The role of TOP3B in AR-mediated transcription and genomic stability in prostate cancer.

Prostate cancer is the second most commonly diagnosed cancer in men, with an estimated 1.1 million diagnoses worldwide in 2012, accounting for 15% of all cancers diagnosed. The clinical diagnosis of prostate cancer is based on three clinical features, which are the digital rectal examination (DRE), prostate-specific antigen (PSA) measurements and prostate biopsies. When diagnosed, prostate cancer is known to be a disease with highly variable outcomes, ranging from indolent localized to highly aggressive metastatic prostate cancer. Both clinically and biologically, we still have limited knowledge what really determines the aggressive behavior of this disease, limiting our possibilities to apply personalized medicine to those patients with lethal disease. The hypotheses formed in this thesis are based on this gap in our knowledge on prostate cancer and the goal of our research was to get a better insight into the clinical and molecular drivers of the lethal prostate cancer phenotype.

For this purpose, in the first part (chapter 3 and 4) of this thesis, we performed two systematic reviews addressing this issue. In the first systematic review, we evaluated whether a negative multiparametric MRI of the prostate could exclude significant prostate cancer on biopsy. This would be highly relevant in the diagnostic setting, since in an ideal world, clinicians would only detect clinically relevant tumors. If a biomarker (such as an MRI) could determine with high certainty that no clinically relevant tumor will be found on biopsy, then performing an invasive test such as transrectal biopsies would be superfluous. Although the question seems simple, based on our review the answer was not that clear. One major problem is that there is no consensus on what significant prostate cancer is, leading to a wide variety of applied definitions in different studies. Based on current evidence, a multiparametric MRI as triage test to decide whether or not to omit biopsies can therefore still not be recommended. Once patients are treated for what is believed to be significant prostate cancer, patients are followed-up by repeated PSA measurements, which is a prostate-specific biomarker. If during follow-up the PSA value rises to a certain threshold, patients have biochemically recurrent disease. However, PSA recurrence is only a surrogate endpoint and its association with hard endpoints, such as the development of distant metastases or prostate cancer specific mortality, is still debated. In light of this, we performed the second systematic review to evaluate whether patients with PSA recurrence have worse survival and if so, which clinical features are most prognostic. Based on this review, we can conclude that PSA recurrence is indeed associated with hard endpoints, but the effect size (the increase in risk when a patient has PSA recurrence) is limited. Some tumor specific factors can indeed prdict the risk of mortality in patients with PSA recurrence, but again its predictive power remains limited.

For this purpose, in the second part (chapter 5) we evaluated whether the Decipher® genomic test, which is a test based on the expression of 22 tumor related genes, could better predict which patients have a poor prognosis after primary treatment with surgery. It was hypothesized that a test that is based on the expression of multiple cancer related genes can better reflect tumor biology when compared to clinical tumor features. For this purpose, we collected samples of two clinically identical cohorts with opposite outcomes (one group developed metastases, the second group did not despite long term follow-up). Therefore, in this study, patients’’ outcomes could not be distinguished based on clinical features. Despite this, indeed we could see that the Decipher® test could still predict whether a patient would develop metastatic recurrence at 10 years follow-up after primary treatment with surgery. As expected, biomarkers that reflect tumor biology have a great potential of identifying patients with lethal disease even when clinical features seem to disagree.

Clearly, developing a better understanding of tumor biology is the basis for the development of better diagnostic, therapeutic and prognostic tools. This was our incentive to perform the studies as described in the third and final part (chapter 6 and 7). In chapter 6, we performed an integrated analysis of tumor copy number aberrations and tumor gene expression, which led us to the identification of AZIN1 as a potential driver of metastatic progression in prostate cancer. Indeed, our in vitro analyses confirmed that modulating AZIN1 expression levels leads to altered proliferation and migration rates, which are features specific for a metastatic phenotype. Furthermore, based on our current experimental results, we hypothesize that AZIN1 exerts its effects by changing the extracellular matrix composition, which is essential for cells to be able to escape the confinement of the prostate. Future experiments will further investigate this research path.

Simplified, the transition of a normal cell to a cancer cell is generally based on the accumulation of genetic aberrations as a result of DNA damage. Therefore, identifying the causes of DNA damage in cancer development are of major interest. Based on a unique prostate cancer case that showed an enormous amount of copy number aberrations and a mutation in TOP3B (a mutation that was predicted to be deleterious to protein function), in chapter 7 we hypothesized that TOP3B is essential in maintaining genomic stability by preventing the accumulation of R-loops during transcription. R-loops are three stranded nucleic acid structures that arise during transcription and are very prone to DNA damage. Based on our research, we showed that in prostate cancer cells, activating androgen-receptor mediated transcription leads to an enormous increase in R-loop formation, which is even more dramatic upon knockdown of TOP3B. These data have paved the way for further research on the effects of TOP3B and R-loops in the development of DNA damage and carcinogenesis.

Collectively, these studies provide a better understanding on the clinical and molecular aspects of significant prostate cancer. Based on our data, we have identified both AZIN1 and TOP3B as new lines of prostate cancer related research.