Prostate cancer is the most common cancer affecting men today. Therefore, there is a strong need for accurate biomarkers and successful therapeutic treatments. A novel approach combining a computationally built protein-protein interaction network of proven microRNA protein targets with high throughput proteomics identified ErbB2 and ErbB3 as key proteins in prostate cancer. These results coupled with microRNA array screening of an androgen-independent prostate cancer progression model, substantiated by single microRNA analysis, suggested miR125b as a key tumor suppressor contributing to prostate cancer progression. miR125b expression was shown to be substantially increased in the non-tumorigenic P69 cell line compared to its highly tumorigenic, metastatic M12 variant. Luciferase reporter gene assays including the entire 3’UTR of either ErbB2 or ErbB3 revealed a 2.8- and 2.4-fold decrease (respectively) compared to control vector. Thus, this combinatorial approach has suggested an additional microRNA and its target involved in prostate tumor progression.

Prostate cancer is the second most common cancer in the United States that affects men today. To better treat this disease accurate biomarkers and successful therapeutic treatments are needed. A novel approach to understand the mechanisms behind prostate cancer tumor formation lies in identifying dysregulated micro-RNAs (miRNAs), which are a class of small (18-24 nucleotides) non-coding RNAs that regulate gene expression posttranscriptionally by either inhibiting protein synthesis or signaling messenger-RNA for degradation. Multiple miRNAs were discovered in our highly tumorigenic and metastatic prostate cancer progression model M12 cell line compared to its weakly tumorigenic P69 parental cell line. Various analyses such as human panel analyses, single-miR analyses and patient tumor biopsy samples were analyzed to determine dysregulated miRNAs that contributed to the progression and metastasis of prostate cancer. Together with performing experiments to identify miRNAs, a de novo next generation sequencing approach was applied to identify miRNAs naturally present in biological fluids of normal and healthy subjects. Since, these miRNAs are highly dysregulated in many diseases, including cancer, they can act as potential biomarkers or therapeutic targets to improve treatments for prostate cancer. Essential miRNAs studied for this research were miR-17-3p that is known to target the ErbB2 mRNA; miR-299-5p that directly targets osteopontin (OPN) mRNA, and miR-147b that directly targets many mRNAs, such as COL4A2, ALDH5A1, NDUFA4, SDHD, and IER5. A wide range of miRNAs were identified in six biological fluids: venous blood, menstrual blood, vaginal fluid, semen, saliva, and feces. There were some miRNAs that were common to all 6 body fluids, some unique to each body fluid, and some miRNAs that literature suggested could potentially be biomarkers or normalizers for body fluid characterization.

In radiation therapy, interfraction organ motion introduces a level of geometric uncertainty into the planning process. Plans, which are typically based upon a single instance of anatomy, must be robust against daily anatomical variations. For this problem, a model of the magnitude, direction, and likelihood of deformation is useful. In this thesis, principal component analysis (PCA) is used to statistically model the 3D organ motion for 19 prostate cancer patients, each with 8-13 fractional computed tomography (CT) images. Deformable image registration and the resultant displacement vector fields (DVFs) are used to quantify the interfraction systematic and random motion. By applying the PCA technique to the random DVFs, principal modes of random tissue deformation were determined for each patient, and a method for sampling synthetic random DVFs was developed. The PCA model was then extended to describe the principal modes of systematic and random organ motion for the population of patients. A leave-one-out study tested both the systematic and random motion model’s ability to represent PCA training set DVFs. The random and systematic DVF PCA models allowed the reconstruction of these data with absolute mean errors between 0.5-0.9 mm and 1-2 mm, respectively. To the best of the author’s knowledge, this study is the first successful effort to build a fully 3D statistical PCA model of systematic tissue deformation in a population of patients. By sampling synthetic systematic and random errors, organ occupancy maps were created for bony and prostate-centroid patient setup processes. By thresholding these maps, PCA-based planning target volume (PTV) was created and tested against conventional margin recipes (van Herk for bony alignment and 5 mm fixed [3 mm posterior] margin for centroid alignment) in a virtual clinical trial for low-risk prostate cancer. Deformably accumulated delivered dose served as a surrogate for clinical outcome. For the bony landmark setup subtrial, the PCA PTV significantly (p30, D20, and D5 to bladder and D50 to rectum, while increasing rectal D20 and D5. For the centroid-aligned setup, the PCA PTV significantly reduced all bladder DVH metrics and trended to lower rectal toxicity metrics. All PTVs covered the prostate with the prescription dose.

The chemokine receptor CCR5 (CCR5) plays an integral role within the inflammatory network of cells. Importantly, CCR5 is a mediator in several disease states and can be targeted using small molecule antagonists. Within this work, CCR5’s role in prostate cancer and HIV/AIDS has been exploited in order to develop potential therapeutics and probes. First, a series of novel compounds was designed by using pharmacophore-based drug design based upon known CCR5 antagonists and molecular modeling studies of the CCR5 receptor’s three-dimensional conformation. Once synthesized, these compounds were tested for their CCR5 antagonism and their anti-proliferative effects in several prostate cancer cell lines. The data from both the calcium mobilization studies and the anti-proliferation studies suggests that the compounds synthesized have activity as CCR5 antagonists and as anti-proliferative agents in certain prostate cancer cell lines. In addition, a bivalent ligand containing both a mu opioid receptor (MOR) and a CCR5 antagonist pharmacophore was designed and synthesized in order to study the pharmacological profile of the putative CCR5-MOR heterodimer and its relation with NeuroAIDS. The structural-activity relationship between the bivalent ligand and the heterodimer was studied with radio-ligand binding assays, functional assays, HIV-1 fusion assays, cell fusion assays, and in silico molecular dynamics. The subsequent bivalent ligand was proven to be a potent inhibitor in both an artificial cell fusion assay mimicking HIV invasion and a native HIV-1 invasion assay using live virus. In all, two novel sets of compounds were synthesized that targeted either CCR5 or the CCR5-MOR heterodimer. For the CCR5 antagonists, as leads for prostate cancer therapeutics, further work needs to be done to ascertain and develop their structure-activity-relationship. This library of novel compounds was shown as promising leads as CCR5 and anti-prostate cancer agents. The bivalent ligand targeting the CCR5-MOR heterodimer proved to be a potent and tissue-specific inhibitor for neuroAIDS where the known treatment, maraviroc, is less efficacious and fails to inhibit virus entry in the presence of morphine. Both projects illustrate the roles that CCR5 plays in these two unique diseases.

This is a compilation of 3 abstracts for the three manuscripts included in this dissertation. I. Meta-Analysis: Racial Disparities in Prostate Cancer Survival: Prostate cancer is the second leading cause of cancer-related mortality in men. Previous studies have drawn inconsistent conclusions on racial differences in prostate cancer survival. This meta-analysis was conducted to investigate the relationship between race and survival from prostate cancer. A systematic review of published articles from 1968 to 2007 assessing survival from prostate cancer among African American and White men was conducted. The search yielded 20 eligible published manuscripts. Analysis of unadjusted studies showed African American men have an increased risk of all-cause mortality (Hazard ratio (HR) = 1.47, 95% confidence interval (CI): 1.31, 1.65, P < 0.001). However, examination of adjusted studies showed no difference (HR = 1.07, 95% CI: 0.94, 1.22, P = 0.308). No statistically significant difference was observed in prostate cancer-specific survival in both analyses using unadjusted (HR = 1.11, 95% CI: 0.94, 1.31, P = 0.209) and adjusted studies (HR = 1.15, 95% CI: 0.95, 1.41, P = 0.157). There was evidence of heterogeneity that was unexplained by factors analyzed in overall survival but explained by stage in prostate cancer-specific survival. This meta-analysis concludes that there are no racial differences in the overall and prostate cancer-specific survival between African American and White men. II. Case-Control study: Association between Family History of Cancers and Prostate Cancer: Family history of prostate cancer is an established risk factor for prostate cancer. However, the relationship between family history of cancers other than prostate cancer and prostate cancer risk is inconclusive. This study sought to examine the association between family history of cancers and prostate cancer. A case-control study was conducted in which cases and controls were randomly selected from a large urology clinic in Central Virginia. Cases were 600 histologically confirmed prostate cancer patients who were diagnosed between January 2000 and December 2005, and controls were 686 patients who visited the clinic during the same period and diagnosed with urological illnesses other than cancers and prostate-related problems. Data on family history of cancers, lifestyle and demographic factors were collected. Unconditional logistic regression analysis was used to estimate the odds ratios and the corresponding 95% confidence intervals after adjustment for potential confounding factors. Multiple comparisons adjustments were made using Bonferroni adjustment. Men with family history of any cancer in first-degree relatives including parents (OR=2.42, 95% CI: 1.53, 3.84) and parents only (OR=1.90, 95% CI: 1.23, 2.94) were at increased risk of developing prostate cancer compared to men with no such family history of cancer. Significant increased risk was also observed with family history of prostate cancer in first-degree relatives (OR=2.68, 95% CI: 1.53, 4.69) and parents only (OR=3.26, 95% CI: 1.71, 6.24) compared to men with no family history of prostate cancer. Even after adjustments for multiple comparisons, the significance persisted both in first-degree relatives (OR=2.68, 95% CI: 1.16, 6.21) and parents alone (OR=3.26, 95% CI: 1.24, 8.63). No association was found with family history of other cancers including breast, colon, lung, skin, digestive tract, stomach, liver, pancreas, female cancers, urogenital, urinary bladder, brain, blood and lymph node and other cancers and risk of prostate cancer. This study demonstrated an increased prostate cancer risk for men with a family history of any cancer or prostate cancer in first-degree relatives including parents and parents alone. Health care providers need to be aware of the potential risk of family history of cancers on prostate cancer. III. Case-Control study: Association between Obesity and Prostate Cancer: Obesity is a major public health problem in the United States. Several studies have investigated the association between obesity and prostate cancer risk. However the impact of early-adult obesity on prostate cancer is not well studied. This study proposes to investigate the relationship between prostate cancer and early-onset obesity and current obesity. A case-control study was conducted to investigate the relationship between obesity and prostate cancer in a large urology clinic population in Central Virginia. Cases included histologically confirmed prostate cancer patients of all stages and grades diagnosed from January 2000 to December 2005. Controls were patients who were diagnosed with urological illness other than cancers and prostate-related problems. Self-reported data was collected on anthropometric, lifestyle and demographic factors through a mail survey. Unconditional logistic regression analysis was conducted to investigate the association between prostate cancer and early-onset obesity (BMI at age 18) and current obesity. Odds ratios and corresponding 95% confidence intervals were calculated after accounting for significant interaction terms and adjusting for potential confounding variables. This study showed statistically significant association between BMI at age 18 and prostate cancer risk in the multivariate analysis when BMI was evaluated as a continuous variable. There was a 7% decrease in the odds of prostate cancer risk for every 1 kg/m(2) increment in BMI at age 18 (OR=0.93, 95% CI: 0.87, 0.98). Analysis of BMI at age 18 as a categorical variable also showed reduced risk though statistically non-significant. Obese men (OR=0.62, 95% CI: 0.12, 3.08) and overweight men (OR=0.60, 95% CI: 0.35, 1.05) had a non-significant decreased risk of developing prostate cancer compared to normal weight men at age 18. Examination of current BMI showed a non-statistically significant decreased risk of prostate cancer when examined as a continuous variable. However, there was significant interaction between current BMI treated categorically and age. This study concludes that there is decreased prostate cancer risk associated with increasing BMI at age 18. Future large prospective studies are needed to better understand the association between early-onset obesity and risk of prostate cancer and explore the biological factors associated especially in the early ages. This document was created in Microsoft Word 2003.

Breast and prostate cancer incidence and mortality have been steadily decreasing. Reasons for these reductions may be related to increased rates of cancer screening and other factors such as improvements in diet, including consumption of fruits and vegetables. We wanted to determine if individuals who get screened for breast and prostate cancer are more or less likely to consume adequate servings of fruit and vegetables. A cross-sectional study using the BRFSS survey was conducted. Individuals included in this study (n=26,222), were asked about their breast or prostate cancer screening history. They were also asked about their servings per day of fruit and vegetables. Statistical analyses were conducted using the SAS 9.2 software program. Logistic regression analyses were conducted on the variables and potential confounders. Over 40% of individuals who did not screen for breast and prostate cancer were in the 50-59 years of age category. A trend was seen with younger age groups being less likely to consume 3 or more daily servings of fruit and vegetables than their older counterparts. Another trend was seen in education levels. Individuals with lower education were less likely to consume at least 3 daily servings of fruit and vegetables. There was a statistically significant association between cancer screening and servings of fruit and vegetables per day. Individuals who were screened for either breast or prostate cancer were 52% more likely to consume 3 or more servings of fruit and vegetables than those who did not screen for either breast or prostate cancer (OR=1.52, 95% CI: 1.29-1.79). Further research needs to be conducted related to how other health behaviors may be related to cancer screening adherence and fruit/vegetable intake.

Abstract 1 – A Geographic Information System for Evaluating Residential Pesticide Exposure and Prostate Cancer Incidence Agricultural pesticide exposure is hypothesized to be a risk factor for prostate cancer, and such exposures are of particular concern for men living in farming communities where large-scale pesticide applications occur. Prostate cancer incidence data were obtained from the State Health Registry of Iowa for the years 1996 through 2006, and county and census tract level age-adjusted incidence rates were calculated. Historical crop-specific land use records and pesticide sales data for the state of Iowa during 1990 were integrated into a geographic information system (GIS), where estimates of predicted exposure to the four most commonly used pesticides in Iowa (atrazine, metolachlor, cyanazine, alachlor) were produced. Ecological correlation between pesticide exposure and prostate cancer incidence was evaluated using Spearman’s (rank) correlation coefficient and linear regression analysis. Statistically significant associations between prostate cancer incidence and percent of acres of corn and soybean crops were found at both the county (r=0.22, p=.031 and r=0.33, p=.001, respectively) and census tract (r=0.10, p=.007 and r=0.13, p<.001, respectively) level. The associations between percent of land exposed to the specific pesticides and prostate cancer were not statistically significant. Our findings suggest that residential proximity to corn and soybean fields, and by association the pesticides used on those crops, is correlated with increased prostate cancer risk, but that the increase in risk is not correlated with exposure to the four most commonly used pesticides in Iowa in 1990. Findings from this study underscore the need for continued investigation of the association between agricultural exposures and prostate cancer incidence. Abstract 2 – Spatial Analysis of Prostate Cancer Incidence and Residential Pesticide Exposure in Iowa A statistically significant positive association between prostate cancer incidence and residential proximity to corn and soybean fields in Iowa exists. Research suggests that exposure to pesticides used on these crops increases prostate cancer risk. The objective of this study was to investigate clustering of prostate cancer risk in the presence of potential exposure to pesticides in Iowa. Prostate cancer incidence data (1996-2006) were obtained from the State Health Registry of Iowa. Using SaTScan software, clusters of high and low prostate cancer risk were identified. Ecological correlation between exposure to the four most commonly used pesticides (atrazine, metolachlor, cyanazine, alachlor) in Iowa during 1990 and residence in a cluster of relatively high or low prostate cancer incidence was evaluated using Pearson’s chi-square test statistic and logistic regression analysis. Clusters of increased prostate cancer risk were associated with a greater percentage of land used for all crops of interest (i.e., corn and soybean farming (p <0.001), corn farming (p <0.001), soybean farming (p <0.001)) and low exposure to alachlor (p =0.032) than did clusters with decreased risk of prostate cancer. After adjustment for percent of land used for each crop type, no association between pesticide exposure and prostate cancer risk was observed. Residence in or near agricultural communities increases prostate cancer risk. Our findings suggest that residential proximity to exposures specific to corn and soybean farming increases prostate cancer risk. Evaluation of exposure to less commonly used pesticides and those used in lower quantities is needed.   Abstract 3 – Multilevel Analysis of Residential Pesticide Exposure and Prostate Cancer Incidence An association between residential exposure to factors specific to corn and soybean farms in Iowa exists. The objectives of this study were to statistically assess spatial autocorrelation in prostate cancer incidence in Iowa and to evaluate the effect of residential exposure to the most commonly used pesticides for corn and soybean farms in Iowa in 1990 on prostate cancer incidence. Prostate cancer incidence data were obtained from the State Health Registry of Iowa for the years 1996 through 2006. Spatial patterning of age-adjusted incidence rates was assessed via Moran’s I global index of spatial autocorrelation. A hierarchical regression modeling approach with an assumed Poisson distribution was used to characterize the relationship between census tract level prostate cancer incidence and exposure to pesticides. Statistically significant spatial patterning of prostate cancer incidence, corn and soybean fields and pesticide use (p<.001 for all variables) was observed. After adjustment for individual and area level characteristics, prostate cancer risk increased by approximately 25% for each percentage point increase in percent of land used for corn and soybean crops. Prostate cancer risk was approximately 25% higher for Black men exposed to corn and soybean fields compared to white men exposed to corn and soybean fields. Results from this study support the need for further evaluation of residential exposure to environmental hazards specific to corn and soybean farming.

A unique computational approach was undertaken to identify targets of miR-17-3p that impart an oncogenic potential to the cells of the prostate. Utilizing this approach, we identified insulin growth factor receptor 1 (IGF1R) as a potential target of miR-17-3p. IGF1R imparts an oncogenic approach to the cells by helping cells escape apoptosis, become hypertrophic and increase the production of extracellular proteases that allow cells to detach from neighbors. The regulation of insulin growth factor receptor 1 by human microRNA-17-3p was evaluated using a western blot analysis of prostate cancer cell lines. Protein levels were compared in a cell line that expressed a non-targeting control RNA and a cell line that expressed microRNA-17-3p. The cell line that expressed the non-targeting control had significantly higher levels of IGF1R protein than the cell line expressing more of the active microRNA. Based on this experiment, it appears that microRNA-17-3p might regulate the insulin growth factor receptor 1.

Prostate cancer (PCa) cells predominantly metastasize to bone and the complex crosstalk between PCa cells and osteoblasts (bone-forming cells) and osteoclasts (bone-destroying cells) leads to increased tumour growth and worsening of bone disease. Understanding the mechanisms of PCa bone metastasis can identify the aggressive fraction of PCa resulting in earlier intervention. The ability of PCa cells to express bone cell-specific features, termed osteomimicry, could potentially explain the osteotropic nature of PCa cells. The aim of this study was to determine the role of osteomimicry in the regulation of epithelial-mesenchymal transition (EMT) in bone metastatic prostate cancer cells. It was demonstrated that the osteoblast-specific marker alkaline phosphatase (bone/liver/kidney) (ALPL) was overexpressed in bone metastatic (ARCaPM), compared to non-metastatic (ARCaPE), human PCa cells. Knockdown of ALPL resulted in decreased cell viability, increased cell death and a change from mesenchymal to epithelial morphology in ARCaPM and PC3 cells, and increased CDH1 expression along with decreased migration in ARCaPM cells. Treatment with extracellular ATP also resulted in decreased viability, increased expression of epithelial markers (CDH1, KRT14) and decreased expression of mesenchymal markers (VIM, ZEB1), and reduced expression of ALPL in ARCaPM cells. Small RNA-sequencing identified microRNAs differentially expressed between ARCaPE and ARCaPM PCa cell lines: miR-373 expression was lower in ARCaPM compared to ARCaPE cells and its overexpression in ARCaPM cells resulted in a change to epithelial morphology, increased expression of the epithelial marker CDH1 and decreased expression of the mesenchymal markers VIM and ZEB1. Finally, the development of a high-throughput screening method to identify novel microRNA regulators of osteomimicry was described, which identified two microRNAs miR-199a-5p and miR-212 as positive regulators of ALP activity. Taken together, this thesis describes the identification of ALPL and ATP as novel regulators of epithelial-mesenchymal transition in PCa cells and high-throughput ALP-activity screening as a powerful tool to identify novel microRNA regulators of ALP expression.

An instrument is being developed at the Department of Biomedical Engineering; Research and Development (MT-FoU), at the University Hospital of Umeå with the aim to detect prostate cancer ex vivo. Using a combination of tactile resonance technology and Raman spectroscopy the instrument is intended to be used in the operating room during radical prostatectomy to identify positive surgical margins. The hypothesis was that the length of the piezoelectric element used in the tactile resonance sensor affects the sensor's sensitivity and reproducibility when measuring the stiffness of soft tissue, and that there might be an optimal impression depth to measure at. The specific aim of this study was to evaluate two piezoelectric elements, of different lengths, by the sensitivity and reproducibility of the measurements they performed. Measurements were performed on five silicone samples of different stiffness, during a 2 mm impression. The standard deviation of the stiffness parameters, the R2 of the linear regression used to determine the stiffness parameter, and the depth at the which the most linear relationship between impression force and frequency shift was found were studied using linear mixed-effects models to identify any significant differences between the elements. The long element had a significantly higher R2 of 0.98 compared to 0.93 for the short element, and a higher measurement depth of 0.47 mm compared to 0.37 mm for the short element. No difference between the elements were found on accuracy as measured by standard deviation of the stiffness parameter. It was concluded that this was not enough to claim that one element was better than the other. / Ett instrument utvecklas på avdelningen för Medicinsk teknik, forskning och utveckling, vid Norrlands universitetssjukhus med målet att detektera prostatacancer ex vivo. Instrumentet kombinerar taktil resonansteknologi med Ramanspektroskopi och är tänkt att användas i operationssalen under radikal prostatektomi för att identifiera positiv kirurgisk marginal. Hypotesen var att längden av det piezoleketriska element som används i den taktila resonanssensorn påverkar sensorns känslighet och reproducerbarhet vid mätning av styvhet av mjukvävnad, och att det kan finnas ett optimalt intryckningsdjup att mäta på. Målet med denna studie var att utvärdera två piezoelektriska element, av olika längd, utifrån känsligheten och reproducerbarheten av mätningarna de utförde. Mätningarna gjordes på fem silikonsprover av olika styvhet, under 2 mm intryckning. Standardavvikelsen av styvhetsparametern, R2 av den linjära regression som användes för att bestämma styvhetsparametern, samt det intryckningsdjup på vilket det mest linjära förhållandet mellan intryckningskraft och frekvensskift hittades, studerade med så kallade linear mixed-effects modeller för att identifiera signifikanta skillnader mellan elementen. Det långa elementet hade ett signifikant högre R2 på 0.98 jämfört med det korta elementets 0.93, och ett högre mätdjup på 0.47 mm jämfört med det korta elementets 0.37 mm. Ingen skillnad mellan elementens standardavvikelser av styvhetsparametern hittades. Slutsatsen drogs att resultatet inte var nog för att påstå att det ena elementet är bättre än det andra.