Functional depletion of mahogunin by cytosolically exposed prion protein contributes to neurodegeneration

Abstract

The pathways leading from aberrant Prion protein (PrP) metabolism to neurodegeneration are poorly understood. Some familial PrP mutants generate increased (Ctm)PrP, a transmembrane isoform associated with disease. In other disease situations, a potentially toxic cytosolic form (termed cyPrP) might be produced. However, the mechanisms by which (Ctm)PrP or cyPrP cause selective neuronal dysfunction are unknown. Here, we show that both (Ctm)PrP and cyPrP can interact with and disrupt the function of Mahogunin (Mgrn), a cytosolic ubiquitin ligase whose loss causes spongiform neurodegeneration. Cultured cells and transgenic mice expressing either (Ctm)PrP-producing mutants or cyPrP partially phenocopy Mgrn depletion, displaying aberrant lysosomal morphology and loss of Mgrn in selected brain regions. These effects were rescued by either Mgrn overexpression, competition for PrP-binding sites, or prevention of cytosolic PrP exposure. Thus, transient or partial exposure of PrP to the cytosol leads to inappropriate Mgrn sequestration that contributes to neuronal dysfunction and disease.

Fig. 1. Mahogunin interacts with cytosolically exposed PrP

(A) Experimental design to detect a potential interaction between two proteins (red and green), one of which remains immobilized upon semi-permeabilization of the plasma membrane with digitonin. (B–D) N2a cells co-transfected with the indicated FP-tagged constructs were imaged before (‘pre-Dig’) or after digitonin semi-permeabilization (‘post-Dig’) for 10 min. Note that Mgrn-RFP is partially retained with aggregates of GFP-PrP_40–231, but not Htt-GFP, after permeabilization. (E) Mgrn-RFP (red) was transfected into cells stably expressing SA-PrP-Cer or PrP-CFP (green) and analyzed by the digitonin co-association assay as in panel B. Images before and after permeabilization are shown. Mgrn-RFP is partially retained by SA-PrP-Cer, but not PrP-CFP. (F) A detergent lysate of normal adult hamster was passed over columns of immobilized BSA or Mgrn, and the bound products (along with different amounts of input brain lysate) were analyzed by immunoblot for PrP. (G) RFP-PrP_40–231 (top panels) or RFP (bottom panels) was co-expressed in N2a cells with the GFP-Mgrn contructs indicated above each lane. The cells were fractionated into a cytosolic (soluble) fraction, Triton-X 100 wash fraction, and insoluble fraction (4-fold more loaded relative to the other fractions), and immunoblotted with anti-GFP (to detect the Mgrn constructs) and anti-RFP (to detect RFP-PrP_40–231 aggregates or RFP).

Fig. 2. Mapping the interaction domains in PrP and Mgrn

(A) Deletion constructs of FP-tagged PrP and Mgrn were assayed for interaction as in Fig. 1. (B) Cytosol from cells co-expressing Mgrn, Mgrn-ΔN, and GFP were incubated with octapeptide-conjugated beads. Aliquots of the input and bound (5-fold excess) fractions were analyzed by immunoblotting with anti-Mgrn and anti-GFP. (C) Cytosol from cells co-expressing Mgrn and Mgrn-GFP were incubated with sepharose beads (‘seph.’) or beads conjugated with the PrP octapeptide (‘pept.’). Aliquots of the input, unbound, and bound (6-fold excess) fractions were analyzed by immunoblotting with anti-Mgrn antibody.

Fig. 3. Sequestration of Mgrn by cytosolically exposed PrP phenocopies Mgrn depletion

(A) HeLa cells transfected with the indicated FP-tagged PrP constructs were analysed by indirect immunofluorescence for endogenous Mgrn. Enlarged views of the areas within the white boxes (insets) are also shown. Note that the normally puntate pattern of Mgrn expression (as in the presence of wtPrP-CFP and GFP-PrP_95–231) was disrupted in GFP-PrP_40–231 expressing cells, where Mgrn is partially sequestered around the aggregates. (B) HeLa cells transfected with Mgrn siRNAs or irrelevant siRNAs were stained with Lysotracker. Mgrn knockdown causes lysosomal enlargement and clustering. (C) HeLa cells transfected with GFP, GFP-PrP_40–231 or GFP-PrP_95–231 were stained to visualize lysosomes as in panel B. Enlarged views reveal several larger lysosomal structures in GFP-PrP_40–231 expressing cells (arrowheads). (D) Histogram plotting the percentage of total lysosomes (Y-axis) at each of the sizes indicated on the X-axis. Over 120 lysosomal structures from at least 25 cells are represented for each condition. (E) HeLa cells transfected with the indicated constructs were analyzed for lysosomal morphology as in panel D. The percent of lysosomal structures that are enlarged (defined as greater than 0.8 um) is plotted. Grey bars indicate PrP constructs that interact with Mgrn.

Fig. 4. Lysosomal morphology defect caused by PrP is mediated via Mgrn

(A) The effect of several constructs on lysosomal morphology was analyzed in N2a cells. The inset shows an anti-Mgrn immunoblot of HeLa and N2a cells that had or had not been transfected with mouse Mgrn (which serves as a positive control). Mgrn was not detectable in N2a cells. Note that human Mgrn is slightly larger than mouse Mgrn. (B) HeLa cells treated with Mgrn siRNAs (upper panels) or irrelevant siRNAs (lower panels) were transfected with GFP-PrP_40–231 (left panels) or GFP-PrP_95–231 (right panels) and stained with Lysotracker. Two fields for each condition are shown. Note that GFP-PrP_40–231 closely phenocopies Mgrn knockdown, and no additional effect is seen when these two treatments are combined. (C) HeLa cells co-transfected with GFP-PrP_40–231 and either empty vector or Mgrn were stained with Lysotracker. The enlarged lysosomal morphology caused by PrP aggregates was reverted by co-expression of Mgrn (quantified in panel D). Transfected (t) and non-transfected (nt) cells are indicated for comparison. (D) GFP-PrP_40–231 was co-transfected with either empty vector, Mgrn, or the catalytically inactive MgrnΔR, and analyzed for lysosomal morphology in HeLa cells. Note that while co-expressing Mgrn rescued the disrupted lysosomal morphology to near wild type levels, MgrnΔR did not. Inset shows comparable expression levels of Mgrn and MgrnΔR in these cells. (E) GFP-PrP_40–231 or GFP-PrP_95–231 were co-transfected with Cerulean (Cer; a variant of CFP) or Mgrn_200–250 -Cer and analysed for lysosomal morphology in HeLa cells. Note that Mgrn_200–250-Cer rescues the enlarged lysosomal phenotype, presumably by shielding Mgrn-binding sites on GFP-PrP_40–231 (see Sup. Fig. S4b)

Fig. 5. Disease-associated PrPs lead to aberrant lysosomes in a Mgrn-dependent manner

(A) HeLa cells transfected with the indicated PrP constructs and either functional or inactive (MgrnΔR) Mgrn were analyzed for lysosomal morphology. (B) The indicated PrP constructs were analyzed as in panel A, except that cells were treated with proteasome inhibitor (10 uM MG132) for 4 hours immediately prior to analysis. Note that while proteasomal inhibition marginally affects the lysosomal size for wtPrP cells, there is an increase in the % of enlarged lysosomes in Ifn-PrP cells that is reverted by Mgrn, but not MgrnΔR.

Fig. 6. Alterations in Mgrn and lysosomes in PrP mutant mice

(A) Age-matched brain sections from the indicated transgenic mice (27 months old) were immunostained with α-Mgrn serum. Three regions of the brain (see Sup. Fig. S6) are shown. Note reduced Mgrn staining in Purkinje cells of the cerebellum in HuPrP(A117V) and Ifn-PrP mice, in cells of the subiculum region near the hippocampus of Ifn-PrP mice, and in cells of the piriform cortex of HuPrP(A117V) mice. On histologic sections, Mgrn typically displays diffuse cytosolic staining with nuclear exclusion (see Sup. Fig. S5). Note that fixation, sectioning, and staining of all sections being compared were performed in parallel, and that imaging conditions were identical among samples. (B) Age-matched brain sections from the indicated transgenic mice were immunostained with anti-Cathepsin D antibody. The cerebellum is shown. Note enhanced accumulation of Cathepsin D in Purkinje cells of the cerebellum in HuPrP(A117V) and Ifn-PrP mice; this is more than the age-dependent accumulation seen in Opn-PrP mice at 27 months. By contrast to the aged mice, levels of Cathepsin D expression was comparable in 4 month old transgenic mice.

Fig. 7. Selective ^CtmPrP reduction rescues Mgrn depletion in cells and mice

(A, B) HeLa cells co-transfected with various PrP constructs and either empty vector, Mgrn, or MgrnΔR were analyzed for lysosomal morphology. The Mgrn-dependent lysosomal phenotypes seen with PrP(AV3) and HuPrP(A117V) are not seen with Prl-PrP(AV3) or Opn-HuPrP(A117V). (C, D) Brain sections from Prl-PrP(AV3) and Opn-HuPrP(A117V) transgenic mice (lines 6 and 33, respectively) at the indicated ages were immunostained for Mgrn. Normal levels of Mgrn expression in Purkinje cells was observed throughout life in both cases, in contrast to HuPrP(A117V) mice (Fig. 6A).