PANDER (pancreatic derived factor, FAM3B) is a novel cytokine, present in insulin secretory granules, that induces apoptosis of alpha and beta cells of mouse, rat, and human islets in a dose- and time-dependent manner, and may be implicated in diabetes. PANDER has the predicted secondary structure of 4 alpha-helical bundles with an up-up-down-down topology, and two disulfide bonds. Eleven mutated PANDERs were constructed and expressed in beta-TC3 cells to identify the essential region of PANDER involved in beta-cell death. Beta-cell function was assessed by assays of cell viability and insulin secretion. Based on quantitative real-time RT-PCR all mutant PANDERs had similar mRNA expression levels in beta-TC3 cells. Immunoblotting showed that ten of eleven mutant PANDER proteins were synthesized and detected in beta-TC3 cells. A mutant PANDER with no signal peptide, however, was not expressed. Truncation of helix D alone caused a 40-50% decrease in PANDER's activity, while truncation of both helices C and D resulted in a 75% loss of activity. In contrast, truncation of the N-terminus of PANDER (helix A, the loop between helices A and B, and the first two cysteines) had no effect on PANDER-induced beta-cell death. The third and fourth cysteines of PANDER, C91 and C229, were shown to form one disulfide bond and be functionally important. Finally, the region between Cys91 and Phe152 constitutes the active part of PANDER, based on the demonstration that mutants with truncation of helix B or C caused decreased beta-cell death and did not inhibit insulin secretion, as compared to wild-type PANDER. Hence, helices B and C and the second disulfide bond of PANDER are essential for PANDER-induced beta-cell death.