Project description:Aim: The goals of this study were to identify transcriptomic changes that arise in basal-like breast cancer cells during the development of resistance to epidermal growth factor receptor inhibitors (EGFRi) and to identify drugs that are cytotoxic once EGFRi resistance occurs. Methods: Human patient-derived xenografts (PDXs) were grown in immunodeficient mice and were treated with a set of EGFRi; the EGFRi erlotinib was selected for more expansive in vivo studies. Single-cell RNA sequencing was performed on mammary tumors from the basal-like PDX WHIM2 that was treated with vehicle or erlotinib for nine weeks. The PDX was then subjected to long-term erlotinib treatment in vivo and through serialserially passaging, an erlotinib-resistant subline of WHIM2 was generated. Bulk RNA-sequencing was performed on parental and erlotinib-resistant tumors. In vitro high-throughput drug screening with >500 clinically used compounds was performed on parental and erlotinib-resistant cells. Previously published bulk gene expression microarray data from MMTV-Wnt1 tumors were contrasted with the WHIM2 PDX data. Results: Erlotinib effectively inhibited WHIM2 tumor growth for ~four weeks. Compared to untreated cells, single-cell RNA sequencing revealed that a greater proportion of erlotinib-treated cells were in the G1 phase of the cell cycle. Comparison of WHIM2 and MMTV-Wnt1 RNA level data revealed a set of 37 overlapping genes that were differentially expressed in the erlotinib-resistant subsets of WHIM2 and MMTV-Wnt1 tumors. Comparison of all three data types revealed five genes that were upregulated across all erlotinib-resistant samples: IL19, KLK7, LCN2, SAA1, and SAA2. Despite transcriptomic differences, parental and erlotinib-resistant WHIM2 displayed similar responses to the majority of drugs assessed for cytotoxicity in vitro. Conclusion: This study identified transcriptomic changes arising in erlotinib-resistant basal-like breast cancer. Interestingly, four of the five genes that were found to be upregulated in the basal-like erlotinib-resistant samples were also upregulated in erlotinib-resistant non-small cell lung cancer samples. These gene signatures could be used to identify basal-like breast cancer patients that would best respond to EGFRi. Future studies should explore the predictive capacity of these gene signatures as well as how these differentially expressed genes contribute to the development of EGFRi resistance.

Dataset Information

Establishment of EGFR-inhibitor resistant triple negative breast cancer models from patient derived xenografts

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure