Objective
The overall aim of the proposed study is understand the changes that take place in the pancreas at genetic, molecular, cellular and physiological levels as the organ transitions from normal to advanced type 1 or type 2 diabetes. As we cannot study a single pancreas gland during this disease process (which would typically take several years), we will compare data generated from pancreas tissue donated by deceased transplant organ donors, that will include entirely healthy donors, as well as donors with type 1 or type 2 diabetes. Importantly, we have access to comprehensive anonymised clinical data for every patient.
The specific objectives of the study are:
1) Generate a unique and comprehensive dataset that comprehensively characterises the pancreas, to unprecedented levels, in healthy donors and donors with diabetes.
2) Create a data analysis-platform to integrate this complex dataset using advanced data science techniques including machine learning.
3) Compare the data from healthy donors with those from donors with diabetes to identify the key changes that take place in the pancreas in patients with diabetes to identify new targets that could be used to diagnose diabetes earlier and to develop new treatments.
This project is designed to generate a unique resource that can be used for researchers worldwide to study diabetes. To this end:
1) We will make all of the data generated during the study freely available to other researchers, so that they can interrogate the date to address other important questions about diabetes.
2) We will make frozen tissue samples from all pancreases available to other researchers, so that the same samples can be used for other tests that could lead to additional discoveries.
Background Rationale
People are currently identified as being healthy or as having diabetes. We believe this 'binary diagnosis' (Diabetes: Yes or No) is too simplistic and leads to the diagnosis of the disease at very late stages. This is particularly so because we rely on changes in the blood (such as levels of glucose) to make the diagnosis. Biologically, it is much more likely (and arguably certain) that the progression from healthy to diabetes is gradual and usually slow. Furthermore, we expect that any abnormal changes start initially in the pancreas, long (perhaps years) before the result of the changes in the pancreas can be seen in the blood. Therefore, studying the abnormal changes that take place in the pancreas tissue itself is fundamental for a better understanding of diabetes - and consequently for earlier diagnosis and for developing better treatments. However, it is very difficult to study the changes that take place in the pancreas, because healthy people or those diagnoses with diabetes do not normally have biopsies performed of their pancreases: researchers rarely have access to tissue samples from the pancreas. We have overcome this important barrier by collecting pancreas tissue samples from deceased transplant organ donors with and without diabetes. Because we also have comprehensive clinical data about every patient (including blood sugar levels) we can correlate the changes we see in the pancreas with the clinical condition of the patients. Moreover, by comparing the changes we see in the blood, we can being to understand how changes in the blood relate to the changes in the pancreas itself. This approach therefore has the potential to allow us to much better understand the pancreas as it starts to transform from being 'normal' to being affected by diabetes and how that relates to changes in the blood. These insights could potentially allow us to diagnose diabetes earlier (based on previously unknown changes in the blood) or identify new therapeutic targets (based on changes that take place in the pancreas during the disease process).
Description of Project
Types 1 and 2 diabetes mellitus (T1DM and T2DM) are leading causes of ill-health and death worldwide. Current methods of diagnosis are based on crude blood tests that rely on the levels of sugar in the blood or how the sugar has affected haemoglobin in red blood cells. Patients are therefore classified into two groups: with or without diabetes. This diagnosis is often very late and only made when the blood tests are very abnormal. It is, however, more likely that progression from 'normal' to 'diabetes' is a slow and continuous process, and it should be possible to make an earlier diagnosis, and hopefully slow or reverse the progression, long before the current blood tests become very abnormal.
We believe that the key to earlier diagnosis, and development of novel therapies is to understand changes associated with diabetes, in blood and in the pancreas (the organ that produces insulin) in much more detail. This detailed information should include a description of the abnormalities in the genes, RNA (the message create from the genes), proteins, lipids (small fat molecules), and metabolites (small signalling molecules) in the pancreas of normal patients and patients with T1DM and T2DM.
This information is currently not available, because it is very difficult to access pancreas tissue from healthy donors or donors with diabetes. To solve this problem, we have collected a very large number of pancreas glands that were originally collected for transplantation but were subsequently not used. These organs were 'declined' for transplantation for a variety of reasons, which include surgical damage, age, being fatty, diabetes or an unrelated disease elsewhere in the donor. Our collection therefore samples the full spectrum of what we expect to be present in the population: from completely healthy to advanced diabetes.
In the proposed study, we will analyse our collected pancreas samples to a very detailed level, to include examination of the genes, RNA, proteins, lipids and metabolites. We also have all of the anonymised relevant clinical data for every patient, which includes their medical history (such time of diagnosis of T1DM or T2DM if relevant), sex, age, weight, height, smoking, alcohol use and medications. This will generate a very unique and comprehensive dataset that we will study using advanced data analysis methods (including artificial intelligence and machine learning approaches) to better understand differences between pancreases from 'healthy' and donors with diabetes. By comparing the data form the pancreas tissue with data obtained from the blood of the same patients, we can understand what 'signals' can be detected in the blood that predict changes in the pancreas.
We believe studying this data will give us new insights on how diabetes arises and progresses, and help us develop methods to diagnose diabetes earlier. It will also help us identify potential new therapies that may slow or reverse the development of diabetes. At the end of the project, we will make all of the data we generate, as well as any extra pancreas samples we have, freely available to the research community worldwide. This will allow other researchers to study our data and to make their own discoveries about the diagnosis and treatment of diabetes.
Anticipated Outcome
The anticipated outcomes of the proposed projects are:
1) Generation of a uniquely comprehensive dataset that is available to the international research community
2) Creation of a large bank of pancreas samples for use by the international research community
3) Identification of at least one new predictor (or biomarker) of diabetes in the blood that can be further investigated in future studies
4) Identification of at least one new therapeutic (or drug) target for diabetes that can be further investigated in future studies
The data and tissue resources that will be created from this study will be a valuable resource for the future studies of diabetes. It is anticipated that these resources will contribute to important discoveries long after the termination of the project.
Relevance to T1D
Much of the data generated in the proposed project will be from donors diagnosed with Type 1 diabetes, including from pancreas tissue donated by patients with Type 1 diabetes. The study will therefore directly improve our understanding of Type 1 diabetes. However, comprehensive understanding of the healthy pancreas is essential as a prelude to understanding how the pancreas is affected by Type 1 diabetes. The study of healthy donor samples is therefore critical for identifying new biomarkers or therapies for Type 1 diabetes. Finally, many patients who are diagnosed with Type 2 diabetes also become dependent on insulin injections. We therefore predict that there are features in the pancreas that are similar or different in Type 1 and Type 2 diabetes. A comparison of the pancreas during Type 1 and Type 2 diabetes will therefore generate additional insights about the development and progression of Type 1 diabetes.