Interactions between heterologous forms of prion protein: binding, inhibition of conversion, and species barriers

Abstract

The self-induced formation of the disease-associated, protease-resistant prion protein (PrP-res) from the normal protease-sensitive isoform (PrP-sen) appears to be a key event in the pathogenesis of transmissible spongiform encephalopathies. The amino acid sequence specificity of PrP-res formation correlates with, and may account for, the species specificity in transmission of transmissible spongiform encephalopathy agents in vivo. To analyze the mechanism controlling the sequence specificity of PrP-res formation, we compared the binding of PrP-sen to PrP-res with its subsequent acquisition of protease resistance by using cell-free systems consisting of heterologous versus homologous mouse and hamster PrP isoforms. Our studies showed that heterologous PrP-sen can bind to PrP-res with little conversion to the protease-resistant state and, in doing so, can interfere with the conversion of homologous PrP-sen. The interference occurred with molar ratios of homologous to heterologous PrP-sen molecules as low as 1:1. The interference was due primarily to the inhibition of conversion, but not the binding, of the homologous PrP-sen to PrP-res. The results provide evidence that the sequence specificity of PrP-res formation in this model is determined more by the conversion to protease resistance than by the initial binding step. These findings also imply that after the initial binding, further intermolecular interactions between PrP-sen and PrP-res are required to complete the process of conversion to the protease-resistant state.

Figure 1

Formation of protease-resistant ³⁵S-PrP by ³⁵S-PrP-sen with homologous and heterologous PrP-res under Gdn⋅HCl-free conditions. (A) The species of PrP-sen and PrP-res are indicated above the figure as H (hamster), M [mouse PrP-res or mouse(3F4)-PrP-sen]. Controls lacking PrP-res are indicated by −. −PK and +PK indicate PK-untreated and PK-treated samples, respectively. The major band of the ³⁵S-PrP-sen (lacking the glycosyl-phosphatidylinosital anchor) was unglycosylated and comigrated with the 28.5-kDa marker. The protease-resistant ³⁵S-PrP bands quantified by PhosphorImager analyses are indicated by the bracket on the right side. The bands lying below the bracket are partial conversion products, the formation of which appears to be less species dependent in this and previous studies (7). Molecular mass markers are on the left (in kDa). (B) Quantitation of the formation of protease-resistant ³⁵S-PrP. The means and standard deviations of five independent experiments are indicated. Statistical significance of the differences between the means is designated by ** and *** for P < 0.01 and P < 0.001, respectively, in an unpaired t test. −, Without PrP-res.

Figure 2

Binding of ³⁵S-PrP-sen to homologous and heterologous PrP-res. Radioactivities of unbound and pelleted (bound) fractions were determined as described in the text. The percentages of the radioactivity in the pelleted fraction relative to total radioactivity (sum of unbound and pelleted fractions) are expressed as ³⁵S-PrP bound to PrP-res in the graph. Without PrP-res (−), <10% of total radioactivity was detected in the pelleted fraction. Means ±SD of four independent experiments are indicated. Marginal statistical significance (P < 0.05) of the differences between the means is designated by *.

Figure 3

Interference of protease-resistant ³⁵S-PrP formation by unlabeled heterologous PrP-sen. The species of the PrP components (³⁵S-PrP-sen, unlabeled PrP-sen, and PrP-res) in the reactions are indicated by H (hamster), M [mouse or mouse(3F4)], and − (solvent control). −PK and +PK indicate PK-untreated and PK-treated samples, respectively. The molar ratio of ³⁵S-PrP-sen to unlabeled PrP-sen was 1:16. The relative amount of ³⁵S-PrP-sen versus unlabeled PrP-sen was determined by quantitative immunoblot analysis.

Figure 4

Dose-dependent interference of protease-resistant ³⁵S-PrP formation by unlabeled (“cold”) heterologous PrP-sen. The left bars show the cell-free conversion reactions with hamster ³⁵S-PrP-sen and PrP-res, whereas the right bars show those with mouse(3F4) ³⁵S-PrP and PrP-res. The conversion reactions were performed in the presence of increasing amounts of unlabeled heterologous PrP-sen as indicated. The percent conversion obtained from the control conversion reactions (without unlabeled PrP-sen, the ratio of ³⁵S-PrP-sen to unlabeled PrP-sen ratio was 1:0) was assigned as 100% and the relative percent conversions as compared with control reactions are shown. Means and standard deviations of triplicate samples are indicated.

Figure 5

Dose-dependent binding of ³⁵S-PrP-sen to PrP-res. A total of 20, 40, and 100 kcpm of hamster (Ha) ³⁵S-PrP-sen or the mixture of hamster and mouse(3F4) (Mo) ³⁵S-PrP-sen (see below) were incubated with (solid lines) or without (dashed lines) hamster PrP-res. The binding of ³⁵S-PrP-sen to PrP-res was analyzed by sedimentation analysis as described in the text. The 100-kcpm sample consisted of 20 kcpm of hamster and 80 kcpm of mouse(3F4) ³⁵S-PrP-sen. The 40-kcpm sample consisted of 20 kcpm of hamster and 20 kcpm of mouse(3F4) ³⁵S-PrP-sen. The 20-kcpm sample consisted only of hamster ³⁵S-PrP-sen. The horizontal axis indicates the ³⁵S-PrP-sen added to the reactions (kcpm), whereas the vertical axis indicates ³⁵S-PrP detected in the pelleted fractions. The recoveries of input ³⁵S-PrP varied from 70% to 85%. Each point represents a mean ±SD of triplicates.

Figure 6

Nucleated polymerization-based models for binding, conversion (Upper), and interference (Lower) phenomena. Single-site (A) and two-binding-site (B) pathways are shown. In the single-site model (Upper, A), rapid binding of PrP-sen (open triangles) is followed by a slower wave of conformational conversion of bound PrP-sen molecules to PrP-res (squares). The inclusion of nonconvertible PrP-sen (black triangles) among convertible PrP-sen molecules prevents propagation of the conversion through the bound PrP-sen molecules (Lower, A). In the two-site models (B), binding of PrP-sen can occur either at the conversion-inducing site (shown at the end of the PrP-res polymer) or at a nonconverting site (e.g., on the sides of the polymer). Interference with conversion (Lower, B) could occur by blockade of the conversion site by nonconvertible PrP-sen without blocking binding of either type of PrP-sen to the nonconverting sites. PK designates a proteinase K digestion step wherein PrP-sen is completely digested and the N-terminal octapeptide repeat domain (residues 23-≈90, the wavy lines in the PrP-sen and PrP-res structures) are removed from PrP-res.