.e., complete. This assumption does not always hold up since RNAi experiments are often incomplete and information for some genes is missing.ResultsIn this article, we develop two methods to construct signaling networks from incomplete RNAi data with the help of a reference network. These methods infer the RNAi constraints for the missing genes such that the inferred network is closest to the reference network. We perform extensive experiments with both real and synthetic networks and demonstrate that these methods produce accurate results efficiently.ConclusionsApplication of our methods to Wnt signal transduction pathway has shown that our methods can be used to construct highly accurate signaling networks from experimental data in less than 100 ms. The two methods that produce accurate results efficiently show great promise of constructing real signaling networks.

Unraveling Kinase Activation Dynamics Using Kinase-Substrate Relationships from Temporal Large-Scale Phosphoproteomics Studies.

In response to stimuli, biological processes are tightly controlled by dynamic cellular signaling mechanisms. Reversible protein phosphorylation occurs on rapid time-scales (milliseconds to seconds), making it an ideal carrier of these signals. Advances in mass spectrometry-based proteomics have led to the identification of many tens of thousands of phosphorylation sites, yet for the majority of these the kinase is unknown and the underlying network topology of signaling networks therefore remains obscured. Identifying kinase substrate relationships (KSRs) is therefore an important goal in cell signaling research. Existing consensus sequence motif based prediction algorithms do not consider the biological context of KSRs, and are therefore insensitive to many other mechanisms guiding kinase-substrate recognition in cellular contexts. Here, we use temporal information to identify biologically relevant KSRs from Large-scale In Vivo Experiments (KSR-LIVE) in a data-dependent and automated fashion. First, we used available phosphorylation databases to construct a repository of existing experimentally-predicted KSRs. For each kinase in this database, we used time-resolved phosphoproteomics data to examine how its substrates changed in phosphorylation over time. Although substrates for a particular kinase clustered together, they often exhibited a different temporal pattern to the phosphorylation of the kinase. Therefore, although phosphorylation regulates kinase activity, our findings imply that substrate phosphorylation likely serve as a better proxy for kinase activity than kinase phosphorylation. KSR-LIVE can thereby infer which kinases are regulated within a biological context. Moreover, KSR-LIVE can also be used to automatically generate positive training sets for the subsequent prediction of novel KSRs using machine learning approaches. We demonstrate that this approach can distinguish between Akt and Rps6kb1, two kinases that share the same linear consensus motif, and provide evidence suggesting IRS-1 S265 as a novel Akt site. KSR-LIVE is an open-access algorithm that allows users to dissect phosphorylation signaling within a specific biological context, with the potential to be included in the standard analysis workflow for studying temporal high-throughput signal transduction data.

Molecular networks in skeletal muscle plasticity.

The skeletal muscle phenotype is subject to considerable malleability depending on use as well as internal and external cues. In humans, low-load endurance-type exercise leads to qualitative changes of muscle tissue characterized by an increase in structures supporting oxygen delivery and consumption, such as capillaries and mitochondria. High-load strength-type exercise leads to growth of muscle fibers dominated by an increase in contractile proteins. In endurance exercise, stress-induced signaling leads to transcriptional upregulation of genes, with Ca(2+) signaling and the energy status of the muscle cells sensed through AMPK being major input determinants. Several interrelated signaling pathways converge on the transcriptional co-activator PGC-1α, perceived to be the coordinator of much of the transcriptional and post-transcriptional processes. Strength training is dominated by a translational upregulation controlled by mTORC1. mTORC1 is mainly regulated by an insulin- and/or growth-factor-dependent signaling cascade as well as mechanical and nutritional cues. Muscle growth is further supported by DNA recruitment through activation and incorporation of satellite cells. In addition, there are several negative regulators of muscle mass. We currently have a good descriptive understanding of the molecular mechanisms controlling the muscle phenotype. The topology of signaling networks seems highly conserved among species, with the signaling outcome being dependent on the particular way individual species make use of the options offered by the multi-nodal networks. As a consequence, muscle structural and functional modifications can be achieved by an almost unlimited combination of inputs and downstream signaling events.

A Divide and Conquer Approach for Construction of Large-Scale Signaling Networks from PPI and RNAi Data Using Linear Programming

Inference of topology of signaling networks from perturbation experiments is a challenging problem. Recently, the inference problem has been formulated as a reference network editing problem and it has been shown that finding the minimum number of edit operations on a reference network to comply with perturbation experiments is an NP-complete problem. In this paper, we propose an integer linear optimization (ILP) model for reconstruction of signaling networks from RNAi data and a reference network. The ILP model guarantees the optimal solution; however, is practical only for small signaling networks of size 10-15 genes due to computational complexity. To scale for large signaling networks, we propose a divide and conquer-based heuristic, in which a given reference network is divided into smaller subnetworks that are solved separately and the solutions are merged together to form the solution for the large network. We validate our proposed approach on real and synthetic data sets, and comparison with the state of the art shows that our proposed approach is able to scale better for large networks while attaining similar or better biological accuracy.

A versatile mass spectrometry-based method to both identify kinase client-relationships and characterize signaling network topology.

While more than a thousand protein kinases (PK) have been identified in the Arabidopsis thaliana genome, relatively little progress has been made toward identifying their individual client proteins. Herein we describe the use of a mass spectrometry-based in vitro phosphorylation strategy, termed Kinase Client assay (KiC assay), to study a targeted-aspect of signaling. A synthetic peptide library comprising 377 in vivo phosphorylation sequences from developing seed was screened using 71 recombinant A. thaliana PK. Among the initial results, we identified 23 proteins as putative clients of 17 PK. In one instance protein phosphatase inhibitor-2 (AtPPI-2) was phosphorylated at multiple-sites by three distinct PK, casein kinase1-like 10, AME3, and a Ser PK-like protein. To confirm this result, full-length recombinant AtPPI-2 was reconstituted with each of these PK. The results confirmed multiple distinct phosphorylation sites within this protein. Biochemical analyses indicate that AtPPI-2 inhibits type 1 protein phosphatase (TOPP) activity, and that the phosphorylated forms of AtPPI-2 are more potent inhibitors. Structural modeling revealed that phosphorylation of AtPPI-2 induces conformational changes that modulate TOPP binding.

Bayesian inference of signaling network topology in a cancer cell line.

Protein signaling networks play a key role in cellular function, and their dysregulation is central to many diseases, including cancer. To shed light on signaling network topology in specific contexts, such as cancer, requires interrogation of multiple proteins through time and statistical approaches to make inferences regarding network structure. In this study, we use dynamic Bayesian networks to make inferences regarding network structure and thereby generate testable hypotheses. We incorporate existing biology using informative network priors, weighted objectively by an empirical Bayes approach, and exploit a connection between variable selection and network inference to enable exact calculation of posterior probabilities of interest. The approach is computationally efficient and essentially free of user-set tuning parameters. Results on data where the true, underlying network is known place the approach favorably relative to existing approaches. We apply these methods to reverse-phase protein array time-course data from a breast cancer cell line (MDA-MB-468) to predict signaling links that we independently validate using targeted inhibition. The methods proposed offer a general approach by which to elucidate molecular networks specific to biological context, including, but not limited to, human cancers. http://mukherjeelab.nki.nl/DBN (code and data).

A Profile of 648 Signaling Network Events Identifies Cell Subsets with Diverse, Abnormal Responses to Lymphocyte Stimuli within Follicular Lymphoma Tumors.

Signal transduction plays a key role in cell survival, and changes to signaling are frequently implicated in tumor initiation and progression. We sought to identify abnormal variation in signaling network activity within primary tumor samples obtained prior to treatment from patients with follicular lymphoma (FL). We previously showed that altered B cell receptor (BCR) signaling distinguishes tumor B cells from the non-malignant host B cells in FL tumors. Here we extend this approach and use flow cytometry to measure 648 signaling events in live lymphoid cells from more than 25 lymphoma specimens and healthy controls. We combined 9 previously identified BCR stimulation conditions with inputs from CD40, interleukin 4, interferons (IFNs), and more than 10 other environmental cues that govern the development and activity of lymphocytes. Fluorescent cell barcoding allowed simultaneous staining and analysis of phospho-protein activation under all 27 stimulation conditions within a single tube. The activation of key phospho-protein nodes throughout lymphocyte signaling networks, including Syk, Erk1/2, Btk, Src family kinases, cCbl, p38, NFkB, Akt, Stat1, Stat3, Stat6, and Stat5, was measured under each of the 27 stimulation conditions. Measurements of phospho-protein responses to stimulation were combined with detection of the Bcl-2 oncogene, B and T cell lineage markers in each cell. This panel allowed us to characterize signaling in the heterogeneous cell subsets found within each patient's tumor sample. Tumor B cells, host tumor infiltrating T cells, non-malignant B cells were all distinguished by contrasting signaling profiles. In some cases, subsets of tumor B cells with differences in signaling network topology were observed within the tumor B cell population. This result suggests that signaling can distinguish between tumor sub-clones and could be used to measure tumor heterogeneity. As previously reported, little variation in signaling was observed among healthy peripheral blood B and T cell samples from different individuals. Abnormally low host T cell signaling was commonly observed within the tumor infiltrating T cells infiltrating FL tumors. Further analysis of tumor T cell subsets indicated that a high proportion of infiltrating T cells expressed CD4 and FoxP3. Taken together, these results support the hypothesis that FL tumor B cells promote suppressed signaling in the T cells of the patient and may modulate the immune response against the tumor. In FL tumor B cells, BCR and IFN signaling frequently triggered Stat5 phosphorylation, but not Stat1 phosphorylation. These results are consistent with the hypothesis that Stat5 initiates genetic programs that support cancer cell survival and proliferation, whereas Stat1 promotes immunogenicity and cooperates with the p53 tumor suppressor protein. In contrast with healthy B cells, loss of the response to CD40L, altered PKC signaling, and variable responses to BCR crosslinking were all seen in FL tumor B cells. The patterns of abnormal signaling we observed in tumor B cells and tumor infiltrating T cells suggest that measuring the activity of key signaling network nodes can identify targets for therapeutic attention in FL.

State-based discovery: a multidimensional screen for small-molecule modulators of EGF signaling

As an alternative to conventional, target-oriented drug discovery, we report a strategy that identifies compounds on the basis of the state that they induce in a signaling network. Immortalized human cells are grown in microtiter plates and treated with compounds from a small-molecule library. The target network is then activated and lysates derived from each sample are arrayed onto glass-supported nitrocellulose pads. By probing these microarrays with antibodies that report on the abundance or phosphorylation state of selected proteins, a global picture of the target network is obtained. As proof of concept, we screened 84 kinase and phosphatase inhibitors for their ability to induce different states in the ErbB signaling network. We observed functional connections between proteins that match our understanding of ErbB signaling, indicating that state-based screens can be used to define the topology of signaling networks. Additionally, compounds sort according to the multidimensional phenotypes they induce, suggesting that state-based screens may inform efforts to identify the targets of biologically active small molecules.

Dynamic stability of signal transduction networks depending on downstream and upstream specificity of protein kinases.

Mathematical methods are used for explaining the structural design of signal transduction networks, e.g. MAP kinase cascades, which control cell proliferation, differentiation or apoptosis. Taking into account protein kinases and phosphatases the interrelation between the topology of signaling networks and the stability of their ground state are analysed. It is shown that the stability is closely related to the system's dimension and to the number of cycles within the network. Systems with a higher number of kinases and/or cycles tend to be more unstable. In contrast to that increasing phosphatase activity stabilises the ground state.