Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
[Preprint]. 2023 Feb 16:2023.02.15.528736.
doi: 10.1101/2023.02.15.528736.

Network-based elucidation of colon cancer drug resistance by phosphoproteomic time-series analysis

George Rosenberger  1 , Wenxue Li  2 , Mikko Turunen  1 , Jing He  1   3 , Prem S Subramaniam  1 , Sergey Pampou  1   4 , Aaron T Griffin  1   5 , Charles Karan  1   4 , Patrick Kerwin  1 , Diana Murray  1 , Barry Honig  1 , Yansheng Liu  2   6 , Andrea Califano  1   4   7   8   9   10
Affiliations

Network-based elucidation of colon cancer drug resistance by phosphoproteomic time-series analysis

George Rosenberger et al. bioRxiv. .

Update in

Abstract

Aberrant signaling pathway activity is a hallmark of tumorigenesis and progression, which has guided targeted inhibitor design for over 30 years. Yet, adaptive resistance mechanisms, induced by rapid, context-specific signaling network rewiring, continue to challenge therapeutic efficacy. By leveraging progress in proteomic technologies and network-based methodologies, over the past decade, we developed VESPA-an algorithm designed to elucidate mechanisms of cell response and adaptation to drug perturbations-and used it to analyze 7-point phosphoproteomic time series from colorectal cancer cells treated with clinically-relevant inhibitors and control media. Interrogation of tumor-specific enzyme/substrate interactions accurately inferred kinase and phosphatase activity, based on their inferred substrate phosphorylation state, effectively accounting for signal cross-talk and sparse phosphoproteome coverage. The analysis elucidated time-dependent signaling pathway response to each drug perturbation and, more importantly, cell adaptive response and rewiring that was experimentally confirmed by CRISPRko assays, suggesting broad applicability to cancer and other diseases.

PubMed Disclaimer

Conflict of interest statement

Declaration of Interests A.C. is founder, equity holder, and consultant of DarwinHealth Inc, a company that has licensed some of the algorithms used in this manuscript from Columbia University. Columbia University is also an equity holder in DarwinHealth Inc and assignee of patent US10,790,040, which covers some components of the algorithms used in this manuscript. The other authors declare no competing interests.

Figures

Figure 1.
Figure 1.. Methodological overview of VESPA.
a) VESPA infers protein signaling activity for kinases and phosphatases on substrate-level (blue box, left panel) and activity-level (pink box, right panel). A quantitative matrix of phosphopeptide or phosphosite abundance across samples or conditions including missing values (black) represents the main input for VESPA. The signaling network reconstruction module uses this matrix together with optional priors from reference networks to assess regulatory relationships by computing the mutual information between enzymatic regulator and target phosphopeptides or phosphosites and the signal transduction Data Processing Inequality (stDPI) to generate signalons for each regulator consisting of interaction probabilistic weight and mode of regulation (kinase activation: red, phosphatase deactivation: blue) with substrate targets. VESPA then uses the substrate-level signalons to infer substrate-level kinase/phosphatase activity (blue box). This quantitative matrix represents the input for activity-level signaling network reconstruction, which uses the Data Processing Inequality (DPI) to generate more abstract and generalized signalons, which are then in turn used to infer protein signaling activity on activity-level (pink box). b) Methodological differences between substrate- and activity-level signaling networks. On substrate-level, ST-Ks (e.g. GSK3A, green) are primarily associated with direct phosphorylation targets, whereas TKs (e.g. ERBB2, orange) can frequently not be directly associated with (unenriched) tyrosine-phosphorylated sites. On activity-level, more abstract “activation/deactivation” events can better associate targets for both ST-Ks and TKs.
Figure 2.
Figure 2.. Benchmark and validation of dVESPA and mVESPA.
a) Comparison of different mutual information (MI) estimation strategies based on imputation (iMI), depletion (dMI) and hybrid partitioning (hpMI) and the MI – Spearman correlation relationship using the CPTAC-S45 dataset. b) Receiver operating characteristic of (signal transduction) Data Processing Inequality (stDPI/DPI) and unprocessed (noDPI) signaling dependencies as evaluated using a ground truth dataset. c) Baseline profiles of six diverse CRC cell lines were acquired and used with the GDSC reference database to identify sensitive and resistant or insensitive cell lines for each covered drug compound. Using mVESPA and CRC-specific signalons, differential comparisons were conducted for each drug compound to identify the top differentially active regulators. d) VESPA (red), consisting of mVESPA and dVESPA, performs substantially better than mVESPA using context unspecific Pathway Commons (blue) SigNets.
Figure 3.
Figure 3.. Representation of CRC subtypes by cell line models and activity-level VESPA.
a) UMAP embedding of KP-enzyme activities and coloration according to different classification systems (phosphoproteome-based VESPA; VC and the CRC Consensus Moleular Signature; CMS). b) The most informative proteins and their VESPA inferred normalized enrichment scores (NES) have been selected for visualization (full datasets: Supplemental Fig. 5-7). CPTAC clinical profiles and cell lines were grouped according to the Consensus Molecular Classifier (CMS) and VESPA clusters (VC). The samples are grouped according to VC. c) Gene Set Enrichment Analysis (GSEA) using a signaling subset of the Reactome database. Only terms significant in at least one sample (BH-adj. p-value < 0.05) are shown. The colors represent GSEA NES and are linked to the legend in b).
Figure 4.
Figure 4.. Targeted drug perturbations of CRC cell lines.
a) A global overview of VESPA inferred normalized enrichment scores (NES) across the full drug perturbation dataset (336 samples), covering six CRC cell lines, 7 drug perturbations and DMSO control across 7 time points. b) Gene Set Enrichment Analysis (GSEA) using a signaling subset of the Reactome database. Only terms significant in at least one sample (BH-adj. p-value < 0.05) are shown. The colors represent GSEA NES and are linked to the legend in a).
Figure 5.
Figure 5.. Temporal VESPA perturbation profiles of known primary drug compound targets.
The VESPA normalized enrichment scores corresponding to Fig. 4 are extracted and visualized for the top 5 downregulated known primary targets, grouped according to drug perturbations and cell lines.
Figure 6.
Figure 6.. Context-specific wiring of signaling pathways.
a) Analysis of the VESPA-inferred activities by the DeMAND algorithm identifies regulators with context-specific dysregulated interactions. The heatmap depicts significance of dysregulation (-log10(BH-adj. p-value). Only known drug targets (bold) and proteins with significant score (black: BH-adj. p-value < 0.05) in the unspecific DeMAND analysis are visualized. b) Grouping of the dysregulated proteins according to drug perturbations and overlay with VESPA-inferred activities of the aggregated early time points. c) Grouping of the dysregulated proteins according to drug perturbations and overlay with VESPA-inferred activities of the aggregated late time points. d) Visualization of network dysregulation and drug compound mechanism of action (MoA) for osimertinib. Nodes indicate the most affected regulators with the inner circos colors indicating cell line type and the outer circos color and node size indicating VESPA activity. The edges indicate dysregulated, undirected interactions between the regulators (Methods). Line thickness indicates significance of dysregulation. Proteins highlighted in green indicate known primary and secondary targets.
Figure 7.
Figure 7.. Context-specific adaptive stress resistance mechanisms.
a) The effect of drug perturbation vs. control in late time points is visualized for each cell line and drug perturbation separately using the differential VESPA paired t-test t -statistic between late-time-point (24h, 48h, 96h) drug perturbation vs. DMSO control samples. The depicted KP-enzymes were selected by selecting those found to be significant (q-value < 0.05, avg(t-statistic) > 0) after integrating all p-values across all comparisons by Stouffer’s method (Methods). b) Receiver operating characteristics (ROC) with area under the curve (AUC) and statistical significance (Mann-Whitney-U test) are depicted for the CRISPRko validation experiment covering HCT-15 linsitinib (LI; C2: 4.0 µM) vs. DMSO and trametinib (TR; C2: 0.7 µM) vs. DMSO, as well as NCI-H508 trametinib (TR; C2: 0.01 µM) vs. DMSO perturbation.
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Kholodenko B. N. Cell-signalling dynamics in time and space. Nature Reviews Molecular Cell Biology 7, 165–176 (2006). - PMC - PubMed
    1. Efeyan A., Comb W. C. & Sabatini D. M. Nutrient-sensing mechanisms and pathways. Nature 517, 302–310 (2015). - PMC - PubMed
    1. Broyde J. et al. Oncoprotein-specific molecular interaction maps (SigMaps) for cancer network analyses. Nature Biotechnology 39, 215–224 (2021). - PMC - PubMed
    1. Hanahan D. & Weinberg R. A. Hallmarks of cancer: The next generation. Cell 144, 646–674 (2011). gr-qc/0208024. - PubMed
    1. Kuipers E. J. et al. Colorectal cancer. Nature Reviews Disease Primers 1, 15065 (2015). - PMC - PubMed

Publication types