CRISPR-Mediated Functional Genomic Screens for Modifiers of Trastuzumab Emtansine Response

Abstract

Amplification of the HER2 oncogene results in breast cancers with an aggressive clinical course. Therapeutic agents targeting the HER2 receptor provide objective responses in some patients and have improved overall survival. However, many tumours are refractory to HER2 blockade or acquire resistance during treatment. Mechanisms of resistance to these agents are poorly understood, particularly for the antibody-drug conjugate trastuzumab emtansine (T-DM1), which is comprised of the potent microtubule inhibitor mertansine (DM1) conjugated to trastuzumab. This project utilised CRISPR-mediated functional genomics to identify genetic determinants of sensitivity and resistance to T-DM1 in breast cancer. A panel of human breast cancer cell lines showing varying patterns of sensitivity to HER2 blockade was studied. HER2 amplification and protein expression were assessed by FISH and immunoblotting, respectively. Drug sensitivity was characterised by CellTiter-Glo and clonogenic cell survival assays. Streptococcus pyogenes Cas9 and the GeCKOv2 whole-genome single guide (sg) RNA library were stably transduced using lentiviral vectors. The resulting genome-scale knockout libraries were functionally validated using 6-thioguanine, for which resistance-conferring mutations are known. Library complexity and treatment-induced sgRNA enrichment were assessed by next-generation sequencing (Illumina NextSeq500). Saturation-scale knockout libraries were used in screens with T-DM1 and free DM1. The cell lines displayed patterns of sensitivity to T-DM1, trastuzumab, lapatinib, and neratinib reflective of clinical response rates. DM1 was universally potent, while T-DM1 retained activity for trastuzumab-refractory cells. Genome-scale knockout libraries showed resistance to 6-thioguanine following drug challenge. Deep sequencing of treatedMDA-MB-453 libraries revealed enrichment of mutations in the 6-thioguanine sensitivity gene HPRT1. Saturation-scale screens with MDA-MB-453 and MDA-MB-361 libraries showed resistance to T-DM1 but not to DM1 after eight and thirteen weeks of escalating drug treatment, respectively. Deep sequencing revealed a number of genes candidates, including ERBB2, SLC46A3, KEAP1, CUL3, SEPSECS, EEFSEC, PSTK, NXPH1, and NUMA1 that are potential regulators of T-DM1 and DM1 response. The genome-wide knockout libraries developed in this study are anticipated to enable identification of genetic mechanisms of treatment resistance, potentially leading to more personalised management of HER2-positive breast cancer.