Landscaping the cell surface proteome of breast cancer: Following pathways through organelles to the plasma membrane

Breast cancer is one of the most common cancers in women. It is most commonly treated by the surgical removal
of the tumour in combination with (neo)-adjuvant therapy (hormone, chemo- or radio- therapy before or after
surgery). However, a large number of patients are over treated, with approximately 60% being given adjuvant
therapy when already cured by surgery alone. This causes many undesirable side effects and cost hence, there is
great need for new diagnostic and prognostic methods to decide if patients are in need of adjuvant therapy.
A number of prognostic and treatment predictive factors have been established such as tumour size, hormone
receptor status, histological grade and age. Hereditary predisposition to developing cancer can be a factor, with
several high penetrance genes identified such as BRCA1 and BRCA2 genes as well as many as a dozen lower
risk genes that are however additive in effect. Molecular subtyping has significant prognostic value allowing the
differentiation of several subtypes having unique survival outcomes, particularly for tumours highly responsive or
nonresponsive to hormonal or targeted drug therapies. Currently the prognostic and treatment predictive factors
are mainly based on the primary tumour status, even though the distant metastases are the main reason for
breast cancer related deaths. Thus, there is need for novel approaches with higher specificity and sensitivity in
newly developed targeted therapies.
The aim of this thesis was to understand the changes in breast cancer tumour cells and tissues, by comparing
protein expression levels in different conditions, using mass spectrometry. Attention was specifically on the
analysis of patient samples, with pairs of primary tumours and metastases, in order to try to understand what
happens to allow tumour cells to be able to metastasise and to identify novel molecular markers. Hence we also
investigated the biological functions of proteins by integrating information about the processes and pathways in
which these proteins take part in order to be able to better understand the contribution of different proteins to
breast cancer development. Further we have explored molecular classification markers for breast cancer tumours
into major intrinsic subtypes. Our results demonstrated a great overlap of subtypes, using gene expression and
protein expression profiling. In conclusion, all our findings together show the great need for improved and early
cancer detection and treatment as well as need for development and promises of personalized medicine.