Dissertation

Identification of Candidate Gene Markers for Biochemical Recurrence in Prostate Cancer

Horning, Aaron Marks

2017

Overview

The typical first-line therapy for prostate cancer is to remove the entire prostate and the tumor within. This radical prostatectomy, however, suffers from over-diagnosis. Approximately 30% of patients who undergo radical prostatectomy will be diagnosed with biochemical recurrent prostate cancer, a rising prostate specific antigen level indicating failure from treatment and eventual metastatic disease. Because the current biomarkers for prostate cancer prognosis are not completely accurate, it is difficult to distinguish between patients who will experience aggressive or indolent disease. Furthermore, the best predictive marker available, Gleason Score, requires an invasive needle biopsy which suffers from sampling bias and can underdiagnose a patient if the tumor in the prostate is missed. Therefore, investigating new non-invasive biomarkers would be beneficial. Also, because prostate cancer exhibits intercellular heterogeneity, characterizing subpopulations may reveal new information about cancer progression and recurrence as well. To discover non-invasive markers for prostate cancer recurrence, I investigated the utility of methylated cell free DNA. Altered DNA methylation patterns at promoter region CpG islands are characteristic of prostate cancer progression. Also, androgen biosynthesis pathway enzymes are known to be expressed differently in prostate cancer tumors with different clinical outcomes. Prostate cancer tumors which became recurrent (n=12) or non-recurrent (n=15), exhibited significantly different levels of methylation of many of the androgen biosynthesis genes’ promoter regions as assessed by methyl-binding domain capture sequencing (p<0.05). Two of the three genes (SRD5A2 and CYP11A1 ) which showed increased methylation in recurrent tumors also showed a significant negative correlation with their gene expression levels in RNA-sequencing data obtained from The Cancer Genome Atlas (p<0.05). The change in methylation of cell free DNA of SRD5A2 and CYP11A1, and the patient’s neoadjuvant therapy history, were useful predictive markers for determining whether or not a patient would have experienced recurrent prostate cancer. Also, because prostate cancer exhibits intercellular heterogeneity, understanding the nature of the cellular subpopulations may reveal new information about how this cancer progresses and recurs as well. Prostate cancer recurrence may be caused by a clonal expansion of a small subpopulation of cells within the original tumor which evade initial therapy. To investigate tumor heterogeneity, a total of 144 single LNCaP cells were analyzed with single-cell RNA-sequencing. A total of 397 distinct genes exhibited significant differences between 0 and 12 hours without androgen and 12 hours with and 12 hours without androgen (7.13×10-10 ≤ p ≤ 0.05). Interestingly, the 144 cells clustered into 8 subclusters, 5 of which contained androgen-treated cells, indicating a heterogeneous response to androgen treatment which would not have been detected in previous bulk-cell experiments. Each of the 5 subclusters exhibited increased activity of distinct molecular pathways. Of these 5 subclusters, one included genes which distinguished recurrent from non-recurrent prostate cancer and showed increased expression of mitosis- and cell cycle- related pathways. When similar cells were isolated from LNCaP, they exhibited some stem-like characteristics which are still being validated. It is possible that this subpopulation of cells could be an early marker for an increased likelihood of developing biochemical recurrence. Altogether, these results suggest a clonal population model for the development of recurrent prostate cancer. Furthermore, these results show the importance of single cell analysis in developing our understanding of cancer progression and biochemical recurrence. In conclusion, these two projects provide alternative approaches to developing predictive markers for biochemical recurrence in prostate cancer. These results emphasize the utility of studying methylated cell free DNA and single cells for capturing the dynamic changes and heterogeneity exhibited by prostate and other cancers.

• Publisher: ProQuest Dissertations & Theses
• Subject: Bioinformatics / Biology / Cellular biology
• ISBN: 9781369813982, 1369813988