Integrative Analysis of Ultra-Deep RNA-Seq Reveals Alternative Promoter Usage as a Mechanism of Activating Oncogenic Programmes During Prostate Cancer Progression - Beyond the Abstract

Despite effective treatments for localized prostate cancer, many patients eventually develop metastatic castration-resistant prostate cancer (mCRPC), a lethal form of the disease. Tumor cells adapt and escape therapy by altering their gene expression profiles through genetic and epigenetic changes.

Therefore, fully characterizing these alterations and identifying targetable vulnerabilities is critical to defining the biology of this disease and identifying new therapeutic targets. With generous funding from the Prostate Cancer Foundation and Stand Up To Cancer, and the dedicated support of patients and their families, the West Coast Dream Team (WCDT) has collected the largest cohort of mCRPC tumor biopsies for molecular characterization to date. We sequenced the transcriptome of 104 mCRPC tumor samples from the WCDT cohort to an unprecedented depth and revealed a previously overlooked regulatory mechanism of oncogenic programs. These biopsies were also analyzed for alterations to their genome, epigenome, chromatin interactome, and transcriptome landscape.1-6 These integrated data are an extremely valuable resource for the research community with multi-omics data in matching metastatic tumor biopsies.

Gene transcription can be compared to making a copy of a recipe from a cookbook. In cells, this process involves copying instructions from DNA (the cookbook) into a molecule called RNA (the recipe copy). This RNA copy carries the information needed to make proteins, which make up the components of cells and control much of their activities. Different numbers of RNA copies are made in different cells under different conditions and are directly associated with the amount of proteins produced. Transcription is carefully regulated in normal cells and can be hijacked by tumor cells to favor their growth and escape therapy. Transcription factors (TFs) and epigenetic modifications are the main regulators of transcription. We and others have previously shown how alterations in these regulators are linked to prostate tumors progression and resistance to AR-targeted therapies. However, how these altered regulators impact the tumor transcriptome remains unclear.

The most important way TFs affect cells is to bind to DNA sequences near the beginning of the genes at locations called promoters. Most human genes have more than one promoter, each initiating distinct isoforms of the same gene. Most isoforms of a gene driven by different promoters share the majority of their coding sequences and only differ in untranslated regions. But from a gene regulatory perspective, each promoter can be bound by distinct regulators because they each harbor distinct TF binding sites, chromatin accessibility, and DNA methylation status. Therefore, the selection of promoters is highly tissue-specific, and alternative promoter usage is an important mechanism for transcriptome diversity and gene expression regulation during development.7,8

However, scientists have only recently linked alternative promoter usage to cancer.9 The mechanisms by which alternative promoters are preferentially activated during tumor progression are not well-studied, and the impact of transcriptomic and epigenomic changes on alternative promoter use associated with tumor formation and progression is unclear. Our work studied ultra-deep RNA sequencing (RNA-seq) data in 274 biopsies of benign prostate tissue, localized prostate tumors, and mCRPC. We also improved a published method for estimating promoter activity using RNA-seq reads. This work characterized the alternative promoter usage landscape across the full spectrum of prostate cancer and revealed the underlying mechanism of alternative promoter activation during tumor progression.

We first observed an increase in alternative promoter usage in tumors. Although each gene has multiple promoters, not all of them are active at the same time. Compared to all genes, the genes with elevated expression levels in tumors and oncogenes were more likely to switch from only one promoter active in benign tissues to multiple promoters active in tumors. This demonstrated that the activation of additional alternative promoters was associated with gene overexpression. We identified statistically significant upregulated alternative promoters in localized prostate tumors and mCRPC, using the benign prostate tissue as the baseline, and found a much higher number of alternative promoters in mCRPC as compared to localized disease. This demonstrated that alternative promoter usage increases as tumors progress. We then asked to what extent alternative promoters contributed to gene overexpression and found that they were responsible for a disproportionately high amount of the increased gene activity compared to canonical promoters.

We then explored the mechanisms of alternative promoter activity by examining TF binding at alternative promoters using publicly available Chromatin Immunoprecipitation sequencing (ChIP-seq) data, and by studying DNA methylation status using sample-matched whole genome DNA methylation data. We first focused on the androgen receptor (AR), the key driver of prostate cancer. We observed a significant correlation between the expression of AR and elevated alternative promoter usage in individual tumors. By integrating the AR ChIP-seq data, we observed an enrichment of AR binding at the upregulated alternative promoters in localized tumors and mCRPC. Then, using an unbiased approach by exploring ChIP-seq data for more than 200 TFs, we identified AR and FOXA1 as the most enriched TFs at upregulated alternative promoters in localized tumors, and MYC, E2F1, and HIFα as the most enriched in mCRPC. Lastly, by integrating matched whole genome methylation data from the mCRPC biopsies, we showed that the DNA methylation at alternative promoters displayed a much larger variability as compared to that at canonical promoters, suggesting that DNA methylation at alternative promoters is more dynamic and plays an important role in regulating gene expression. By correlating alternative promoter activity and DNA methylation changes in matching samples, we discovered that DNA methylation is a likely mechanism for alternative promoter activation during tumor progression and lineage plasticity.

Overall, this work added a missing puzzle piece to the molecular characterization of mCRPC and depicted a clear linkage between the frequently altered TFs and DNA methylation to changes in the transcriptome. These findings advanced our understanding of the biology underlying prostate tumor progression on treatment and highlighted the importance of alternative promoters and the genes driven by alternative promoters as potential novel therapeutic targets for overcoming therapy resistance.

Screenshot_2025-01-06_at_1.24.17_PM.png

Written by:

  • David Quigley, PhD, Helen Diller Family Comprehensive Cancer Center, Department of Urology, Department of Epidemiology & Biostatistics, University of California at San Francisco, San Francisco, CA
  • Meng Zhang, PhD, Helen Diller Family Comprehensive Cancer Center, Department of Radiation Oncology, University of California at San Francisco, San Francisco, CA, USA.
References:

  1. Quigley, D. A. et al. Genomic Hallmarks and Structural Variation in Metastatic Prostate Cancer. Cell 174, 758-769.e759 (2018). https://doi.org/10.1016/j.cell.2018.06.039
  2. Aggarwal, R. R. et al. Whole-Genome and Transcriptional Analysis of Treatment-Emergent Small-Cell Neuroendocrine Prostate Cancer Demonstrates Intraclass Heterogeneity. Mol Cancer Res 17, 1235-1240 (2019). https://doi.org/10.1158/1541-7786.MCR-18-1101
  3. Zhao, S. G. et al. The DNA methylation landscape of advanced prostate cancer. Nat Genet 52, 778-789 (2020). https://doi.org/10.1038/s41588-020-0648-8
  4. Sjöström, M. et al. The 5-Hydroxymethylcytosine Landscape of Prostate Cancer. Cancer Res, Of1-of15 (2022). https://doi.org/10.1158/0008-5472.can-22-1123
  5. Zhao, S. G. et al. Integrated analyses highlight interactions between the three-dimensional genome and DNA, RNA, and epigenomic alterations in metastatic prostate cancer. Nat Genet (2024). https://doi.org/10.1038/s41588-024-01826-3
  6. Shrestha, R. et al. An Atlas of Accessible Chromatin in Advanced Prostate Cancer Reveals the Epigenetic Evolution during Tumor Progression. Cancer Res 84, 3086-3100 (2024). https://doi.org/10.1158/0008-5472.can-24-0890
  7. Landry, J. R., Mager, D. L. & Wilhelm, B. T. Complex controls: the role of alternative promoters in mammalian genomes. Trends Genet 19, 640-648 (2003). https://doi.org/10.1016/j.tig.2003.09.014
  8. Davuluri, R. V., Suzuki, Y., Sugano, S., Plass, C. & Huang, T. H. The functional consequences of alternative promoter use in mammalian genomes. Trends Genet 24, 167-177 (2008). https://doi.org/10.1016/j.tig.2008.01.008
  9. Demircioğlu, D. et al. A Pan-cancer Transcriptome Analysis Reveals Pervasive Regulation through Alternative Promoters. Cell 178, 1465-1477.e1417 (2019). https://doi.org/10.1016/j.cell.2019.08.018
Read the Abstract