Type 1 diabetes (T1D) is typically diagnosed when patients lose their ability to produce sufficient amounts of insulin. However, this is merely the final stage in a long process which involves the gradual loss of beta cell mass due to an autoimmune attack which specifically targets insulin producing beta cells in the pancreatic islets. Recently, the FDA approved Teplizumab–a drug which can delay the onset of insulin dependence by two years if administered before the typical point of diagnosis. Administration of Teplizumab therefore requires prediction of imminent T1D, which is based on the appearance of autoantibodies against four different islet–associated proteins (antigens) in a person’s blood. However, the predictive value of these autoantibodies is limited, as they provide a very wide time window for disease onset, which can span a decade or more. The problem is even worse in adult onset T1D, where many patients do not develop autoantibodies against the known antigens. Here, we intend to improve early T1D diagnosis by expanding the repertoire of known autoantibodies. To discover novel antigens targeted by autoantibodies in early T1D, we will use spatial transcriptomics on cadaveric pancreata of patients with recent onset T1D, and sequence the antigen–recognition site in antibody producing molecules (B cell receptors–BCR) generated near their islets. We will then create synthetic antibodies with identical sequences, and use them to “fish” their cognate antigen. Ultimately, we will use blood samples from live donors to explore whether autoantibodies against the newly discovered antigens are unique to T1D patients, and examine their potential for accurate and reliable early diagnosis.
