Tumour promoting HER2 splice variant 16HER2 : regulation and implication in breast cancer

Abstract

Results from high-throughput studies of the last decade have shown an important connection between regulated pre-mRNA splicing and tumorigenesis as well as therapy resistance. To date only relatively few such interactions have been studied in depth. The human epidermal growth factor receptor 2 (HER2) is an important biomarker in cancer, especially in 20-30% of breast cancer cases. This subset of breast cancer patients are treated with HER2-targeting therapies in addition to chemotherapy. Although these therapies are very successful, in a subset of patients with primary tumours as well as in advanced metastatic breast cancer resistance occurs. The underlying mechanisms are not well understood; one potential cause is the alternative splicing of HER2 pre-mRNA yielding protein variants which can evade or prevent action of these therapeutics. This PhD thesis details progress in understanding the mechanisms regulating the production of the HER2 variant Δ16HER2, which has been shown to have high oncogenic potential. Four key splicing factors that regulate this splicing event were identified as well as a specific RNA:protein binding site in the Δ16HER2 splicing region. In addition, a potential novel HER2 mRNA transcript was identified involving retention of intron 25 (termed I25HER2). I25HER2 was shown to be expressed in both normal human tissues and breast cancer tissue samples. Interestingly, the expression level of I25HER2 is not proportional to wild type HER2 in all tissues. Although, I25HER2 would be expected to undergo nonsense-mediated decay, this study indicated that I25HER2 mRNA has similar stability to Δ16HER2 mRNA. An ongoing challenge in both clinical and non-clinical studies is the inability to distinguish different HER2 protein variants. Initial results presented in this thesis, from a comparative study between immunohistochemistry (using three different HER2 antibodies, including the clinically used HercepTestTM) and HER2 gene amplification (using HER2 CISH pharmDxTM) in a cohort of primary human ductal carcinomas showed a subset of discordant cases. In a few of these tumours elevated levels of HER2 splice variants could be detected. The iii possible implication for successful treatment of these patients’ highlights the importance of linking the basic processing of HER2 transcripts and protein variants with pathological information from patients.