Integrative analysis of multi-platform reverse-phase protein array data for the pharmacodynamic assessment of response to targeted therapies

Leanne De Koning,Neil O Carragher,Bryan Serrels,Vonick Sibut,Stephan Bernhardt,Bérèngere Ouine,Aurélie Cartier,Kenneth G Macleod,Ulrike Korf,Adam Byron

doi:10.1038/s41598-020-77335-0

Abstract

Reverse-phase protein array (RPPA) technology uses panels of high-specificity antibodies to measure proteins and protein post-translational modifications in cells and tissues. The approach offers sensitive and precise quantification of large numbers of samples and has thus found applications in the analysis of clinical and pre-clinical samples. For effective integration into drug development and clinical practice, robust assays with consistent results are essential. Leveraging a collaborative RPPA model, we set out to assess the variability between three different RPPA platforms using distinct instrument set-ups and workflows. Employing multiple RPPA-based approaches operated across distinct laboratories, we characterised a range of human breast cancer cells and their protein-level responses to two clinically relevant cancer drugs. We integrated multi-platform RPPA data and used unsupervised learning to identify protein expression and phosphorylation signatures that were not dependent on RPPA platform and analysis workflow. Our findings indicate that proteomic analyses of cancer cell lines using different RPPA platforms can identify concordant profiles of response to pharmacological inhibition, including when using different antibodies to measure the same target antigens. These results highlight the robustness and the reproducibility of RPPA technology and its capacity to identify protein markers of disease or response to therapy.

Highlights

Reverse-phase protein array (RPPA) technology uses panels of high-specificity antibodies to measure proteins and protein post-translational modifications in cells and tissues
We selected for this study six breast cancer cell lines, encompassing different breast cancer molecular subtypes and presenting distinct drug sensitivities (Supplementary Table S1)
Samples were analysed at each site using the respective in-house RPPA platforms, the set-up of which differed at many stages of the RPPA analysis workflow, including slide type, the number of technical replicates and dilutions per sample, read-out dye, scanner, image analysis software and normalisation procedure (Supplementary Table S2), enabling the capture of variation between RPPA platforms operated in different laboratories

Summary

Introduction

Reverse-phase protein array (RPPA) technology uses panels of high-specificity antibodies to measure proteins and protein post-translational modifications in cells and tissues. Our findings indicate that proteomic analyses of cancer cell lines using different RPPA platforms can identify concordant profiles of response to pharmacological inhibition, including when using different antibodies to measure the same target antigens These results highlight the robustness and the reproducibility of RPPA technology and its capacity to identify protein markers of disease or response to therapy. We use collaborative RPPA-based proteomics, employing distinct RPPA workflows at multiple research sites in different countries, to characterise a range of human breast cancer cell lines and their biochemical responses to two clinically relevant cancer drugs. We combine data derived from different RPPA platforms to assess inter-platform variation and use integrative analysis to identify platform-independent protein markers of response to drug treatment Such cross-platform validation may have utility in multi-centre evaluation of robust protein markers of disease and therapeutic response

Methods

Results

Conclusion