In silico approach to predict pancreatic β-cells classically secreted proteins

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningfagfællebedømt


Pancreatic β-cells, residents of the islets of Langerhans, are the unique insulin-producers in the body. Their physiology is a topic of intensive studies aiming to understand the biology of insulin production and its role in diabetes pathology. However, investigations about these cells' subset of secreted proteins, the secretome, are surprisingly scarce and a list describing islet/β-cell secretome upon glucose-stimulation is not yet available. In silico predictions of secretomes are an interesting approach that can be employed to forecast proteins likely to be secreted. In this context, using the rationale behind classical secretion of proteins through the secretory pathway, a Python tool capable of predicting classically secreted proteins was developed. This tool was applied to different available proteomic data (human and rodent islets, isolated β-cells, β-cell secretory granules, and β-cells supernatant), filtering them in order to selectively list only classically secreted proteins. The method presented here can retrieve, organize, search and filter proteomic lists using UniProtKB as a central database. It provides analysis by overlaying different sets of information, filtering out potential contaminants and clustering the identified proteins into functional groups. A range of 70-92% of the original proteomes analyzed was reduced generating predicted secretomes. Islet and β-cell signal peptide-containing proteins, and endoplasmic reticulum-resident proteins were identified and quantified. From the predicted secretomes, exemplary conservational patterns were inferred, as well as the signaling pathways enriched within them. Such a technique proves to be an effective approach to reduce the horizon of plausible targets for drug development or biomarkers identification.

TidsskriftBioscience Reports
Udgave nummer2
Antal sider13
StatusUdgivet - 1 jan. 2020

Antal downloads er baseret på statistik fra Google Scholar og

Ingen data tilgængelig

ID: 240156671