Systems modelling of the EGFR-PYK2-c-Met interaction network predicts and prioritizes synergistic drug combinations for triple-negative breast cancer

Sung-Young Shin, Anna Katharina Müller, Nandini Verma, Sima Lev, Lan Nguyen

Research output: Contribution to journalArticleResearchpeer-review

3 Citations (Scopus)

Abstract

Prediction of drug combinations that effectively target cancer cells is a critical challenge for cancer therapy, in particular for triple-negative breast cancer (TNBC), a highly aggressive breast cancer subtype with no effective targeted treatment. As signalling pathway networks critically control cancer cell behaviour, analysis of signalling network activity and crosstalk can help predict potent drug combinations and rational stratification of patients, thus bringing therapeutic and prognostic values. We have previously showed that the non-receptor tyrosine kinase PYK2 is a downstream effector of EGFR and c-Met and demonstrated their crosstalk signalling in basal-like TNBC. Here we applied a systems modelling approach and developed a mechanistic model of the integrated EGFR-PYK2-c-Met signalling network to identify and prioritize potent drug combinations for TNBC. Model predictions validated by experimental data revealed that among six potential combinations of drug pairs targeting the central nodes of the network, including EGFR, c-Met, PYK2 and STAT3, co-targeting of EGFR and PYK2 and to a lesser extent of EGFR and c-Met yielded strongest synergistic effect. Importantly, the synergy in co-targeting EGFR and PYK2 was linked to switch-like cell proliferation-associated responses. Moreover, simulations of patient-specific models using public gene expression data of TNBC patients led to predictive stratification of patients into subgroups displaying distinct susceptibility to specific drug combinations. These results suggest that mechanistic systems modelling is a powerful approach for the rational design, prediction and prioritization of potent combination therapies for individual patients, thus providing a concrete step towards personalized treatment for TNBC and other tumour types.

Original languageEnglish
Article numbere1006192
Number of pages30
JournalPLoS Computational Biology
Volume14
Issue number6
DOIs
Publication statusPublished - 19 Jun 2018

Keywords

  • simulation and modeling
  • drug therapy
  • signalling networks
  • cancer treatment
  • network analysis
  • STAT signaling
  • dose prediction methods
  • signal inhibition

Cite this

@article{5c46d3f0392e448aa5116ba86fae8267,
title = "Systems modelling of the EGFR-PYK2-c-Met interaction network predicts and prioritizes synergistic drug combinations for triple-negative breast cancer",
abstract = "Prediction of drug combinations that effectively target cancer cells is a critical challenge for cancer therapy, in particular for triple-negative breast cancer (TNBC), a highly aggressive breast cancer subtype with no effective targeted treatment. As signalling pathway networks critically control cancer cell behaviour, analysis of signalling network activity and crosstalk can help predict potent drug combinations and rational stratification of patients, thus bringing therapeutic and prognostic values. We have previously showed that the non-receptor tyrosine kinase PYK2 is a downstream effector of EGFR and c-Met and demonstrated their crosstalk signalling in basal-like TNBC. Here we applied a systems modelling approach and developed a mechanistic model of the integrated EGFR-PYK2-c-Met signalling network to identify and prioritize potent drug combinations for TNBC. Model predictions validated by experimental data revealed that among six potential combinations of drug pairs targeting the central nodes of the network, including EGFR, c-Met, PYK2 and STAT3, co-targeting of EGFR and PYK2 and to a lesser extent of EGFR and c-Met yielded strongest synergistic effect. Importantly, the synergy in co-targeting EGFR and PYK2 was linked to switch-like cell proliferation-associated responses. Moreover, simulations of patient-specific models using public gene expression data of TNBC patients led to predictive stratification of patients into subgroups displaying distinct susceptibility to specific drug combinations. These results suggest that mechanistic systems modelling is a powerful approach for the rational design, prediction and prioritization of potent combination therapies for individual patients, thus providing a concrete step towards personalized treatment for TNBC and other tumour types.",
keywords = "simulation and modeling, drug therapy, signalling networks, cancer treatment, network analysis, STAT signaling, dose prediction methods, signal inhibition",
author = "Sung-Young Shin and M{\"u}ller, {Anna Katharina} and Nandini Verma and Sima Lev and Lan Nguyen",
year = "2018",
month = "6",
day = "19",
doi = "10.1371/journal.pcbi.1006192",
language = "English",
volume = "14",
journal = "PLoS Computational Biology",
issn = "1553-7358",
publisher = "Public Library of Science",
number = "6",

}

Systems modelling of the EGFR-PYK2-c-Met interaction network predicts and prioritizes synergistic drug combinations for triple-negative breast cancer. / Shin, Sung-Young; Müller, Anna Katharina; Verma, Nandini; Lev, Sima; Nguyen, Lan.

In: PLoS Computational Biology, Vol. 14, No. 6, e1006192, 19.06.2018.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Systems modelling of the EGFR-PYK2-c-Met interaction network predicts and prioritizes synergistic drug combinations for triple-negative breast cancer

AU - Shin, Sung-Young

AU - Müller, Anna Katharina

AU - Verma, Nandini

AU - Lev, Sima

AU - Nguyen, Lan

PY - 2018/6/19

Y1 - 2018/6/19

N2 - Prediction of drug combinations that effectively target cancer cells is a critical challenge for cancer therapy, in particular for triple-negative breast cancer (TNBC), a highly aggressive breast cancer subtype with no effective targeted treatment. As signalling pathway networks critically control cancer cell behaviour, analysis of signalling network activity and crosstalk can help predict potent drug combinations and rational stratification of patients, thus bringing therapeutic and prognostic values. We have previously showed that the non-receptor tyrosine kinase PYK2 is a downstream effector of EGFR and c-Met and demonstrated their crosstalk signalling in basal-like TNBC. Here we applied a systems modelling approach and developed a mechanistic model of the integrated EGFR-PYK2-c-Met signalling network to identify and prioritize potent drug combinations for TNBC. Model predictions validated by experimental data revealed that among six potential combinations of drug pairs targeting the central nodes of the network, including EGFR, c-Met, PYK2 and STAT3, co-targeting of EGFR and PYK2 and to a lesser extent of EGFR and c-Met yielded strongest synergistic effect. Importantly, the synergy in co-targeting EGFR and PYK2 was linked to switch-like cell proliferation-associated responses. Moreover, simulations of patient-specific models using public gene expression data of TNBC patients led to predictive stratification of patients into subgroups displaying distinct susceptibility to specific drug combinations. These results suggest that mechanistic systems modelling is a powerful approach for the rational design, prediction and prioritization of potent combination therapies for individual patients, thus providing a concrete step towards personalized treatment for TNBC and other tumour types.

AB - Prediction of drug combinations that effectively target cancer cells is a critical challenge for cancer therapy, in particular for triple-negative breast cancer (TNBC), a highly aggressive breast cancer subtype with no effective targeted treatment. As signalling pathway networks critically control cancer cell behaviour, analysis of signalling network activity and crosstalk can help predict potent drug combinations and rational stratification of patients, thus bringing therapeutic and prognostic values. We have previously showed that the non-receptor tyrosine kinase PYK2 is a downstream effector of EGFR and c-Met and demonstrated their crosstalk signalling in basal-like TNBC. Here we applied a systems modelling approach and developed a mechanistic model of the integrated EGFR-PYK2-c-Met signalling network to identify and prioritize potent drug combinations for TNBC. Model predictions validated by experimental data revealed that among six potential combinations of drug pairs targeting the central nodes of the network, including EGFR, c-Met, PYK2 and STAT3, co-targeting of EGFR and PYK2 and to a lesser extent of EGFR and c-Met yielded strongest synergistic effect. Importantly, the synergy in co-targeting EGFR and PYK2 was linked to switch-like cell proliferation-associated responses. Moreover, simulations of patient-specific models using public gene expression data of TNBC patients led to predictive stratification of patients into subgroups displaying distinct susceptibility to specific drug combinations. These results suggest that mechanistic systems modelling is a powerful approach for the rational design, prediction and prioritization of potent combination therapies for individual patients, thus providing a concrete step towards personalized treatment for TNBC and other tumour types.

KW - simulation and modeling

KW - drug therapy

KW - signalling networks

KW - cancer treatment

KW - network analysis

KW - STAT signaling

KW - dose prediction methods

KW - signal inhibition

UR - http://www.scopus.com/inward/record.url?scp=85049380391&partnerID=8YFLogxK

U2 - 10.1371/journal.pcbi.1006192

DO - 10.1371/journal.pcbi.1006192

M3 - Article

VL - 14

JO - PLoS Computational Biology

JF - PLoS Computational Biology

SN - 1553-7358

IS - 6

M1 - e1006192

ER -