Cancer is a major public health problem worldwide and a leading cause of death in developed countries. Recently, collaborative work between the World Health Organization and the International Agency for Research on Cancer reported that more than ten million new cancers were diagnosed in 2000 and that there were over six million deaths from the disease (Stewart and Kleihues 2003). Nowadays, the incidence of cancer continues to increase slightly. However, the mortality rate from all cancers decreases by about 1% every year in developed countries (Jemal et al. 2005), essentially due to early detection.
Cancers are heterogeneous abnormalities that develop through various well-identified early genetic changes. These drive the progressive, multistep transformation of normal cells into malignant, cancerous cells. It is suggested that neoplastic cells from most of the different cancers acquire the same set of functional capabilities during their development. These include self-sufficiency in growth signals, insensitivity to antigrowth signals, evading apoptosis, sustained angiogenesis, limitless replicative potential and tissue invasion and metastasis (Hanahan and Weinberg 2000). All these events lead to modifications of the proteome that are due to the reprogramming of cancer cells and also to the body's response to cancer.
Proteomics can contribute to cancer research at several levels, namely for diagnosis, prognosis, monitoring treatment response and the identification of targets for cancer prevention and treatment. Depending on their interests, investigators can use different proteomic approaches to answer their biological questions and to highlight proteins that are present in greater or lesser quantities. Investigations often focus on comparing control tissue versus cancerous tissue, discovery of proteins specific for one state and changes in protein post-translational modifications between states. Nowadays, two main methods are used in cancer proteomics: biomarker discovery and proteomic profiling.
In biomarker discovery, the focus is on identifying proteins whose expression levels or modifications are specific for a disease state. Historically, investigators have used 2-DE as the primary tool, but in recent years, new proteomic tools have been developed for biomarker discovery in tissues and cell lines.
The comparison of proteomes from normal versus malignant tissues has been widely used to highlight proteins that could be involved in disease establishment and progression. For tissue-based studies, sample preparation is of particular importance to avoid protein degradation prior to proteomic analysis. Numerous studies have successfully examined the whole proteome of cancer tissues. However, tumours are heterogeneous samples containing various proportions of cell types such as fibroblasts, endothelial cells, normal epithelial cells, immune cells and others. This heterogeneity has a major impact on comparative studies. This is why different techniques have been set up to selectively enrich samples for cells of interest, allowing the analysis of individual cell types.
Franzen et al. (1995) addressed this critical issue in several publications and demonstrated that non-enzymatic methods for the preparation of tumoral cells, including fine-needle aspiration, scraping or squeezing tissue biopsies, had advantages over methods using enzymatic extraction of cells. Non-enzymatic methods were shown to be rapid and to reduce loss of high molecular weight proteins. These methods did not require the separation of viable and nonviable cells by Percoll gradient centrifugation. They also analysed qualitative aspects of tissue preparation in relation to the histopathology of lung cancer, and examined the relationship between histopathological findings and 2-DE gel quality. They concluded that histopathological features, such as a local homogeneity, and the amounts of connective tissue and serum proteins were critical factors for the successful preparation of the sample and the high quality of overall protein separation and analysis. They clearly overcame some major technical difficulties. As a result of their work, clear guidelines are now available for sample preparation of patient cells and biopsies (Franzen et al. 1995).
Page et al. (1999) used immunoselection of cells before proteomic analysis. They used a double antibody magnetic affinity cell sorting technique to purify normal human luminal and myoepithelial breast cells from the reduction mammoplasties of ten premenopausal women. For this, two antibodies were used, one produced in rat and directed to the luminal epithelial marker EMA and the other produced in mouse and directed to the myoepithelial antigen CD-10. The use of antirat and antimouse magnetic beads allowed the separation of EMA and CD-10 expressing cells. Myoepithelial cells were then purified a second time using anti-CD-10 and anti-FAP antibodies to free them from F-19 positive fibroblasts. Purified luminal and myoepithelial cells were then subjected to proteomic analyses using 2-DE. One hundred and seventy different proteins were found to be differentially expressed between the two breast cell types and 51 of them were identified using MS/MS. This work forms the basis for future studies of purified breast cancer cells.
Another technique, called laser-capture microdissection, is widely used in studies of cancer tissues (Jain 2002). This technique employs a pulsed infrared laser to activate a transfer film placed over the tissue of interest, causing the film to become fused to the cells. With a diameter measuring only few microns, laser-capture microdissection allows single cells to be extracted from heterogeneous tumour samples. Interestingly, this technique permits tumour cells and normal cells to be isolated from the same biopsy without using chemicals or physical agents that might modify the proteome. An example of this technique is the work of Li et al. (2004) and their study of hepatocellular carcinoma (HCC). They used laser-capture microdissection to isolate HCC and non-HCC hepatocytes, and compared them using cleavable ICAT and 2-D LC-MS/MS. A total of 644 proteins were identified and 261 differentially expressed proteins were described. These results provided a new basis for understanding the mechanism of HCC, and identified potential markers and drug targets that could be useful for disease diagnosis and treatment.
Cell lines have been extensively used for the proteomic analysis of cancer. In several respects, they are much more useful than biopsies. In contrast to tissues, the quantity of cells is unlimited. This allows extensive experiments.
Cell lines also represent a pure cell population and allow investigators to manipulate growth conditions. However, it is important to mention that in vitro cell lines are not identical to the corresponding cells in vivo since they are removed from their native environment. For example, Ornstein et al. (2000) compared the proteome of in vivo prostate cancer cells with in vitro prostate cell lines and found that less than 20% of proteins were shared by both cell types.
Cell lines have proven to be very useful for the unravelling of pathways perturbed by particular genetic alterations. For example, new insights into the development of nervous system tumours in patients bearing the NF1 tumour-predisposition syndrome have been recently published (Dasgupta et al. 2005). For this, the investigators generated transgenic mice producing NF1-/— astrocytes. The proteome of these astrocytes was compared with that of NF1+/+ astrocytes using 2-DE. The MS identification of differentially expressed proteins between these two cell lines revealed that the mammalian target of the rapamycin pathway is hyperactivated in NF1-/- astrocytes. Furthermore, the inhibition of this pathway in NF1-/- astrocytes restores a normal proliferative rate, suggesting that this pathway could represent a target for therapy of brain tumours in patients bearing the NF1 tumour-predisposition syndrome.
Resistance to therapy is the main cause of therapeutic failure and death in cancer patients. Cell lines represent a powerful tool for the study of therapy-resistant cells. Brown and Fenselau (2004) have addressed the question of doxorubicin resistance in breast cancer cells. For this, they used shotgun pro-teomics with proteolytic 18O labelling to compare the cytosolic proteome of a doxorubicin resistant MCF-7 cell line with the drug-sensitive MCF-7 cell line. This study identified several proteins with altered expression levels in the drug-resistant cell line. A number of these proteins may represent key factors explaining doxorubicin resistance of breast cancer cells.
An advantage of cell lines for cancer proteomics, as compared with tissue samples, is the opportunity for efficient subcellular fractionation. Subcellular fractionation allows the direct investigation of functional cell compartments instead of complete proteomes. This enriches low-abundance proteins, and allows the focus on subcellular compartments with potential value in tumori-genesis and cancer progression. This includes the plasma membrane (Zhao et al. 2004) or nucleoli (Scherl et al. 2002).
Biomarker discovery using proteomics has proven to be a valuable means to identify markers that could become useful in diagnosis, prognosis or treatment of cancer. To date, dozens of markers have been identified using pro-teomic strategies. The most promising ones are now undergoing validation.
Was this article helpful?