Isolation of Tumor- and Cell Type-Specific Exosomes Using Quenchable Nanoprobes

Pancreas cancer is often diagnosed at a locally advanced stage, which involves critical structures around the pancreas, making it impossible to safely achieve complete resection. Such advanced diseases are treated with chemotherapy and radiation (chemoradiation) to shrink the tumor down to a size that can be surgically approached. Unfortunately, the response rate of pancreas cancer to chemoradiation remains as low as 25%. However, interestingly, surgeons have started to realize that some of the tumors that received chemoradiation mostly contain scar tissue with minimal cancer cells, even if they appear unchanged in imaging studies. Thus, some surgeons have now started to aggressively attempt surgical resection of advanced pancreas cancer after chemoradiation. The challenge, however, is to determine which tumor is largely replaced by scar tissue.

We envision that a system that analyzes molecules in the bloodstream released from various tumor cell types will help us determine the cellular composition of the tumor. The system will also serve as an easy and cost-effective way to achieve early tumor diagnosis, another major unmet need in pancreas cancer management. Here, we will build a foundation to develop the system by establishing a method to accurately sample tumor-derived molecules from the bloodstream. We will focus on exosomes, nanoscale extracellular vesicles (EVs) that are released from cells, which inherit genetic and protein information from the cells of origin. The challenge is to distinguish tumor-derived EVs from other EVs since any cell in the body can produce EVs.

We will use unique nanoprobes, which appear to label tumor-derived EVs when they are

delivered into pancreas cancer tissue. The nanoprobes are special in that their fluorescent signals can be instantly switched off in the body with a non-toxic chemical. The nanoprobes are selectively delivered deep into pancreas cancer tissue with a small piece of protein called the iRGD peptide. Once the delivery is achieved, the fluorescence of the nanoprobes that did not reach the tumor is switched off. The result is an extremely tumor-specific signal. Our recent serendipitous finding has been that the nanoprobes delivered into the cells within the tumor tissue are released into the bloodstream after being encapsulated in EVs, suggesting that the nanoprobes may serve as a tool to selectively label tumor-derived EVs.

We will optimize the nanosystem to effectively and selectivity label tumor-derived EVs. We will identify a condition that delivers the nanoprobes maximally into pancreas cancer and minimally into non-tumor tissue in mice by using various methods ranging from whole animal imaging to single cell analysis. We will characterize the cell populations that were targeted by the nanoprobes, as they are likely the cells that release nanoprobe-loaded EVs. We will then isolate EVs from the bloodstream of mice and identify a time point that provides nanoprobe-loaded EVs most abundantly in pancreas cancer mice, and minimally in normal mice. The EVs will be analyzed for their genetic and protein molecules to identify signatures found only in nanoprobe-labeled EVs collected from pancreas cancer mice. After studying the cells of origin of the signatures, we will finally test whether the presence of pancreas cancer and the cellular composition within the cancer can be determined in mice using the signatures.

Although this study is highly basic in nature, once we successfully determine pancreas cancer- and cell type-specific signatures that can be sampled from the blood, it will be straightforward to apply the findings to patient care. The next step beyond this study is to collect blood samples from patients with pancreas cancer before and after chemoradiation, premalignant pancreas lesions, and benign pancreas diseases such as pancreatitis, to test whether the identified signatures can be used to correctly diagnose pancreas cancer and the degree of malignancy of the tumor. As noted in the attached letter, our Pancreas Center at Columbia University is highly supportive of advancing this project into a translational study that involves patient-derived samples.

The long-term goal of this study is to increase the population that successfully undergoes surgical treatment of pancreas cancer, either by accurately identifying operable advanced tumors that have been replaced by scar tissue after chemoradiation, or by diagnosing pancreas cancer early enough that surgery can be performed without chemoradiation. Military personnel have increased risk of pancreas cancer, and many of them have limited access to advanced health care systems. The proposed system should translate into a simple and readily accessible technology, which helps increase the number of military personnel who undergo successful surgical therapy to facilitate their return to duty. Finally, the scientific information gained through this study will greatly facilitate the understanding of pancreas cancer biology and provide deeper insight into the mechanism of cancer development and the response to chemoradiation, benefiting the general population.