Genetic instability, a hallmark feature of human cancers including prostatic adenocarcinomas, is considered a driver of metastasis. Somatic copy number alterations are found in most aggressive primary human prostate cancers, and the overall number of such changes is increased in metastases. Chromosome 10q23 deletions, encompassing PTEN, and amplification of 8q24, harboring MYC, are frequently observed, and the presence of both together portends a high risk of prostate cancer-specific mortality. In extant genetically engineered mouse prostate cancer models (GEMMs), isolated MYC overexpression or targeted Pten loss can each produce early prostate adenocarcinomas, but are not sufficient to induce genetic instability or metastases with high penetrance. While a previous study showed that combining Pten loss with focal MYC overexpression in a small fraction of prostatic epithelial cells exhibits cooperativity in GEMMs, additional targeted Tp53 disruption was required for formation of metastases. We hypothesized that driving combined MYC overexpression and Pten loss using recently characterized Hoxb13 transcriptional control elements that are active in prostate luminal epithelial cells would induce the development of genomic instability and aggressive disease with metastatic potential. Neoplastic lesions that developed with either MYC activation alone (Hoxb13-MYC) or Pten loss alone (Hoxb13-Cre|PtenFl/Fl) failed to progress beyond PIN and did not harbor genomic copy number alterations. By contrast, mice with both alterations (Hoxb13-MYC|Hoxb13-Cre|PtenFl/Fl or BMPC) developed lethal adenocarcinoma with distant metastases and widespread genome copy number alterations that were independent of forced disruption of Tp53 and telomere shortening. BMPC cancers lacked neuroendocrine or sarcomatoid differentiation, features uncommon in human disease but common in other models of prostate cancer that metastasize. These data show that combined MYC activation and Pten loss driven by the Hoxb13 regulatory locus synergize to induce genomic instability and aggressive prostate cancer that phenocopies the human disease at the histological and genomic levels.

NKX3.1 is a homeodomain protein that functions as a dosage sensitive prostate-specific transcription factor. Diminished NKX3.1 expression is associated with prostate epithelial cell proliferation in vitro and with increasing Gleason grade in patient samples. Mouse Nkx3.1 also functions as a negative regulator of prostate cell growth in prostate cancer models. Identifying biological and environmental factors that modulate NKX3.1 accumulation is therefore central to efforts aimed at elucidating prostate growth control mechanisms. To determine the effect of inflammation on Nxk3.1 accumulation, bacterial prostatitis was induced by intraurethral inoculation of a uropathogenic E. coli strain in mice. Nkx3.1 expression was profoundly reduced in infected prostate lobes and correlated with increased expression of a proliferation marker. Androgen receptor levels were also reduced in concert with Nkx3.1, and a marked increase in the basal cell marker p63 was observed. Analyses of the inflammatory infiltrate revealed a classic acute inflammatory response that attained characteristics of a chronic state within fourteen days postinoculation. Comparison of the four prostate lobes revealed clear differences in the extent of inflammation. These data demonstrate that acute inflammation in response to a bacterial agent in the prostate is associated with a significant diminution in the level of a key regulator of prostate cell proliferation. These observations provide a plausible mechanism whereby prostate inflammation may establish a local environment conducive to epithelial cell growth.

The spatial distribution and early postnatal onset of Vpp1 expression is consistent with a role for this gene in prostate regionalization. The absolute prostate specificity of Vpp1 expression may allow this gene to serve as a paradigm to study the molecular basis of gene expression that is restricted exclusively to the prostate gland.

There is a pressing need for animal models that recapitulate key pathological features of lethal human prostate adenocarcinoma. The availability of these models would facilitate studies to elucidate the molecular events underlying disease progression, and would be invaluable for evaluating new treatment efficacy. Within extant mouse models, metastasis is rare and typically occurs at an advanced age. Using a bacterial artificial chromosome-based transgenic approach, we generated mice in which Hoxb13 regulatory elements drive expression of human MYC (Hoxb13MYC+) or Cre recombinase (Hoxb13Cre+). Hoxb13Cre+ mice were bred to mice carrying floxed Pten (Ptenfl) alleles to generate focal loss of Pten in the prostate epithelium. We then generated mice carrying both the Hoxb13MYC+ and Hoxb13Cre+ transgenes that also harbored two Ptenfl alleles. Phenotypic analyses of Hoxb13MYC+|Hoxb13Cre+| Ptenfl/fl males revealed pervasive low-grade mouse-PIN (mPIN) in all prostate lobes emerging at two weeks of age that became pervasive that progressed to high-grade mPINby four weeks. The lesions bore hallmarks of human PIN including increased nuclear and nucleolar size and dispersed chromatin. Immunostaining revealed MYC and androgen receptor expression and focal Pten loss within the epithelial compartment. mPIN incidence high-gradewas higher in ventral and anterior lobes. By twelve weeks, lesions progressed to cribriform mPIN/intraductal carcinoma was widespread and scattered intraductal carcinoma lesions were observed. At sixteen weeks, in all animals, large frank invasive adenocarcinoma lesions involving multiple prostate lobes were was present as well as, and histological analyses confirmed the presence of invasive poorly differentiated adenocarcinoma. Interestingly, by 16 weeks, metastasis to pelvic and peritoneal lymph nodes occurred with 100% penetrance. In survival analyses, all Hoxb13MYC+|Hoxb13Cre+| Ptenfl/fl mice (N=12) reached criteria for euthanasia within 28 weeks. Liver and lung metastases were observed with high penetrance (>80%). In one case, metastatic spread to a thoracic vertebra was observed, which may represent the first hematogenous metastasis to bone in a mouse prostate cancer model. Copy number analyses using array-based comparative genomic hybridization revealed clonal gene gains and losses in primary prostate cancers and matched metastatic lesions, indicative of genomic instability. The incidence of aggressive cancer and the reproducible chronology of events will enable this model to serve as a robust platform to test strategies to interfere with disease initiation and progression. The lethal phenotype will permit analyses of potentially life extending therapies. In addition, these features will enable studies to dissect the pivotal molecular events that drive the stepwise progression/evolution of this disease. Citation Format: Gretchen K. Hubbard, Laura N. Mutton, May Khalili, Ryan P. McMullin, Jessica Hicks, Daniella Bianchi-Frias, Peter S. Nelson, Srinivasan Yegnasubramanian, Angelo M. De Marzo, Charles J. Bieberich. MYC overexpression combined with Pten loss generates genomic instability and rapid metastasis in a new mouse model of lethal prostate adenocarcinoma. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 1086. doi:10.1158/1538-7445.AM2013-1086

<p>Supplementary Figure 1 Morphology of PIN in B13-MYC mice. Supplementary Figure 2 Phenotypic analysis of PIN in a B13-MYC mouse (ventral lobe, 52 weeks old). Supplementary Figure 3 Additional phenotypic analysis of B-MYC/Pten mouse PIN (anterior lobe, 8 weeks). Supplementary Figure 4 Invasive adenocarcinoma lesion in a 14 week old B-MYC/Pten animal. Supplementary Figure 5 Chromogranin (a) staining in a primary tumor and Foxa2 (b) staining in a lymph node metastasis from a 24 week-old B-MYC/Pten animal shows focal scattered tumor cells staining positively (arrows). Supplementary Figure 6 Morphological features of castrated B-MYC/Pten animals. Supplementary Figure 7. Genes within regions of copy number gains and losses are enriched in specific gene ontology (GO) gene sets. Supplementary Table 1. Prostate Phenotype in B-MYC/PtenFl/Fl mice and wild type (FVB/N) mice. Supplementary Table 2. Metastatic Spread in B-MYC/PtenFl/Fl mice and wild type (FVB/N) mice. Supplementary Table 3. Prostate Phenotype in B-MYC/PtenFl/+ mice and wild type mice. Supplementary Table 4. Effects of androgen withdrawal/castration in B-MYC/PtenFl/Fl mice and wild type (FVB/N) mice. Supplementary Table 5. Metastatic Spread in B-MYC/PtenFl/Fl mice after androgen withdrawal. Supplementary Table 6. Somatic copy number gains and losses in each sample. Supplementary Table 7. Regions of Copy Number alterations in each sample. Supplementary Table 8. Regions with recurrent gains and losses in at least two mice. Supplementary Table 9. Regions of high copy number gain and low copy number loss in any tumor specimen.</p>