Die funktionelle Charakterisierung des putativen Tumorsuppressors DKK3 im humanen Mammakarzinom sowie die Identifizierung potentieller SFRP1-Mimetika für die Brustkrebstherapie

The clinical need to develop new drugs for the treatment of human breast cancer is emphasized by the frequent occurrence of resistances against currently used pharmaceuticals as well as a lack of effective targeted therapies, especially for mammary carcinomas of the basal subtype. Targeting of the tumor-promoting WNT signaling pathway activity might represent a promising therapeutic strategy. A deregulation of this pathway is observed in many tumor entities, including breast cancer, and is associated with an unfavorable patient prognosis. Aberrant WNT pathway activation is frequently attributed to an expression loss of its antagonists, including DKK3 and SFRP1.Profound understanding of the molecular alterations underlying the particular subtype-specific breast carcinogenesis is essential for the development of new therapeutic strategies. To contribute to this goal, one aim of the present study was the comprehensive subtype-specific characterization of the putative tumor suppressor DKK3 in human breast cancer. A subtype-specific expression analysis using three independent tissue cohorts (n=30, n=463 and n=791) revealed a significant loss of DKK3 on mRNA and protein level in all clinically relevant breast cancer subtypes compared to healthy breast tissue controls, with the most abundant reduction in basal breast cancer cases respectively. To investigate the molecular mechanisms underlying the differential expression loss observed, a methylation analysis of both transcription-regulating CpG-islands in the DKK3 promoter and its proximal regions was performed using two independent patient cohorts (n=33 and n=497). While methylation in this region was significantly correlated with a loss of DKK3 expression in non-basal cases, basal carcinomas surprisingly presented the lowest methylation frequency indicating a different mechanism of DKK3 inactivation in these tumors. A significant correlation between a low DKK3 mRNA expression and a reduced recurrence-free survival of luminal and basal like breast cancer cases suggested a possible causal involvement of DKK3 in the tumorigenesis of these two breast cancer subtypes. This hypothesis was addressed by performing functional in vitro tests using two luminal and two basal stably transfected gain-of-function breast cancer models. Interestingly, ectopic DKK3 expression resulted in most pronounced effects on cell growth and apoptosis in basal breast cancer cells. In addition, DKK3 reexpression mediated inhibition of progression-associated characteristics of basal breast cancer cell lines resulting in a modified cell morphology, an enhanced cell-matrix adhesion and a reduced cell migration. These observations indicated a possible partial mesenchymal-to-epithelial transition in DKK3-transfected breast cancer cells, supported by an increased expression of epithelial markers like E-Cadherin and Claudin 1 as well as loss of the mesenchymal transcription factor Snail 1. In conclusion, the findings of this study suggest a possible functional involvement of a DKK3 expression loss especially in the tumorigenesis of basal breast cancers. These data provide the basis for future studies investigating the underlying molecular effects of DKK3 in detail and might therefore contribute to the development of targeted therapies for basal breast carcinomas.The second sub-project of the present dissertation was realized in cooperation with the Lead Discovery Center (LDC) in Dortmund. The aim of this study was the identification and functional characterization of specific substances mimicking the tumor suppressive effects of SRFP1. These so called SFRP1-mimetics should conduce to the future development of new drugs for the treatment of SFRP1-deficient breast cancers. The LDC initially identified twelve potential SFRP1-mimetics in a cellular primary screen. In the present study, these candidate substances were further validated in additional experiments regarding their potential SFRP1-analogous effects. Based on the results of an initial cell growth analysis, four promising molecules of these twelve putative SFRP1-mimetics were selected for a subsequent comprehensive functional in vitro analysis. All four candidates exhibited a differential, dose-dependent, growth-inhibitory capacity on SFRP1-negative and SFRP1-positive breast cancer cell culture models, with a pronounced effect on SFRP1-deficient cells. Moreover, all candidate substances influenced the expression of PCDH10, a marker gene of the SFRP1-function identified in a recent microarray analysis, in a dose-dependent manner. These data were in line with a potential SFRP1-analogous mode of action of the candidate molecules. In addition, further tumor-suppressive effects were verified for all four potential SFRP1-mimetics resulting in a decreased colony growth, an increased apoptosis and a reduced cell migration. However, the substances had varying effects on the different tumor cell characteristics analyzed so that further studies are needed to investigate their defined mechanisms of action. In conclusion, four promising, putative SFRP1-mimetics, namely 064, 366, 848 and 994 were identified in the present study that have the potential to be used as initial substances for the development of new drugs for the therapy of SFRP1-deficient mammary carcinomas