Our first aim is to define germline-encoded modifier loci for tumor progression and metastasis in the TRAMP model of prostate tumorigenesis. We are initially approaching this by crossing TRAMP mice to the eight progenitor strains of the Collaborative Cross (CC) recombinant inbred (RI) panel. The goal here is to prove that the introduction of germline polymorphism through breeding influences prostate tumor progression and metastasis in mice. These experiments are at an early stage, primarily because TRAMP mice are aged 30 weeks prior to euthanasia. However, preliminary data are proving potentially exciting. For example, TRAMP x C57BL/B6J F1 crosses show a low level of tumorigenicity (average tumor weight = 0.62 g +/- 1.08 g;n=5) and low frequency of metastasis (overt metastasis in 0/5). However, other crosses such as TRAMP x NOD/ShiLtJ show a high level of tumorigenicity (average tumor weight = 9.43 g +/- 6.10 g;n=6) and high frequency of metastasis (overt metastasis in 5/6 mice). A greater number of animals will be required to achieve statistical significance, and earlier studies have shown that approximately 25 mice per group will be required. The breeding phase of this experiment has been completed with over 25 TRAMP F1 mice being available for each of the eight CC progenitor strains. Following completion of this, we will define quantitative trait loci (QTLs) for these traits by crossing a mouse model of prostate cancer to CC RI strains. Ideally, these experiments will be performed using the Pten micro-RNA (miRNA) mouse model of prostate tumorigenesis (see below). However, the fact that we are observing profound differences in tumorigenesis and metastasis in the TRAMP F1 crosses raises the possibility of performing an F2 intercross to define QTLs. Such an approach would not only prove faster than using the CC x TRAMP F1 mice to characterize QTLs, but also would be much less costly. We plan to initiate these intercross experiments in upcoming months. However, criticism has been leveled at the TRAMP model since the molecular pathways involved in tumor initiation are somewhat different from those involved in human prostate cancer tumor initiation. To address this, we intend to develop a mouse model that more faithfully recapitulates the molecular mechanisms seen in human prostate cancer initiation. To this end, we are developing a transgenic mouse that over-expresses miRNAs targeting Phosphatase and Tensin homolog (Pten) within the prostatic epithelium. The first phase of this project has been to transiently transfect NIH/3T3 with a variety of Pten-specific miRNAs followed by quantification of Pten expression. These miRNAs are co-expressed with GFP from a RNA polymerase II (CMV) promoter within a commercially available vector backbone. Thus far, we have isolated a number of miRNAs that substantially reduce Pten levels. The next phase of this project will be to replace the CMV promoter with a prostate-specific promoter, and we intend to assess the efficacy of a variety of promoters and chose the sequence that induces the highest level of prostate-specific miRNA expression. Furthermore, we intend to replace the GFP co-expression marker with a luciferase reporter to facilitate quantification of in vivo bioluminescence in transgenic animals. Following completion of the construct, transgene DNA will be introduced into fertilized mouse oocytes and the phenotypic effects of the Pten-specific miRNA transgene assessed. We are continuing to assess the role of Ribosomal RNA Processing 1 Homolog B (Rrp1b) in tumor progression and metastasis in breast cancer. One approach we are using is to employ mouse models to assess the effects of Rrp1b dysregulation upon tumor progression and metastasis. We have utilized Rrp1b BAC transgenic mice to assess the effects of Rrp1b over-expression on polyoma middle T (PyMT)-induced mammary tumorigenesis. This year, we have demonstrated that Rrp1b is a novel metastasis suppressor (i.e. Rrp1b suppresses pulmonary metastasis but has no overall effect upon tumor growth kinetics) by breeding Rrp1b transgenic mice to PyMT transgenic mice. Furthermore, we are developing a conditional knockout mouse for Rrp1b in collaboration with Pamela Schwartzberg, PhD. Here we are using a recombineering-based approach to ablate mammary epithelial Rrp1b expression. We are in the process of developing targeting constructs to achieve this and expect to introduce these into mouse embryonic stem cells shortly. Rrp1b knockout mice will subsequently be crossed to PyMT mice to assess the effect of loss of Rrp1b expression upon PyMT-induced mammary tumorigenesis and metastasis. The impact of Rrp1b dysregulation on the in vitro and in vivo cellular behaviors of a variety of cell lines is also being assessed. We are utilizing mouse (Mvt1, 4T1) and human (MDA-MD-231) cell lines to evaluate the effects of ectopic expression and knockdown of Rrp1b. Of particular interest here are the differential effects of ectopic expression of the RRP1B C1421T allelic variants (rs9306160), the variant allele of which has be associated with an improved survival in multiple human breast cancer cohorts. These cell lines are in the final stages of production and we will shortly begin assessing the differences between them by performing in vitro invasion and migration assays, and by implanting them orthotopically into the mammary fat pads of syngenic and immunocompromised mice. Furthermore, we are also in the process of developing cell lines that over-express RRP1B deletion mutants as well as RRP1B knockdown cell lines. The functional properties will be assessed in a similar manner. We are using a number of approaches to assess the role of RRP1B in transcriptional regulation and alternative mRNA splicing. We have utilized ChIP-Seq to assess chromatin regions bound by RRP1B. Specifically, we have used an endogenous RRP1B antibody to immunoprecipitate RRP1B-bound chromatin and analyzed these sequences using next generation sequencing technology. Sequence analysis has been performed by the NIH Intramural Sequencing Core (NISC) and data analysis performed in collaboration with Tyra Wolfsberg, PhD and the NHGRI Bioinformatics and Scientific Programming Core. We have been able to identify numerous novel RRP1B-chromatin interactions and are in the process of confirming these interactions through ChIP-qPCR and qPCR gene expression analyses. A homologous recombination strategy is being employed to introduce a FLAG-tag into the endogenous RRP1B locus of MDA-MB-231 cells since the RRP1B-specific antibody has not been used previously for ChIP. The targeting vector for this has been completed and cells have been infected with adeno-associated viral particles produced from this vector. We expect these cell lines to be available in upcoming months, and intend to use them to confirm our initial findings. We are also using a homologous recombination-based knock-in approach to change the genotype of C1421T polymorphism in MDA-MB-231 cells. These cell lines, which will endogenously express the RRP1B 1421C wildtype allele instead of the 1421T allele seen in unaltered MDA-MB-231, will prove valuable in ChIP experiments and will allow the effects of the two allelic variants upon transcription to be assessed. Here, we are approaching completion of the knock-in targeting vector and expect to generate viral particles shortly. Finally, we intend to assess the consequences of RRP1B knockdown by using mRNA-Seq. Here, we expect that we will be able to use data from these experiments to assess the role of RRP1B in both transcription and alternative mRNA splicing. These cell lines are in the final stages of generation and we expect to be sending RNA samples to NISC imminently. Data analysis will be performed by the NHGRI Bioinformatics and Scientific Programming Core.
