Fig. 4

The glycopeptide sequence contains a second aggregation sequence. a Full-size provasopressin and deletion mutants lacking the glycopeptide (∆gp) or the C-terminal half of the protein (1–75), with or without the ∆E47 point mutation or the Pro1, Pro2, or Pro3 replacement (as indicated) were constructed to localize the second aggregation sequence. The truncated proteins were provided with a His₆ tag (black bar). b Constructs were expressed in HN10 cells and analyzed by immunofluorescence as shown for selected examples. Bar: 10 μm. c The fraction of expressing cells forming ER aggregates was quantified as before (mean and individual values of three or four independent transfections (as indicated), analyzing ~200 expressing cells per transfection). For comparison, values for Pro1/2/3, 1–75, and 1–75Pro1/2/3 from Figs. 1c and 3c, respectively, are also shown. d, e Cells expressing construct Pro1, which lacks the N-terminal aggregation sequence, were analyzed by electron microscopy with immunogold labeling. Aggregates formed via the glycopeptide sequence showed different degrees of compaction (see also Additional file 2: Figure S2). Some aggregates contained dense and light regions (d), while others were almost completely dense (e). Bars: 500 nm. Enlargements of the boxed regions below reveal a fibrillar network within the aggregates