Classification of prostate cancer using a protease activity nanosensor library
Menée à partir de l'analyse protéomique ou transcriptomique de tumeurs de la prostate pour identifier des protéases puis menée à l'aide de lignées de cellules cancéreuses humaines et de xénogreffes sur des modèles murins, cette étude présente une approche méthodologique consistant à développer, à l'aide d'une banque de peptides, des biocapteurs permettant de mesurer par fluorescence l'activité des protéases tumorales et de diagnostiquer avec précision un cancer invasif agressif
Prostate cancer is the most common noncutaneous cancer in men, but there is a need for better biomarkers that can identify aggressive disease. Here, we describe a bottom-up approach to design nanosensors to detect and classify prostate cancer. We used transcriptomic and proteomic analysis to identify proteolytic enzymes that are dysregulated in human prostate cancer and built a library of nanosensors to measure their activity in vivo using multiplexed urinary readouts. In mouse models, we demonstrated that these nanosensors could classify aggressive prostate cancer and outperform a serum biomarker. This library could be deployed as a screening test to identify patients with higher-risk prostate cancer at the time of screening.Improved biomarkers are needed for prostate cancer, as the current gold standards have poor predictive value. Tests for circulating prostate-specific antigen (PSA) levels are susceptible to various noncancer comorbidities in the prostate and do not provide prognostic information, whereas physical biopsies are invasive, must be performed repeatedly, and only sample a fraction of the prostate. Injectable biosensors may provide a new paradigm for prostate cancer biomarkers by querying the status of the prostate via a noninvasive readout. Proteases are an important class of enzymes that play a role in every hallmark of cancer; their activities could be leveraged as biomarkers. We identified a panel of prostate cancer proteases through transcriptomic and proteomic analysis. Using this panel, we developed a nanosensor library that measures protease activity in vitro using fluorescence and in vivo using urinary readouts. In xenograft mouse models, we applied this nanosensor library to classify aggressive prostate cancer and to select predictive substrates. Last, we coformulated a subset of nanosensors with integrin-targeting ligands to increase sensitivity. These targeted nanosensors robustly classified prostate cancer aggressiveness and outperformed PSA. This activity-based nanosensor library could be useful throughout clinical management of prostate cancer, with both diagnostic and prognostic utility.
Proceedings of the National Academy of Sciences , résumé, 2017