Martin Klammer

Phosphosignature Predicts Dasatinib Response in Non-small Cell Lung Cancer

Targeted drugs are less toxic than traditional chemotherapeutic therapies; however, the proportion of patients that benefit from these drugs is often smaller. A marker that confidently predicts patient response to a specific therapy would allow an individual therapy selection most likely to benefit the patient. Here, we used quantitative mass spectrometry to globally profile the basal phosphoproteome of a panel of non-small cell lung cancer cell lines. The effect of the kinase inhibitor dasatinib on cellular growth was tested against the same panel. From the phosphoproteome profiles, we identified 58 phosphorylation sites, which consistently differ between sensitive and resistant cell lines. Many of the corresponding proteins are involved in cell adhesion and cytoskeleton organization. We showed that a signature of only 12 phosphorylation sites is sufficient to accurately predict dasatinib sensitivity. Four of the phosphorylation sites belong to integrin β4, a protein that mediates cell-matrix or cell-cell adhesion. The signature was validated in cross-validation and label switch experiments and in six independently profiled breast cancer cell lines. The study supports that the phosphorylation of integrin β4, as well as eight further proteins comprising the signature, are candidate biomarkers for predicting response to dasatinib in solid tumors. Furthermore, our results show that identifying predictive phosphorylation signatures from global, quantitative phosphoproteomic data is possible and can open a new path to discovering molecular markers for response prediction.

Identifying differentially regulated subnetworks from phosphoproteomic data

BackgroundVarious high throughput methods are available for detecting regulations at the level of transcription, translation or posttranslation (e.g. phosphorylation). Integrating these data with protein networks should make it possible to identify subnetworks that are significantly regulated. Furthermore, such integration can support identification of regulated entities from often noisy high throughput data. In particular, processing mass spectrometry-based phosphoproteomic data in this manner may expose signal transduction pathways and, in the case of experiments with drug-treated cells, reveal the drugs mode of action.ResultsHere, we introduce SubExtractor, an algorithm that combines phosphoproteomic data with protein network information from STRING to identify differentially regulated subnetworks and individual proteins. The method is based on a Bayesian probabilistic model combined with a genetic algorithm and rigorous significance testing. The Bayesian model accounts for information about both differential regulation and network topology. The method was tested with artificial data and subsequently applied to a comprehensive phosphoproteomics study investigating the mode of action of sorafenib, a small molecule kinase inhibitor.ConclusionsSubExtractor reliably identifies differentially regulated subnetworks from phosphoproteomic data by integrating protein networks. The method can also be applied to gene or protein expression data.

MetaTM - a consensus method for transmembrane protein topology prediction

BackgroundTransmembrane (TM) proteins are proteins that span a biological membrane one or more times. As their 3-D structures are hard to determine, experiments focus on identifying their topology (i. e. which parts of the amino acid sequence are buried in the membrane and which are located on either side of the membrane), but only a few topologies are known. Consequently, various computational TM topology predictors have been developed, but their accuracies are far from perfect. The prediction quality can be improved by applying a consensus approach, which combines results of several predictors to yield a more reliable result.ResultsA novel TM consensus method, named MetaTM, is proposed in this work. MetaTM is based on support vector machine models and combines the results of six TM topology predictors and two signal peptide predictors. On a large data set comprising 1460 sequences of TM proteins with known topologies and 2362 globular protein sequences it correctly predicts 86.7% of all topologies.ConclusionCombining several TM predictors in a consensus prediction framework improves overall accuracy compared to any of the individual methods. Our proposed SVM-based system also has higher accuracy than a previous consensus predictor. MetaTM is made available both as downloadable source code and as DAS server at http://MetaTM.sbc.su.se

Global Quantitative Phosphoproteome Analysis of Human Tumor Xenografts Treated with a CD44 Antagonist

The cell surface glycoprotein CD44 plays an important role in the development and progression of various tumor types. RG7356 is a humanized antibody targeting the constant region of CD44 that shows antitumor efficacy in mice implanted with CD44-expressing tumors such as MDA-MB-231 breast cancer cells. CD44 receptor seems to function as the main receptor for hyaluronic acid and osteopontin, serving as coreceptor for growth factor pathways like cMet, EGFR, HER-2, and VEGFR and by cytoskeletal modulation via ERM and Rho kinase signaling. To assess the direct impact of RG7356 binding to the CD44 receptor, a global mass spectrometry-based phosphoproteomics approach was applied to freshly isolated MDA-MB-231 tumor xenografts. Results from a global phosphoproteomics screen were further corroborated by Western blot and ELISA analyses of tumor lysates from CD44-expressing tumors. Short-term treatment of tumor-bearing mice with RG7356 resulted in modifications of the MAPK pathway in the responsive model, although no effects on downstream phosphorylation were observed in a nonresponsive xenograft model. Taken together, our approach augments the value of other high throughput techniques to identify biomarkers for clinical development of targeted agents.

jSquid: a Java applet for graphical on-line network exploration

UNLABELLED jSquid is a graph visualization tool for exploring graphs from protein-protein interaction or functional coupling networks. The tool was designed for the FunCoup web site, but can be used for any similar network exploring purpose. The program offers various visualization and graph manipulation techniques to increase the utility for the user. AVAILABILITY jSquid is available for direct usage and download at http://jSquid.sbc.su.se including source code under the GPLv3 license, and input examples. It requires Java version 5 or higher to run properly. CONTACT erik.sonnhammer@sbc.su.se SUPPLEMENTARY INFORMATION Supplementary data are available at Bioinformatics online.

Systematic evaluation of label‐free and super‐SILAC quantification for proteome expression analysis

RATIONALE Advanced implementations of mass spectrometry (MS)-based proteomics allow for comprehensive proteome expression profiling across many biological samples. The outcome of such studies critically depends on accurate and precise quantification, which has to be ensured for high-coverage proteome analysis possible on fast and sensitive mass spectrometers such as quadrupole orbitrap instruments. METHODS We conducted ultra-high-performance liquid chromatography (UHPLC)/MS experiments on a Q Exactive to systematically compare label-free proteome quantification across six human cancer cell lines with quantification against a shared reference mix generated by stable isotope labeling with amino acids in cell culture (super-SILAC). RESULTS Single-shot experiments identified on average about 5000 proteins in the label-free compared to about 3500 in super-SILAC experiments. Label-free quantification was slightly less precise than super-SILAC in replicate measurements, verifying previous results obtained for lower proteome coverage. Due to the higher number of quantified proteins, more significant differences were detected in label-free cell line comparisons, whereas a higher percentage of quantified proteins was identified as differentially expressed in super-SILAC experiments. Additional label-free replicate analyses effectively compensated for lower precision of quantification. Finally, peptide fractionation by high pH reversed-phase chromatography prior to LC/MS analysis further increased the robustness and precision of label-free quantification in conjunction with higher proteome coverage. CONCLUSIONS Our results benchmark and highlight the utility of label-free proteome quantification for applications such as target and biomarker discovery on state-of-the-art UHPLC/MS workflows.

Comparative proteome analysis across non-small cell lung cancer cell lines.

Non-small cell lung cancer (NSCLC) cell lines are widely used model systems to study molecular aspects of lung cancer. Comparative and in-depth proteome expression data across many NSCLC cell lines has not been generated yet, but would be of utility for the investigation of candidate targets and markers in oncogenesis. We employed a SILAC reference approach to perform replicate proteome quantifications across 23 distinct NSCLC cell lines. On average, close to 4000 distinct proteins were identified and quantified per cell line. These included many known targets and diagnostic markers, indicating that our proteome expression data represents a useful resource for NSCLC pre-clinical research. To assess proteome diversity within the NSCLC cell line panel, we performed hierarchical clustering and principal component analysis of proteome expression data. Our results indicate that general proteome diversity among NSCLC cell lines supersedes potential effects common to K-Ras or epidermal growth factor receptor (EGFR) oncoprotein expression. However, we observed partial segregation of EGFR or KRAS mutant cell lines for certain principal components, which reflected biological differences according to gene ontology enrichment analyses. Moreover, statistical analysis revealed several proteins that were significantly overexpressed in KRAS or EGFR mutant cell lines.

Pareto Optimization Identifies Diverse Set of Phosphorylation Signatures Predicting Response to Treatment with Dasatinib.

Multivariate biomarkers that can predict the effectiveness of targeted therapy in individual patients are highly desired. Previous biomarker discovery studies have largely focused on the identification of single biomarker signatures, aimed at maximizing prediction accuracy. Here, we present a different approach that identifies multiple biomarkers by simultaneously optimizing their predictive power, number of features, and proximity to the drug target in a protein-protein interaction network. To this end, we incorporated NSGA-II, a fast and elitist multi-objective optimization algorithm that is based on the principle of Pareto optimality, into the biomarker discovery workflow. The method was applied to quantitative phosphoproteome data of 19 non-small cell lung cancer (NSCLC) cell lines from a previous biomarker study. The algorithm successfully identified a total of 77 candidate biomarker signatures predicting response to treatment with dasatinib. Through filtering and similarity clustering, this set was trimmed to four final biomarker signatures, which then were validated on an independent set of breast cancer cell lines. All four candidates reached the same good prediction accuracy (83%) as the originally published biomarker. Although the newly discovered signatures were diverse in their composition and in their size, the central protein of the originally published signature — integrin β4 (ITGB4) — was also present in all four Pareto signatures, confirming its pivotal role in predicting dasatinib response in NSCLC cell lines. In summary, the method presented here allows for a robust and simultaneous identification of multiple multivariate biomarkers that are optimized for prediction performance, size, and relevance.

Abstract LB-325: Global analysis of the phosphoproteome of human blasts reveals predictive phosphorylation markers for the treatment of acute myeloid leukemia with quizartinib

Proceedings: AACR Annual Meeting 2014; April 5-9, 2014; San Diego, CA Acute Myeloid Leukemia (AML) results from a combination of oncogenic events that can involve multiple signal transduction pathways including mutation-induced activation of tyrosine kinases. Kinase inhibitors are increasingly studied as promising targeted approaches either alone or in combination with other agents. However, only subsets of patients respond to respective targeted therapies. Internal tandem duplication (ITD) of FLT3 is one of the most common mutations in AML. It causes constitutive activation of FLT3. Quizartinib (AC220) is an example of a potent FLT3 inhibitor that was studied in a recent phase II open-label study in patients with relapsed/refractory AML. However, the presence of activating mutations within FLT3 can predict response to a certain extent only. Here, we investigated whether large-scale analyses of phosphorylation-based signalling events allows identification of more accurate markers based on the hypothesis that the read-out is closer to the mode of action of FLT3 inhibitors. Therefore, we applied quantitative mass-spectrometry to globally profile the phosphoproteome of 12 pre-treatment bone marrow aspirates obtained from AML patients treated with the quizartinib. A signature derived from this analysis consists of five phospho-sites within the proteins EEPD1, BCL11A, RANBP3, RP3, and LMN1 and it accurately predicted response to treatment with AC220 as revealed by validation in additional independent nine AML patients. Although the combined signature of five phospho-sites showed the highest prediction accuracy, we could demonstrate that in particular phosphorylation of S640 on BCL11A and S333 on RANBP3 lead to almost equally good predictions if used as individual markers. Furthermore, we could show that in case of BCL11A, EEPD1, and LMN1 the expression of the total protein correlates with its phosphorylation and thus with response. The phosphorylation markers were identified and validated in bone marrow aspirates. Although, it is clinical standard procedure to use bone marrow aspirates for diagnosis of AML patients, a predictive test that can be applied to peripheral blood samples would have many advantages. Indeed, we could show that the phosphorylation of the marker proteins strongly correlate between bone marrow and peripheral blood samples from the same patients, suggesting that the phosphorylation or protein markers can be measured and are predictive in both, bone marrow and peripheral blood samples. Citation Format: Christoph Schaab, Felix Oppermann, Martin Klammer, Heike Pfeifer, Andreas Tebbe, Thomas Oellerich, Juergen Krauter, Mark Levis, Alexander Perl, Henrik Daub, Bjoern Steffen, Klaus Godl, Hubert Serve. Global analysis of the phosphoproteome of human blasts reveals predictive phosphorylation markers for the treatment of acute myeloid leukemia with quizartinib. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr LB-325. doi:10.1158/1538-7445.AM2014-LB-325