ARTICLES NATURE CELL BIOLOGY ADVANCE ONLINE PUBLICATION 1 Polypyrimidine tract-binding protein promotes insulin secretory granule biogenesis Klaus-Peter Knoch 1 , Hendrik Bergert 1,2 , Barbara Borgonovo 1 , Hans-Detlev Saeger 2 , Anke Altkrüger 1 , Paul Verkade 3 and Michele Solimena 1,3,4 Pancreatic β-cells store insulin in secretory granules that undergo exocytosis upon glucose stimulation. Sustained stimulation depletes β-cells of their granule pool, which must be quickly restored. However, the factors promoting rapid granule biogenesis are unknown. Here we show that β-cell stimulation induces the nucleocytoplasmic translocation of polypyrimidine tract-binding protein (PTB). Activated cytosolic PTB binds and stabilizes mRNAs encoding proteins of secretory granules, thus increasing their translation, whereas knockdown of PTB expression by RNA interference (RNAi) results in the depletion of secretory granules. These findings may provide insight for the understanding and treatment of diabetes, in which insulin secretion is typically impaired. Secretory granules store peptide hormones in peptide-secreting endocrine cells. Various stimuli induce secretory granules to fuse with the plasma membrane and release their content into the extracellular space. Similarly to most secretory proteins, peptide hormones are co- translationally translocated into the lumen of the rough endoplasmic reticulum, transported to the Golgi complex and then sorted into nas- cent secretory granules. Along this route peptide hormones can undergo multiple post-translational modifications, including pro- teolytic cleavage and glycosylation. Thus, the generation of secre- tory granules is a slow process, requiring in excess of 30 min. As sustained stimulation progressively depletes cells of secretory gran- ules, transcriptional and post-transcriptional mechanisms should be quickly activated to renew secretory granule stores. The existence of one or more factors that coordinate the synthesis of secretory 1 Experimental Diabetology and 2 Department of Surgery, Carl Gustav Carus Medical School, University of Technology Dresden, Dresden 01307, Germany. 3 Max Planck Institute for Molecular Cell Biology and Genetics, Dresden 01307, Germany. 4 Correspondence should be addressed to M.S. (e-mail: michele.solimena@mailbox.tu-dresden.de) Published online: 22 February 2004; DOI: 10.1038/ncb1099 Table 1 Morphometry of INS-1 cells treated with siRNA oligonucleotides n Minimum number Maximum number % Mean number Mean (µm 2 ) of SGs per of SGs per of SGs per surface area cell section cell section section ± sem ± sem Untreated cells 71 7 59 100 18.8 ± 1.1 90.0 ± 5.6 Cells treated with 140 0 47 100 6.6 ± 0.8 84.1 ± 2.7 siRNA oligos 1+ 2 (P < 0.000005) (P < 0.30) Cells treated with 85 0 2 60.7 0.3 ± 0.1 84.1 ± 3.6 siRNA oligos 1+ 2 with ≤ 2 SGs (P < 0.000005) (P < 0.33) (P < 0.000005)* (P = 1)* Cells treated with 55 7 47 39.3 16.6 ± 1.1 84.1 ± 4.1 siRNA oligos 1+ 2 with > 2 SGs (P < 0.1) (P < 0.35) Cells treated with 25 6 54 100 18.7 ± 2.5 nd scrambled siRNA oligo (P = 0.37) Cells treated with 25 6 56 100 18.1 ± 2.5 nd siRNA oligo for F-Luc (P = 0.28) The total number of cells per group (n) was from three independent experiments. As no differences between these experiments were found by analysis of variance (ANOVA), all data were pooled. Statistical analysis was performed using a t-test or, in case variances were not equal, a Welch test. Cells treated with siRNA oligonucleotides were compared for secretory granule content and size (P values in parenthesis) with cells in the untreated group. Each of the two distinct groups of cells treated with siRNA oligonucleotides 1 and 2 for PTB was also independently compared with the untreated group (P values in parenthesis) and with each other (P values with an asterisk). SG, secretory granule; nd, not determined. ©2004 Nature Publishing Group