Fig. 1. Induction of insoluble PrP aggregates and reduction of PrP^Sc by Suramin. (A) Induction of insoluble PrP in N2a and ScN2a cells and reduction of PrP^Sc determined using a solubility assay. N2a (N; lanes 1, 3, 5 and 7) and ScN2a (S; lanes 2, 4, 6 and 8) cells were treated with Suramin for 7 days (200 µg/ml; lanes 3, 4, 7 and 8) or mock treated (lanes 1, 2, 5 and 6). Cells were lysed and postnuclear lysates were subjected to a solubility assay. PrP was analyzed by immunoblotting using the anti-PrP antibody SAF 70. Detergent-insoluble fractions are shown on the left (pellet; lanes 1–4), soluble fractions on the right (sup; lanes 5–8). Molecular size markers are depicted on the left (kDa); bars on the right indicate the six PrP-specific bands (full-length PrP and N-terminally truncated PrP^Sc). (B) Reduction of PrP^Sc and induction of insoluble PrP in ScGT1 cells. Panel (a) shows an immunoblot analysis. ScGT1 cells were treated with Suramin for 1 or 7 days (200 µg/ml; lanes 4–6 and 7–9, respectively) or mock treated (lanes 1–3). Cells were lysed, postnuclear lysates were divided into two parts, and either subjected to digestion with PK (lanes 2, 5 and 8: 20 µg/ml; lanes 3, 6 and 9: 40 µg/ml for 30 min) or proceeded further without PK treatment (lanes 1, 4 and 7). PrP was analyzed by immunoblotting using SAF 70. Panel (b) shows a solubility assay (lanes 10–12). ScGT1 cells were either mock treated (lane 10) or treated for 3 or 5 days with Suramin (200 µg/ml; lanes 11 and 12, respectively). Cells were lysed and postnuclear lysates subjected to a solubility assay. Detergent-insoluble fractions are shown. Bars on the right indicate PrP-specific bands.

Fig. 2. Detergent-insoluble PrP induced by Suramin is not PK resistant, nor is it N-terminally truncated, and it has a half-life time comparable to PrP^c. (A) PrP aggregates formed in the presence of Suramin are completely sensitive to proteolytic digestion. 3F4-ScN2a cells incubated with Suramin (200 µg/ml) for 7 days (+Sur; lanes 1–4) were compared with mock-treated cells (lanes 5–8). Cells were lysed and postnuclear lysates subjected to proteolytic digestion with various amounts of PK as indicated. PK digestion was terminated by the addition of proteinase inhibitors, lysates were adjusted to 1% sarcosyl and centrifuged for 1 h at 100 000 g. An immunoblot analysis of the remaining PrP present in the detergent-insoluble fraction is shown using the anti-PrP antibody 3F4. Molecular size markers are shown on the left (kDa); bars on the right indicate PrP-specific bands. (B) Suramin induces aggregation of full-length glycosylated PrP. 3F4-ScN2a cells were mock treated or incubated in the presence of 200 µg/ml Suramin for 7 days. Detergent-insoluble fractions were prepared as described in (A), and incubated with PNGase F or left untreated. PrP was analyzed by immunoblotting using 3F4 (3F4; lanes 1–4) or the anti-PrP antibody 4F2 (anti-N; lanes 5–8), directed to the N-terminus of PrP. Bars on the left indicate the six PrP-specific bands [glycoforms of full-length PrP (aa 23–231) and N-terminally truncated PrP^Sc (aa ∼90–231)]. (C) Detergent-insoluble PrP does not accumulate in Suramin-treated cells. 3F4-N2a cells were mock treated [panel (b); –Sur/sup] or treated for 1 day with Suramin [panel (a); +Sur/pel] and metabolically labeled for 2 h with [³⁵S]Met (+/– Suramin). One plate was harvested directly (chase: 0); the other plates were incubated further in [³⁵S]Met-free medium for 2 and 6 h, respectively. Cells were harvested and detergent-soluble (panel b) and -insoluble fractions (panel a) prepared as described. PrP was immunoprecipitated using 3F4. Lane 1 is the pellet fraction of non-Suramin-treated cells. Molecular size markers are shown on the left (kDa).

Fig. 3. Suramin induces intracellular re-routing of PrP from post-ER compartments to acidic vesicles. (A) PrP is not detectable by surface FACS analysis in Suramin-treated cells. The left panel shows the surface FACS analysis of 3F4-N2a cells [the bold line is PrP reactivity without Suramin; the broken line is in the presence of Suramin (200 µg/ml for 24 h)]. The y-axis denotes the number of cells, the x-axis the fluorescence intensity (lg-scale). The polyclonal rabbit anti-PrP antibody A4 was used. The right panel shows the identical analysis with permeabilized cells (intracellular FACS). (B) No release of surface PrP by PIPLC in cells treated with Suramin. 3F4-N2a cells were incubated for 3 days with Suramin (200 µg/ml medium), treated with PIPLC, and the cellular lysate preparations analyzed in immunoblot (3F4). In non-Suramin-treated cells, PIPLC treatment results in the expected reduction of PrP in the cellular lysate fraction (lane 2); in the presence of Suramin, PIPLC treatment has no effect (lane 4). (C) Suramin induces intracellular retention of PrP in the Golgi/TGN and re-routing to acidic compartments. 3F4-N2a cells were mock treated (panel a) or incubated in the presence of Suramin (100µg/ml for 24 h) [panels (b) and (c)] and analyzed in indirect immunofluorescence experiments using confocal microscopy. (Panel a) PrP is excluded from the cell surface of Suramin-treated cells. Cell surface proteins of fixed and non-permeabilized 3F4-N2a cells were unspecifically surface-labeled with biotin; at the same time PrP present at the cell surface was specifically labeled with the antibody 3F4. The left panel shows surface biotinylated proteins (biotin), the middle panel cell surface PrP (3F4), and the right panel the overlay of both signals in yellowish color (merge). The upper right cell does not express detectable levels of 3F4-PrP, which indicates the specificity of the PrP antibody used. (Panel b) Suramin induces the retention of PrP in a Golgi/TGN compartment. 3F4-N2a cells were treated with Suramin, fixed, permeabilized and incubated with a polyclonal anti-mannosidase II antiserum to localize the medial to trans-Golgi compartments (left panel, manII). In parallel, PrP was analyzed with 3F4 (middle panel). In the right panel the overlay of both signals indicates co-localization (merge). (Panel c) PrP is re-routed to acidic compartments in Suramin-treated cells. Cells and treatment as in panel (b), with the exception that leupeptin was added. To stain for acidic endosomal/lysosomal compartments, an antibody against the acidotropic amine DAMP was used. The left panel shows the DAMP staining, the middle panel is PrP (3F4), and the right panel denotes the overlay of both signals (merge).

Fig. 4. In vitro aggregation of PrP in the presence of Suramin without inducing a conformational shift in PrP from α-helix to β-sheet. (A) Suramin induces PrP aggregation in crude cellular lysates. 3F4-N2a cells were scraped off the cell culture dish and either lysed in lysis buffer (0.5% Triton X-100 and 0.5% DOC in PBS; lanes 3 and 4) or resuspended in PBS and disrupted by freeze–thawing (lanes 1 and 2). The crude lysates were incubated for 16 h at 4°C with (lanes 2 and 4) or without Suramin (lanes 1 and 3). All samples were adjusted to 0.5% Triton X-100 and DOC each, and centrifuged at 15 000 g for 20 min at 4°C. PrP present in the detergent-soluble (sup) and -insoluble phase (pellet) was analyzed by immunoblotting using the anti-PrP antiserum A4. (B) Suramin induces aggregation of purified recombinant PrP. In panels (a) and (b), studies with CD spectroscopy of recombinant PrP and re-aggregated PrP 27–30 are shown. Recombinant Syrian hamster PrP 29–231 (panel a) and PrP 90–231 (panel b) were either mock treated (bold line) or incubated with Suramin (7.5 µM shown; broken line) in sodium phosphate buffer pH 7 and directly analyzed by CD. The molar ellipticity is plotted against the wave length. In panel (b), the CD signal obtained for re-aggregated Syrian hamster PrP 27–30 is overlaid (dotted line; taken from Post et al., 1998, Figure 2). Aggregation of PrP is in good agreement with the quenching of Trp fluorescence (panel c). Recombinant PrP 90–231 was incubated with various concentrations of Suramin (0–10 µM), and the fluorescence emission was measured between 340 and 500 nm using an excitation wavelength of 280 nm.

Fig. 5. Suramin delays onset of prion disease in scrapie-infected mice. Days to onset of terminal prion disorder of mock-treated Tg20 mice (control; n = 18), compared with mice treated with 2 mg (n = 17), and mice treated with 5 mg of Suramin (n = 10), respectively. Individual data points represent the percentage of mice with symptoms of the total number of mice for that group, and may include more than one mouse.

Fig. 6. Model of Suramin effects on PrP biogenesis and propagation of PrP^Sc. The biosynthesis of PrP^c in the exocytotic pathway and its recycling and degradation in the endocytotic pathway are depicted (–Sur; PrP^c shown by shaded circles). The putative subcellular compartment of conversion of PrP^c into PrP^Sc in the early endocytotic pathway is marked (rafts or caveolae; Taraboulos et al., 1995). PrP^Sc, which is not efficiently degraded in lysosomes, is indicated by black boxes. The induction of detergent-insoluble PrP aggregates by Suramin in post-ER compartments and the direct re-routing to acidic vesicles are indicated (+Sur; red arrow; PrP depicted by red boxes). As a result of this re-routing, the plasma membrane localization and the putative compartment of prion conversion are bypassed.