Signaling mechanisms in progestin-induced canine mammary tumorigenesis

A. Gracanin

    Research output: ThesisDoctoral thesis 1 (Research UU / Graduation UU)


    Breast cancer is the most prominent cancer in women of the Western world with over 1.7 million women newly diagnosed and >500,000 deaths annually. The majority of breast cancers are hormone receptor positive and potentially bear sensitivity towards endocrine treatment. Despite initially effective, still some 30% of women with breast cancer eventually die of their disease due to therapy resistant metastases. Therefore, gaining insight into the development of metastasis and therapy resistance is crucial, especially in initial endocrine-responsive disease. Active Wnt signaling contributes to stemness and metastasis, and is upregulated in models of acquired resistance to hormonal therapies. Whereas spontaneous mammary tumors in rodents are either not hormone-sensitive (mice) or do not metastasize (rats) they are not ideally suited for research on hormone-dependent breast cancer. Female dogs are well-known for their sensitivity to hormone-induced mammary cancer and early castration leads to a decreased incidence of spontaneous mammary carcinomas. Our findings on the role of progestin-induced growth hormone (GH) expression in the canine mammary gland showed GH to play also a role in human breast epithelium, where it results in increased proliferation of stem/progenitor cells. In this thesis we focused on three aspects regarding progestin-induced canine mammary tumorigenesis: 1) Studies focusing on expression and function of human and murine PR isoforms have revealed high degree of conservation but also species-specific features. We first addressed the function and expression of canine PR isoforms. We show that dogs (and other canids) are unique among mammals in that they have multiple mutations within AF3 domain resulting in a poor PR-B related transcriptional activity. Moreover, dogs appear to have low (or no) expression of PR-B. This data is further discussed in relation to unique progesterone-related reproductive biology of canids. 2) A subset of canine mammary tumors expresses mammary GH in the absence of PR, suggesting a loss of direct progesterone control and/or cross-regulation by additional pathways. Similarly, canonical Wnt activity in human breast cancer was shown to be responsive to deregulation of multiple signaling pathways. We explored the possibility of direct interaction between GH and canonical Wnt signaling in PR- canine mammary tumor cell lines. Using luciferase reporters and gene expression analysis we observed no direct cross-regulation, suggesting that the progesterone-dependent regulation of the two pathways is independent of each other. 3) Aberrant activation of canonical Wnt signaling has been suggested in canine mammary tumors based on limited immunohistochemical data and gene expression profiling. We quantitatively assessed the activity of canonical Wnt signaling in canine mammary tumor cell lines as well as the underlying mechanism of activation. Luciferase based TCF-reporter assay identified three groups of cell lines exhibiting high, moderate and low canonical Wnt activity. Moderate canonical Wnt activity was dependent on Wnt ligands and could be explained by overexpression of Wnt ligands and/or loss of sFRP1 antagonist. In contrast, high canonical Wnt activity was ligand-independent and not caused by known mutations in pathway components. Our data suggest a novel ligand-independent mechanism of canonical Wnt activation that may involve overexpression of a TCF-family member, LEF1.
    Original languageEnglish
    Awarding Institution
    • Utrecht University
    • Clevers, H.C., Primary supervisor
    • Mol, J.A., Co-supervisor
    Award date8 Sept 2015
    Print ISBNs978-90-393-6376-8
    Publication statusPublished - 8 Sept 2015


    • Canine
    • mammary tumor
    • progesterone
    • PR-B, Wnt signaling
    • GH signaling


    Dive into the research topics of 'Signaling mechanisms in progestin-induced canine mammary tumorigenesis'. Together they form a unique fingerprint.

    Cite this