Using single cell transcriptomic and blood proteomic data to understand how pancreatic cancer causes diabetes and to find ways to detect the cancer earlier

Pancreatic cancer is the deadliest of the common cancers. Five-year survival remains low at 12%. Sadly for 80% of patients, by the time their cancer is diagnosed it has spread to distant organs making them ineligible for potentially curative surgery. Methods to detect pancreatic cancer earlier are badly needed ¹.

An unexplained feature of pancreatic cancer is its ability to cause hyperglycaemia (raised blood glucose). A small proportion of patients experience no disruption to blood glucose. For others, hyperglycaemia begins three years prior to cancer diagnosis. Over 45% of patients have diabetes at the time of cancer diagnosis.

We hypothesise that pancreatic cancers that do not raise blood glucose (non-diabetogenic tumours) have different molecular programs compared to their diabetogenic (diabetes-causing) counterparts.

The Costello Laboratory, focussed on early detection of pancreatic cancer, leads a Cancer Research UK-funded programme (aimed at building resources for earlier detection of pancreatic cancer in individuals with new-onset diabetes), the United Kingdom Initiative for Early Detection of Pancreatic Cancer, UK-EDI ². As part of this work, we are researching mass spectrometry-based biomarker analysis of blood proteins of pancreatic cancer patients with/without diabetes ³. In parallel, we are currently undertaking single cell gene expression analysis to explore differences in cell types and patterns of gene expression in pancreatic tumours from patients with/without diabetes at the time of their pancreatic cancer diagnosis (Pancreatic Cancer UK-funded) ⁴.

Objectives

• To merge single cell gene expression and blood-based proteomics outputs, selecting candidate markers that best exemplify differences between pancreatic cancer that is/is not associated with diabetes.
• Undertake pathway analysis investigating novel mechanisms involved in pancreatic cancer diabetes
• Validate molecular findings in independent samples.

Novelty

Comparing gene expression in single cells from diabetogenic and non-diabetogenic pancreatic tumours will allow us to identify molecular pathways associated with pancreatic cancer-related diabetes and develop new theories about how pancreatic cancer causes diabetes and the associated molecular consequences. Comparing tissue and blood data will allow potential biomarkers for early detection to be identified.

Timeliness

New-onset diabetes is an early warning sign of pancreatic cancer, and individuals with new-onset diabetes are the largest high-risk group for this cancer. Currently we do not know how to screen this group. This studentship will synergise with the UK-EDI³ programme and our Pancreatic Cancer UK-funded research.

Experimental Approach

We welcome applications from enthusiastic and ambitious candidates. Laboratory research experience is advantageous but is not a conditional requirement for applying. You will be trained in the analysis of large transcriptomic and proteomic datasets at the University of Liverpool’s Computational Biology Facility ⁵. The integration of tissue-derived data alongside blood protein data will provide a unique and comprehensive view of molecular pathways underpinning pancreatic cancer-related diabetes.

Differences in gene expression and protein level between diabetogenic and non-diabetogenic tumours will be validated by targeted analysis in independent cancers. This will include use of standard methodologies (Western blotting, ELISA) as well as a novel technology, mass cytometry-based Proximity Ligation AssaY for Rna, (PLAYR).

How to apply

If you are interested in this project, please contact Prof Eithne Costello directly (by email to ecostell@liverpool.ac.uk) stating your background and interest and please attach your CV. Please do not make an online AY application at this stage.

The online application is something that will happen after we have interviewed and selected a candidate.