PrP(106-126) does not interact with membranes under physiological conditions

Abstract

Transmissible spongiform encephalopathies are neurodegenerative diseases characterized by the accumulation of an abnormal isoform of the prion protein PrP(Sc). Its fragment 106-126 has been reported to maintain most of the pathological features of PrP(Sc), and a role in neurodegeneration has been proposed based on the modulation of membrane properties and channel formation. The ability of PrP(Sc) to modulate membranes and/or form channels in membranes has not been clearly demonstrated; however, if these processes are important, peptide-membrane interactions would be a key feature in the toxicity of PrP(Sc). In this work, the interaction of PrP(106-126) with model membranes comprising typical lipid identities, as well as more specialized lipids such as phosphatidylserine and GM1 ganglioside, was examined using surface plasmon resonance and fluorescence methodologies. This comprehensive study examines different parameters relevant to characterization of peptide-membrane interactions, including membrane charge, viscosity, lipid composition, pH, and ionic strength. We report that PrP(106-126) has a low affinity for lipid membranes under physiological conditions without evidence of membrane disturbances. Membrane insertion and leakage occur only under conditions in which strong electrostatic interactions operate. These results support the hypothesis that the physiological prion protein PrP(C) mediates PrP(106-126) toxic effects in neuronal cells.

FIGURE 1

Identification of β-structures in the PrP(106-126) by CR absorbance. Absorbance spectra of 5 μM CR in the presence of PrP(106-126) 0–50 μM (A) at pH 7.4, 150 mM NaCl, and (B) at pH 5, 150 mM NaCl. Absorbance was normalized to highlight the red shift at pH 7.4 and the blue shift at pH 5.0 upon peptide addition. (Inset) Initial CR absorbance spectrum was subtracted to all the spectra obtained after peptide addition at pH 7.4 or pH 5. At pH 7.4 an increase is seen in the CR absorbance at 535 nm with peptide concentration, and at pH 5 there is an increase at 496 nm. This indicates that PrP(106-126) is interacting with CR, which suggests that PrP(106-126) has a β-structure. This effect is stronger at pH 7.4 compared to pH 5.

FIGURE 2

Aggregation of PrP(106-126) evaluated by ANS fluorescence properties. The effect of peptide concentration in 12.5 μM ANS fluorescence emission spectrum (λ_excitation = 369 nm, pH 7.4, 150 mM NaCl). Spectra were normalized to highlight the ANS blue shift upon interaction with PrP(106-126). (Inset) Dependence of integrated fluorescence intensity of ANS with peptide concentration. A significant blue shift and a concomitant increase in ANS fluorescence intensity indicate that this peptide is in an aggregated form.

FIGURE 3

CD spectra of 100 μM PrP(106-126) in the presence and absence of POPC/POPG (4:1) LUVs ([Lipid] = 2 mM). Three conditions were tested: 10 mM HEPES buffer, pH 7.4, 150 mM NaF; 20 mM acetate buffer, pH 5, 150 mM NaF; and 20 mM acetate buffer, pH 5, 10 mM NaF. No α-helix (double minima bands at 208 and 222 nm and a positive band at 192 nm) is observed for PrP(106-126) under any conditions. At pH 5 (but not pH 7.4), the CD spectra have a strong negative band at 198 nm (random coil structure), with a shift upon interaction with membranes to 204 nm. Both minima in the presence and absence of membranes remain at the same wavelength for pH 7.4 conditions (204 nm).

FIGURE 4

Influence of lipid and buffer composition on peptide affinity for membranes. (A) pH and ionic strength effect on PrP(106-126) (25 μM) interaction with POPC membranes immobilized on the surface of an L1 chip. HEPES buffer pH 7.4, 150 mM NaCl; acetate buffer pH 5, 150 mM NaCl; and acetate buffer pH 5, 10 mM NaCl were used. The peptide does not show a significant affinity for membranes under physiological conditions. At acidic pH a marked increase in the membrane-bound peptide is observed, which is significantly enhanced at low ionic strength. (B) The affinity of 25 μM PrP(106-126) for membrane surfaces on L1 chips is shown for different lipidic compositions at pH 5, 10 mM NaCl. Compositions are: POPC, POPC/POPG (4:1), POPC/Chol (2:1), and POPC/GM1 (9:1). PrP(106-126) has a lower binding to POPC/Chol and POPC/GM1 relative to POPC and the highest affinity for the anionic membrane.

FIGURE 5

Global two-state kinetic analysis of SPR data PrP(106-126) in the presence of (A) POPC or (B) POPC/POPG (4:1) membranes captured on L1 chip surfaces. Peptide samples (5, 10, 15, 20, and 25 μM) were prepared in acetate buffer (pH 5, 10 mM NaCl) and injected at 20 μL/min flow rate. Sensorgrams were corrected for bulk shift effect (shaded lines) and a two-step binding model was fitted to data with BIAevaluation version 4.1 (solid lines). Kinetic and affinity constants are presented in Table 2.

FIGURE 6

Cross-bilayer channel formation induced by PrP(106-126) in 100 μM POPC/POPG (1:1) vesicles followed by NBD quenching by Co²⁺. The ratio of NBD fluorescence emission (λ_excitation = 460 nm) in the absence of quencher (I₀) and the presence of 20 mM Co²⁺ (I), for the vesicles with Co²⁺ accessible to both layers (open columns) and for the vesicles where Co²⁺ is only accessible to the outer layer (solid columns), is presented for different peptide concentrations. These experiments were carried at pH 5 (A) 150 mM NaCl or (B) 10 mM NaCl. A comparison of the results (solid columns) with the positive control (open columns) demonstrates that pores are not formed, even at a high peptide/lipid ratio (1:2, for 50 μM PrP(106-126)), at 150 mM NaCl. At 10 mM NaCl there is increased leakage of Co²⁺ with increasing peptide concentration.

FIGURE 7

PrP(106-126) effect in the dipolar potential of POPC/POPG (1:1) vesicles followed by fluorescence difference spectra of Di-8-ANEPPS-labeled vesicles. The excitation spectrum obtained in the absence of peptide was subtracted to the spectrum obtained in the presence of 25 μM PrP(106-126); both spectra were normalized to the integrated areas to reflect only the spectral shift. The difference spectrum obtained in acetate buffer, pH 5 (10 mM NaCl), has a more pronounced shift than the other four difference spectra obtained with 150 mM NaCl (pH 7.4, pH 5) or with a pH gradient pH 7.4/pH 5 (in/out) and pH 5/pH 7.4 (in/out). PrP(106-126)'s effect on the dipolar potential is therefore independent of pH or pH gradient, but depends on ionic strength.

FIGURE 8

Theoretical analysis of PrP(106-126) partition into interfacial membrane region-based free-energy change ΔG_wif from water transfer to the lipid membrane interface (see White and Wimley (62)). Residues with values ΔG_wif < 0 have a tendency to be transferred from the water phase to the membrane.