A collaborative study has applied analytical tools to lethal cases of metastatic castration resistant-prostate cancer in order to develop a detailed map of network interactions among genes and proteins involved in enabling prostate cancer cells growth and survival.
The team, from both UC Santa Cruz and UCLA (both CA, USA), also developed a computational approach for analyzing patient-specific data that may assist clinicians in individualizing treatment.
The study, published recently in Cell, utilized clinical tissue samples from patients who died of metastatic prostate cancer, which underwent a range of novel analyses to characterize the cancer cells. The resulting databases produced personalized maps of the cancer signaling pathways in each patient – a strategy that could suggest potential targets for therapy .
“This is our dream for personalized cancer therapy,” commented Josh Stuart (UC Santa Cruz Genomics Institute). He went on to state how these findings could in future help clinicians to “choose drug targets based on what’s driving that patient’s cancer.”
A major component of the study was an analysis of the phosphoproteome of prostate cancer tumors and cells which provided the data on the phosphorylation changes of cellular proteins. The collated phosphoproteomic data was integrated with genomic and gene expression datasets, utilizing MSigDB hallmark gene sets, to provide a unified view of activated signaling pathways in late-stage prostate cancer.
Utilizing this data enabled the study to “get a more comprehensive view of aberrant signaling in this disease,” noted co-first author Evan Paull (Columbia University [NY, USA], previously UC Santa Cruz).
Currently the main treatment for advanced prostate cancer is androgen deprivation, targeting either androgen receptors or androgen synthesis; however, in most cases metastatic cancer becomes resistant. The study demonstrated some of the mechanisms involved in antiandrogen resistance, either mutations to the androgen receptor protein or alternative kinase signaling pathways.
The Patient Cancer Hallmark Integrated Phospho Signature tool enabled the researchers to generate the individual profiles based on tumor cell analysis, utilizing integrated datasets to build generic signaling network models, allowing users to create patient-specific network predictions and visualize results. Utilizing these methods, the researchers achieved accurate predictions of drug sensitivity using either genomics or phosphoproteomics data alone.
“For now it’s a research tool, but the hope is to have a strategy like this to use in the clinic,” concluded Stuart.