Community detection in empirical kinase networks identifies new potential members of signalling pathways.

Phosphoproteomics allows one to measure the activity of kinases that drive the fluxes of signal transduction pathways involved in biological processes such as immune function, senescence and cell growth. However, deriving knowledge of signalling network circuitry from these data is challenging due to a scarcity of phosphorylation sites that define kinase-kinase relationships. To address this issue, we previously identified around 6,000 phosphorylation sites as markers of kinase-kinase relationships (that may be conceptualised as network edges), from which empirical cell-model-specific weighted kinase networks may be reconstructed. Here, we assess whether the application of community detection algorithms to such networks can identify new components linked to canonical signalling pathways. Phosphoproteomics data from acute myeloid leukaemia (AML) cells treated separately with PI3K, AKT, MEK and ERK inhibitors were used to reconstruct individual kinase networks. We used modularity maximisation to detect communities in each network, and selected the community containing the main target of the inhibitor used to treat cells. These analyses returned communities that contained known canonical signalling components. Interestingly, in addition to canonical PI3K/AKT/mTOR members, the community assignments returned TTK (also known as MPS1) as a likely component of PI3K/AKT/mTOR signalling. We drew similar insights from an external phosphoproteomics dataset from breast cancer cells treated with rapamycin and oestrogen. We confirmed this observation with wet-lab laboratory experiments showing that TTK phosphorylation was decreased in AML cells treated with AKT and MTOR inhibitors. This study illustrates the application of community detection algorithms to the analysis of empirical kinase networks to uncover new members linked to canonical signalling pathways.

Driving role of head and neck cancer cell secretome on the invasion of stromal fibroblasts: Mechanistic insights by phosphoproteomics.

BACKGROUND: Cancer-associated fibroblasts (CAFs) are major players in tumor-stroma communication, and participate in several cancer hallmarks to drive tumor progression and metastatic dissemination. This study investigates the driving effects of tumor-secreted factors on CAF biology, with the ultimate goal of identifying effective therapeutic targets/strategies for head and neck squamous cell carcinomas (HNSCC). METHODS: Functionally, conditioned media (CM) from different HNSCC-derived cell lines and normal keratinocytes (Kc) were tested on the growth and invasion of populations of primary CAFs and normal fibroblasts (NFs) using 3D invasion assays in collagen matrices. The changes in MMPs expression were evaluated by RT-qPCR and kinase enrichment was analyzed using mass spectrometry phosphoproteomics. RESULTS: Our results consistently demonstrate that HNSCC-secreted factors (but not Kc CM) specifically and robustly promoted pro-invasive properties in both CAFs and NFs, thereby reflecting the plasticity of fibroblast subtypes. Concomitantly, HNSCC-secreted factors massively increased metalloproteinases levels in CAFs and NFs. By contrast, HNSCC CM and Kc CM exhibited comparable growth-promoting effects on stromal fibroblasts. Mechanistically, phosphoproteomic analysis predominantly revealed phosphorylation changes in fibroblasts upon treatment with HNSCC CM, and various promising kinases were identified: MKK7, MKK4, ASK1, RAF1, BRAF, ARAF, COT, PDK1, RSK2 and AKT1. Interestingly, pharmacologic inhibition of RAF1/BRAF using sorafenib emerged as the most effective drug to block tumor-promoted fibroblast invasion without affecting fibroblast viability CONCLUSIONS: Our findings demonstrate that HNSCC-secreted factors specifically fine tune the invasive potential of stromal fibroblasts, thereby generating tumor-driven pro-invasive niches, which in turn to ultimately facilitate cancer cell dissemination. Furthermore, the RAF/BRAF inhibitor sorafenib was identified as a promising candidate to effectively target the onset of pro-invasive clusters of stromal fibroblasts in the HNSCC microenvironment.

eEF2K Activity Determines Synergy to Cotreatment of Cancer Cells With PI3K and MEK Inhibitors.

PI3K-mammalian target of rapamycin and MAPK/ERK kinase (MEK)/mitogen-activated protein kinase (MAPK) are the most frequently dysregulated signaling pathways in cancer. A problem that limits the success of therapies that target individual PI3K-MAPK members is that these pathways converge to regulate downstream functions and often compensate each other, leading to drug resistance and transient responses to therapy. In order to overcome resistance, therapies based on cotreatments with PI3K/AKT and MEK/MAPK inhibitors are now being investigated in clinical trials, but the mechanisms of sensitivity to cotreatment are not fully understood. Using LC-MS/MS-based phosphoproteomics, we found that eukaryotic elongation factor 2 kinase (eEF2K), a key convergence point downstream of MAPK and PI3K pathways, mediates synergism to cotreatment with trametinib plus pictilisib (which target MEK1/2 and PI3Kα/δ, respectively). Inhibition of eEF2K by siRNA or with a small molecule inhibitor reversed the antiproliferative effects of the cotreatment with PI3K plus MEK inhibitors in a cell model-specific manner. Systematic analysis in 12 acute myeloid leukemia cell lines revealed that eEF2K activity was increased in cells for which PI3K plus MEKi cotreatment is synergistic, while PKC potentially mediated resistance to such cotreatment. Together, our study uncovers eEF2K activity as a key mediator of responses to PI3Ki plus MEKi and as a potential biomarker to predict synergy to cotreatment in cancer cells.

Targeting the lysine-specific demethylase 1 rewires kinase networks and primes leukemia cells for kinase inhibitor treatment.

Most tumor types either fail to respond or become resistant to kinase inhibitors, often because of compensatory prosurvival pathways in the cancer cell's broader signaling circuitry. Here, we found that intrinsic resistance to kinase inhibitors in cultured primary acute myeloid leukemia (AML) cells may be overcome by reshaping kinase networks into topologies that confer drug sensitivity. We identified several antagonists of chromatin-modifying enzymes that sensitized AML cell lines to kinase inhibitors. Of these, we confirmed that inhibitors of the lysine-specific demethylase (LSD1; also known as KDM1A) rewired kinase signaling in AML cells in a way that increased the activity of the kinase MEK and that broadly suppressed the activity of other kinases and feedback loops. As a result, AML cell lines and about half of primary human AML samples were primed for sensitivity to the MEK inhibitor trametinib. Primary human cells with KRAS mutations and those with high MEK pathway activity were the best responders to sequential treatment with LSD1 inhibitors then trametinib, whereas those with NRAS mutations and high mTOR activity were poor responders. Overall, our study reveals the MEK pathway as a mechanism of resistance to LSD1 inhibitors in AML and shows a way to modulate kinase network circuitry to potentially overcome therapeutic resistance to kinase inhibitors.

Implementation of Clinical Phosphoproteomics and Proteomics for Personalized Medicine.

The identification of biomarkers for companion diagnostics is revolutionizing the development of treatments tailored to individual patients in different disease areas including cancer. Precision medicine is most frequently based on the detection of genomic markers that correlate with the efficacy of selected targeted therapies. However, since nongenetic mechanisms also contribute to disease biology, there is a considerable interest of using proteomic techniques as additional source of biomarkers to personalize therapies. In this chapter, we describe label-free mass spectrometry methods for proteomic and phosphoproteomic analysis compatible with routine analysis of clinical samples. We also outline bioinformatic pipelines based on statistical learning that use these proteomics datasets as input to quantify kinase activities and predict drug responses in cancer cells.

Drug ranking using machine learning systematically predicts the efficacy of anti-cancer drugs.

Artificial intelligence and machine learning (ML) promise to transform cancer therapies by accurately predicting the most appropriate therapies to treat individual patients. Here, we present an approach, named Drug Ranking Using ML (DRUML), which uses omics data to produce ordered lists of >400 drugs based on their anti-proliferative efficacy in cancer cells. To reduce noise and increase predictive robustness, instead of individual features, DRUML uses internally normalized distance metrics of drug response as features for ML model generation. DRUML is trained using in-house proteomics and phosphoproteomics data derived from 48 cell lines, and it is verified with data comprised of 53 cellular models from 12 independent laboratories. We show that DRUML predicts drug responses in independent verification datasets with low error (mean squared error < 0.1 and mean Spearman's rank 0.7). In addition, we demonstrate that DRUML predictions of cytarabine sensitivity in clinical leukemia samples are prognostic of patient survival (Log rank p < 0.005). Our results indicate that DRUML accurately ranks anti-cancer drugs by their efficacy across a wide range of pathologies.

PARP-1 activation after oxidative insult promotes energy stress-dependent phosphorylation of YAP1 and reduces cell viability.

Poly(ADP-ribose) polymerase 1 (PARP-1) is a nuclear enzyme that catalyze the transfer of ADP-ribose units from NAD+ to several target proteins involved in cellular stress responses. Using WRL68 (HeLa derivate) cells, we previously showed that PARP-1 activation induced by oxidative stress after H2O2 treatment lead to depletion of cellular NAD+ and ATP, which promoted cell death. In this work, LC-MS/MS-based phosphoproteomics in WRL68 cells showed that the oxidative damage induced by H2O2 increased the phosphorylation of YAP1, a transcriptional co-activator involved in cell survival, and modified the phosphorylation of other proteins involved in transcription. Genetic or pharmacological inhibition of PARP-1 in H2O2-treated cells reduced YAP1 phosphorylation and degradation and increased cell viability. YAP1 silencing abrogated the protective effect of PARP-1 inhibition, indicating that YAP1 is important for the survival of WRL68 cells exposed to oxidative damage. Supplementation of NAD+ also reduced YAP1 phosphorylation, suggesting that the loss of cellular NAD+ caused by PARP-1 activation after oxidative treatment is responsible for the phosphorylation of YAP1. Finally, PARP-1 silencing after oxidative treatment diminished the activation of the metabolic sensor AMPK. Since NAD+ supplementation reduced the phosphorylation of some AMPK substrates, we hypothesized that the loss of cellular NAD+ after PARP-1 activation may induce an energy stress that activates AMPK. In summary, we showed a new crucial role of PARP-1 in the response to oxidative stress in which PARP-1 activation reduced cell viability by promoting the phosphorylation and degradation of YAP1 through a mechanism that involves the depletion of NAD+.

Prediction of Signed Protein Kinase Regulatory Circuits.

Complex networks of regulatory relationships between protein kinases comprise a major component of intracellular signaling. Although many kinase-kinase regulatory relationships have been described in detail, these tend to be limited to well-studied kinases whereas the majority of possible relationships remains unexplored. Here, we implement a data-driven, supervised machine learning method to predict human kinase-kinase regulatory relationships and whether they have activating or inhibiting effects. We incorporate high-throughput data, kinase specificity profiles, and structural information to produce our predictions. The results successfully recapitulate previously annotated regulatory relationships and can reconstruct known signaling pathways from the ground up. The full network of predictions is relatively sparse, with the vast majority of relationships assigned low probabilities. However, it nevertheless suggests denser modes of inter-kinase regulation than normally considered in intracellular signaling research. A record of this paper's transparent peer review process is included in the Supplemental Information.

Cancer Burden Is Controlled by Mural Cell-ß3-Integrin Regulated Crosstalk with Tumor Cells.

Enhanced blood vessel (BV) formation is thought to drive tumor growth through elevated nutrient delivery. However, this observation has overlooked potential roles for mural cells in directly affecting tumor growth independent of BV function. Here we provide clinical data correlating high percentages of mural-ß3-integrin-negative tumor BVs with increased tumor sizes but no effect on BV numbers. Mural-ß3-integrin loss also enhances tumor growth in implanted and autochthonous mouse tumor models with no detectable effects on BV numbers or function. At a molecular level, mural-cell ß3-integrin loss enhances signaling via FAK-p-HGFR-p-Akt-p-p65, driving CXCL1, CCL2, and TIMP-1 production. In particular, mural-cell-derived CCL2 stimulates tumor cell MEK1-ERK1/2-ROCK2-dependent signaling and enhances tumor cell survival and tumor growth. Overall, our data indicate that mural cells can control tumor growth via paracrine signals regulated by ß3-integrin, providing a previously unrecognized mechanism of cancer growth control.

Reconstructing kinase network topologies from phosphoproteomics data reveals cancer-associated rewiring.

Understanding how oncogenic mutations rewire regulatory-protein networks is important for rationalizing the mechanisms of oncogenesis and for individualizing anticancer treatments. We report a chemical phosphoproteomics method to elucidate the topology of kinase-signaling networks in mammalian cells. We identified >6,000 protein phosphorylation sites that can be used to infer >1,500 kinase-kinase interactions and devised algorithms that can reconstruct kinase network topologies from these phosphoproteomics data. Application of our methods to primary acute myeloid leukemia and breast cancer tumors quantified the relationship between kinase expression and activity, and enabled the identification of hitherto unknown kinase network topologies associated with drug-resistant phenotypes or specific genetic mutations. Using orthogonal methods we validated that PIK3CA wild-type cells adopt MAPK-dependent circuitries in breast cancer cells and that the kinase TTK is important in acute myeloid leukemia. Our phosphoproteomic signatures of network circuitry can identify kinase topologies associated with both phenotypes and genotypes of cancer cells.

Restrained Phosphatidylcholine Synthesis in a Cellular Model of Down's Syndrome is Associated with the Overexpression of Dyrk1A.

Aberrant formation of the cerebral cortex could be attributed to the lack of suitable substrates that direct the migration of neurons. Previous work carried out at our laboratory has shown that oleic acid is a neurotrophic factor. In order to characterize the effect of oleic acid in a cellular model of Down's syndrome (DS), here, we used immortalized cell lines derived from the cortex of trisomy Ts16 and euploid mice. We report that in the plasma membrane of euploid cells, an increase in phosphatidylcholine concentrations occurs in the presence of oleic acid. However, in trisomic cells, oleic acid failed to increase phosphatidylcholine incorporation into the plasma membrane. Gene expression analysis of trisomic cells revealed that the phosphatidylcholine biosynthetic pathway was deregulated. Taken together, these results suggest that the overdose of specific genes in trisomic lines delays differentiation in the presence of oleic acid. The dual-specificity tyrosine (Y) phosphorylation-regulated kinase 1A (DYRK1A) gene is located on human chromosome 21. DYRK1A contributes to intellectual disability and the early onset of Alzheimer's disease in DS patients. Here, we explored the potential role of Dyrk1A in the reduction of phosphatidylcholine concentrations in trisomic cells in the presence of oleic acid. The downregulation of Dyrk1A by small interfering RNA (siRNA) in trisomic cells returned phosphatidylcholine concentrations up to similar levels to those of euploid cells in the presence of oleic acid. Thus, our results highlight the role of Dyrk1A in brain development through the modulation of phosphatidylcholine location, levels and synthesis.

Impact of phosphoproteomics in the translation of kinase-targeted therapies.

Signaling pathways driven by protein and lipid kinases are altered in most human diseases. Therefore, pharmacological inhibitors of cell signaling are one of the most intensively pursued therapeutic approaches for the treatment of diseases such as cancer, neurodegeneration, and metabolic syndromes. Phosphoproteomics is a technique that measures the products of kinase activities and, with the appropriate bioinformatics techniques, the methodology can also provide measures of kinase pathway activation and network circuitry. Hence, due to recent technological advantages, LC-MS-based quantitative phosphoproteomics provides relevant information for the design and implementation of kinase inhibitor based therapies. Here, we review how phosphoproteome profiling is being used in translational research as a means to identify drug targets and biomarkers for personalizing therapies based on kinase inhibitors.

Overexpression of DYRK1A inhibits choline acetyltransferase induction by oleic acid in cellular models of Down syndrome.

Histological brain studies of individuals with DS have revealed an aberrant formation of the cerebral cortex. Previous work from our laboratory has shown that oleic acid acts as a neurotrophic factor and induces neuronal differentiation. In order to characterize the effects of oleic acid in a cellular model of DS, immortalized cell lines derived from the cortex of trisomy Ts16 (CTb) and normal mice (CNh) were incubated in the absence or presence of oleic acid. Oleic acid increased choline acetyltransferase expression (ChAT), a marker of cholinergic differentiation in CNh cells. However, in trisomic cells (CTb line) oleic acid failed to increase ChAT expression. These results suggest that the overdose of specific genes in trisomic lines delays differentiation in the presence of oleic acid by inhibiting acetylcholine production mediated by ChAT. The dual-specificity tyrosine (Y) phosphorylation-regulated kinase 1A (DYRK1A) gene is located on human chromosome 21 and encodes a proline-directed protein kinase. It has been proposed that DYRK1A plays a prominent role in several biological functions, leading to mental retardation in DS patients. Here we explored the potential role of DYRK1A in the modulation of ChAT expression in trisomic cells and in the signaling pathways of oleic acid. Down-regulation of DYRK1A by siRNA in trisomic CTb cells rescued ChAT expression up to levels similar to those of normal cells in the presence of oleic acid. In agreement with these results, oleic acid was unable to increase ChAT expression in neuronal cultures of transgenic mice overexpressing DYRK1A. In summary, our results highlight the role played by DYRK1A in brain development through the control of ChAT expression. In addition, the overexpression of DYRK1A in DS models prevented the neurotrophic effect of oleic acid, a fact that may account for mental retardation in DS patients.