Characterizing the in vivo mechanism of action of vitamin D in the TRAMP model: A bioinformatics based approach
Kirk, Jason S.
MetadataShow full item record
Prostate cancer (CaP) is currently the most common form of non-cutaneous cancer with an estimated 217,730 new cases and 32,050 cancer deaths in American men for 2010. Causes for CaP are poorly understood but family history, diet, race, and geographic location are all documented contributors to risk. Current therapeutic options for localized early stage CaP include surgical resection and radiation treatment. If these forms of treatment fail, hormone ablation is the second line of defense. However, ablation therapy in most cases fails. Once CaP advances to become hormone insensitive, treatment options are limited. Because of limited therapeutic options, and due to the observed anti-neoplastic activity of vitamin D, there have been many studies investigating its therapeutic potential in prostate cancer. Calcitriol is the most active form of vitamin D with actions being mediated by the vitamin D receptor (VDR), a steroid nuclear receptor responsible for regulating transcription of vitamin D target genes. Activating vitamin D signaling pathways can lead to induction of apoptosis, cell cycle arrest, differentiation, and inhibition of metastasis and angiogenesis, making it a very attractive anti-neoplastic agent. In addition to treating advanced stage prostate cancer, vitamin D is also an attractive chemopreventative agent because it is a compound naturally present within the body. Vitamin D physiologically regulates calcium and phosphate levels, but studies also suggest a role in maintaining cellular differentiation and preventing/lowering the risk of some forms of cancer, including prostate. Prevention studies from our lab demonstrated a dual biology associated with in vivo calcitriol treatment in the transgenic adenocarcinoma of mouse prostate (TRAMP) model. When mice were treated from 4 to 18 weeks of age calcitriol inhibited tumor weight. However, when treatment continued past 18 weeks of age, calcitriol treatment resulted in increased distant organ metastasis. These contradictory results led us to hypothesize that calcitriol produces dual responses that are dependent on the stage of CaP progression. Therefore, we used Affymetrix Mouse Gene 1.0 St expression arrays to characterize TRAMP mRNA expression profiles (transcriptomes) after in vivo calcitriol treatment. Characterization of calcitriol response was performed in both Early and Late Stage TRAMP disease. Early Stage response was characterized by treating 10 week old TRAMP mice 3 times in one week on a Monday/Wednesday/Friday (MWF) schedule, and harvesting tissue 12 and 24 hours post treatment. Late Stage (poorly-differentiated palpable tumors, 20–25 weeks of age) calcitriol response was characterized by both acute and chronic calcitriol treatments. Acute response was assessed by treating TRAMP mice with Late Stage disease 3 times per week on a MWF schedule. Chronic response was assessed by treating TRAMP mice 3 times per week on a MWF schedule from 4 weeks of age until development of Late Stage palpable tumors, all tissue characterizing Late Stage response (acute and chronic) was harvested 24 hours post treatment. In addition to analyzing Early and Late Stage calcitriol response, we also characterized changes in expression profiles associated with TRAMP progression from Early (10 weeks of age) to Late Stage (poorly-differentiated palpable tumor, 20–25 weeks of age) disease. Progression analysis was done to better understand the potential differences in calcitriol response between Early and Late Stage time points. Expression arrays were validated using various combinations of quantitative real-time PCR (qRT-PCR), Western blotting, and gene set enrichment analysis (GSEA). In order to aid in identification of the mechanism of action associated with calcitriol treatment, expression datasets were analyzed using the online bioinformatics software developed by GeneGo (www.genego.com), which identified deregulation of key pathways and signaling molecules associated with calcitriol response and progression. Overall, we have demonstrated that chronic calcitriol treatment resulted in increased metastatic disease. Gene expression changes associated with TRAMP model progression suggest a loss of vitamin D responsiveness with disease progression, which is further supported by examination of VDR target gene activation. Analysis of results associated with calcitriol treatment suggested that calcitriol mediates growth control in TRAMP through deregulation of androgen receptor (AR) signaling, which subsequently down-regulates T-antigen (Tag) signaling, and inhibits TRAMP model progression. The regulation of AR signaling occurs at both Early and Late Stage time points during disease progression, but is lost with chronic treatment, suggesting developed resistance to growth control. No clear mechanism responsible for the increased metastatic incidence with chronic calcitriol treatment was found, but results are highly reproducible. Regulation of AR signaling by vitamin D has important clinical implications for prostate cancer treatment, suggesting that treatment of hormone refractory patients with calcitriol may not be the best approach. Furthermore, if the loss of vitamin D responsiveness with TRAMP progression also occurs in human disease, it would suggest that early intervention in disease progression might be more beneficial.