Resistance to HER2-targeted therapies in HER2-positive breast cancer
Item statusRestricted Access
Embargo end date06/07/2020
Breast cancer accounts for 522,000 deaths worldwide each year and is the most common cancer in women. It is classified according to the cell of origin and its expression of several receptors: oestrogen and progesterone receptors, and human epidermal growth factor receptor 2 (HER2). Historically, HER2-positive breast cancers had a worse survival prognosis than oestrogen or progesterone receptor-positive cancers, but the development of HER2-targeted therapies led to significant survival improvements. Despite this, patients often present with de novo resistance, or will develop acquired resistance to targeted therapies. Several resistance mechanisms have been identified but attempts to target them have failed. Thus, it is of paramount importance to identify the mechanisms used, to prevent development of resistance or resensitise tumours to HER2-targeted therapies. Objectives of this study were: to understand the link between epithelial to mesenchymal transition (EMT) and loss of HER2, seen in a model of acquired resistance to the HER2-targeted therapy, sapatinib, and to characterise and validate tumours from a sapatinib-treated spontaneous mouse model of HER2-positive breast cancer. The EMT-linked transcription factor ZEB1 was associated with acquired resistance to sapatinib in tumours that had undergone EMT and concurrently lost HER2. Generation of drug resistant cell lines failed to recapitulate the in vivo phenotype. Transient overexpression of ZEB1 in vitro did not induce clear EMT or loss of HER2, despite a trend towards lower HER2 expression. However, we found that treatment of cells with ERBB2 shRNA, the gene encoding HER2, increased levels of ZEB1 and enhanced migration, but did not induce overt EMT. This may be the result of differing PTEN status between in vivo and in vitro models. Treatment of a spontaneous mouse model of HER2-positive breast cancer with sapatinib revealed that progressing tumours had an increase in proteins associated with cellular iron homeostasis. Further investigation revealed increased heme oxygenase-1 (HO-1), iron exporter ferroportin and altered iron storage. To ascertain if modulation of dietary iron intake could affect the development of resistance to sapatinib, mice were given a control or iron-deficient diet and treated with vehicle or sapatinib. This showed that in sapatinib-treated mice fed an iron-deficient diet, HO-1 was not increased as in tumours from mice fed a iron-low control diet. We looked at the possibility of HER2-targeting therapies inducing ferroptosis, an iron-dependent form of cell death. Sapatinib-treated tumours from mice on a iron-low control diet had increased cyclooxygenase 2 (COX2), a marker of ferroptosis, which was not seen in sapatinib-treated tumours from mice on an iron-deficient diet. Additionally, in vitro drug treatments with HER2-targeting agents showed that SKBR3 cell death could be rescued by iron chelation. HO-1 overexpression in SKBR3 cells revealed increased autophagic flux and resistance to HER2-targeted therapies. Inhibition of autophagy reversed resistance, rendering them susceptible to sapatinib- and lapatinib-induced cell death. Further, increased autophagic flux was seen in all progressive tumours on sapatinib. The increased resistance to sapatinib in mice fed an iron-deficient diet was also associated with increased autophagic flux, although this was HO-1-independent. Taken together, the results presented here provide a novel mechanism of cell death induced by HER2-targeting agents in vitro and in vivo. We have shown that increased HO-1 and reducing dietary iron can affect the development of resistance to sapatinib, which is reliant on autophagy induction. Further, inhibiting autophagy can resensitise cells to sapatinib and lapatinib treatment.