Cellular prion protein mediates impairment of synaptic plasticity by amyloid-beta oligomers

Abstract

A pathological hallmark of Alzheimer's disease is an accumulation of insoluble plaque containing the amyloid-beta peptide of 40-42 amino acid residues. Prefibrillar, soluble oligomers of amyloid-beta have been recognized to be early and key intermediates in Alzheimer's-disease-related synaptic dysfunction. At nanomolar concentrations, soluble amyloid-beta oligomers block hippocampal long-term potentiation, cause dendritic spine retraction from pyramidal cells and impair rodent spatial memory. Soluble amyloid-beta oligomers have been prepared from chemical syntheses, transfected cell culture supernatants, transgenic mouse brain and human Alzheimer's disease brain. Together, these data imply a high-affinity cell-surface receptor for soluble amyloid-beta oligomers on neurons-one that is central to the pathophysiological process in Alzheimer's disease. Here we identify the cellular prion protein (PrP(C)) as an amyloid-beta-oligomer receptor by expression cloning. Amyloid-beta oligomers bind with nanomolar affinity to PrP(C), but the interaction does not require the infectious PrP(Sc) conformation. Synaptic responsiveness in hippocampal slices from young adult PrP null mice is normal, but the amyloid-beta oligomer blockade of long-term potentiation is absent. Anti-PrP antibodies prevent amyloid-beta-oligomer binding to PrP(C) and rescue synaptic plasticity in hippocampal slices from oligomeric amyloid-beta. Thus, PrP(C) is a mediator of amyloid-beta-oligomer-induced synaptic dysfunction, and PrP(C)-specific pharmaceuticals may have therapeutic potential for Alzheimer's disease.

Figure 1. Oligomeric Aβ42 binds to neurons and to cells expressing PrP^C

a, Freshly prepared, oligomeric, or fibrillary preparations of Aβ42 were examined by transmission electron microscopy with negative staining. The arrows indicate globular oligomers in the middle segment and a fibril in the lower segment. Scale bar, 25 nm. b, Oligomeric Aβ42 peptide was analyzed by size exclusion chromatography, monitoring absorbance at 220 nm (black) and light scattering (red). The void volume (V_o) and elution of bovine serum albumin (BSA) from a separate run are shown. c, Oligomeric Aβ42 peptide (200 nM total peptide) binds to 21 DIV hippocampal neurons, whereas fresh Aβ42 (200 nM) does not. Bound biotin-Aβ42 was visualized by alkaline phosphatase conjugated streptavidin. d, Dose dependence of oligomeric Aβ42 binding to hippocampal neurons.e, The binding of 40 nM oligomeric or freshly prepared Aβ42 to COS-7 expressing PrP^C. f, g, Fresh or oligomeric Aβ42 binding to PrP^C-expressing COS-7 cells as a function of Aβ42 total concentration (monomer equivalent for oligomer preparations). Data are mean ± sem, and the Scatchard analysis is presented in g. Scale bars, 100 µm for c and e.

Figure 2. Characterization of Aβ42 oligomer binding sites

a, The table summarizes the Aβ42 binding to COS-7 cells expressing the indicated proteins. K_D values are mean ± sem. APLP1 and TMEM30B were identified by examination of individual clones from a 352-member pre-existing collection of expression vectors for transmembrane proteins (Origene). *, For K_D values indicated with an asterisk, binding of Aβ42 was detected at 2 µM ligand, but binding was not saturated. Oligomer specificity is the ratio of monomer K_D to oligomer K_D. b, Oligomeric Aβ42 (30 nM) binding to hippocampal neurons after culture for the indicated days. Neuronal cell density is similar in the three panels. Scale bar, 100 µm. c, Total protein (20 µg) from hippocampal cultures or from whole brain of the indicated genotype or from HEK293T cells transfected with a PrP expression vector was analyzed by immunoblot with anti-PrP antibody (8H4), with anti-βIII tubulin antibody or with anti-GAPDH antibody. Samples for the middle panel were pretreated with endoglycosidase (PNGase F) before gel electrophoresis through a 4–20% polyacrylamide gel in tris-glycine-SDS. Mol wt standards at left. d, Hippocampal neurons from E18 mice after 21 DIV were incubated with biotinylated Aβ42-oligomer (130 nM monomer equivalent) for 1 hour at 37C and then fixed. Bound Aβ was detected with fluorescent avidin (green), and PrP^C with anti-PrP immunocytochemistry (red). Scale bar, 100 µm. e, Cultures were prepared from wild type or Prnp −/− mice and then binding of Aβ42-oligomer (130 nM monomer equivalent) was detected as in d. Scale bar, 10µm. f, Binding of 100 nM Aβ42-oligomers to puncta in hippocampal cultures as shown in e. Data from n = 6 pairs of cultures from wild type and Prnp−/− embryos. Mean ± sem; *, P < 0.05, two-tailed t test.

Figure 3. Aβ42 oligomers bind to residues 95–110 of PrP^C

a, COS-7 cells were transfected with expression plasmids directing the expression of each of the indicated PrP deletion mutants (green, octapeptide repeats; red, globular domain). Transfected cells were assessed for binding of oligomeric Aβ42, or by live cell immunocytochemistry with 6D11 and 7D9 anti-PrP antibodies. Mean ± sem from 4 experiments. b, Schematic of antibody epitopes. c, d, PrP^C-expressing COS-7 cells were analyzed for oligomeric Aβ42 binding after exposure to the various anti-PrP antibodies for one hour. 6D11 and 8G8 but not other antibodies block Aβ42-oligomer binding. Data are mean ± sem from 4 experiments. Inhibitalicion of binding by 6D11 or 8G8 is significant (*, P < 0.02, ANOVA).

Figure 4. PrP^Cis required for Aβ42 oligomer inhibition of hippocampal long-term potentiation

a, Field potentials were recorded from the CA1 region of hippocampal slices from adult wild-type mice with or without the addition of 500 nM oligomeric Aβ42 to the perfusion 20–40 minutes prior to theta burst stimulation (TBS). The top panels show traces before and after TBS. The slope of the EPSP relative to the pre-TBS level is a plotted as a function of time in the lower panel. Data are mean ± sem from separate slices. For no peptide, n = 12 slices from 9 mice and for Aβ-oligomer, n = 31 slices from 14 mice. b, CA1 potentials were recorded from slices of mice lacking PrP expression by the same method as in a. There is no significant inhibition of LTP by oligomeric Aβ42. For no peptide, n = 10 slices from 7 mice and for Aβ-oligomer, n = 35 slices from 15 mice. c, The CA1 EPSP slope in wild-type and Prnp −/− slices was recorded in the presence of oligomeric Aβ42 by an observer blind to the genotype and is replotted from panels a and b. For the values 30–60 minutes post-TBS, the EPSPs were significantly greater in the Prnp−/− slices by Repeated Measures ANOVA, P = 0.005. For WT, n = 31 slices from 14 mice and for Prnp −/−, n = 35 slices from 15 mice. d, The magnitude of LTP between 30–60 minutes is plotted as a function of genotype, the addition of 6D11 antibody, control IgG and/or Aβ42 oligomer prior to the induction of LTP. Data are mean ± sem. For 6D11 without Aβ, n = 7, and for 6D11 plus Aβ, n= 6. For IgG without Aβ, n = 8, and for IgG plus Aβ, n= 6. The indicated comparisons are significant at **P < 0.01 or *P < 0.05, ANOVA. Untreated, IgG and 6D11 slices without Aβ42 did not differ significantly.