INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

Transmissible spongiform encephalopathies (TSE)¹ are fatal neurodegenerative diseases including sporadic and familial Creutzfeldt-Jakob disease, kuru and Gerstmann-Sträussler-Scheinker syndrome in humans, scrapie in sheep, and bovine spongiform encephalopathy in cattle (1, 2). TSE are characterized by the formation and accumulation of an abnormal proteinase K-resistant isoform (PrPres) of normal protease-sensitive host-encoded prion protein (PrPsen). PrPres is formed from PrPsen by a post-translational process involving conformational changes, a higher -sheet content, and the formation of macromolecular aggregates (3-5). PrPres is closely associated with TSE-mediated brain pathology, suggesting a crucial role of PrPres in the pathogenic process. However, in the absence of PrPsen expression in the brain tissue of recipients of scrapie-infected neurografts, no pathology outside the graft itself was observed, demonstrating that PrPres and PrPsen are both required for the pathology (6). Both cross-species transmission in vivo and efficient PrPres accumulation in vitro require a high degree of sequence homology between PrPres and PrPsen molecules. For example, in scrapie-infected neuroblastoma cells, point mutations at three positions, 108, 111, and 138, in the mouse PrP sequence are sufficient to completely block the formation of PrPres (7, 8). In a cell-free in vitro system, PrPres induces the conversion of PrPsen to PrPres (9). This interaction between PrPres and PrPsen occurs with species and strain specificities that mimic TSE species barrier effects and strain differences in vivo (10-13). Collectively, these in vivo and in vitro observations strongly suggest that precise direct interactions between PrPres and PrPsen are critical in PrPres formation and TSE pathogenesis.

In previous experiments, small synthetic peptides containing certain PrP sequences spontaneously aggregated to form fibrils with a high degree of -sheet secondary structure (14-17). Moreover, other synthetic PrP peptides were shown to interact with PrPsen molecules, forming an aggregated complex with increased protease resistance (18, 19). In the current study, peptides from a variety of locations in the hamster PrP sequence were studied for their ability to influence the generation of PrPres under cell-free conditions. A group of peptides from the central portion of the PrP molecule having high -sheet content displayed a strong inhibitory effect on the cell-free conversion assay. Based on these results, we propose the existence of sites on the central part of the PrP molecule that allow a specific binding between PrPsen and PrPres as a first step in the conversion of PrPsen to PrPres.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Peptides-- The following peptides were synthesized by the laboratory of molecular structure of NIAID, NIH, Rockville MD: hamster P106-128 (KTNMKHMAGAAAAGAVVGGLGGY), P109-141 (MKHMAGAAAAGAVVGGLGGYMLGSAMSRPMMHF), P113-141 (AGAAAAGAVVGGLGGYMLGSAMSRPMMHF), and P121-141 (VVGGLGGYMLGSAMSRPMMHF). Peptides were >95% pure, and analysis by high pressure liquid chromatography revealed only a single peak. Alzheimer's disease amyloid- protein fragment 1-40 (A-(1-40)) was purchased from Sigma. Lyophilized peptides were dissolved in deionized water at a concentration of 2 mM, distributed into 20-µl aliquots, and stored at 20 °C.

Labeling and Purification of PrPsen-- The radiolabeling and the purification of the ³⁵S-PrPsen were performed as described previously by Raymond et al. (13). Briefly, 90% confluent cells were cultured for 30 min at 37 °C in 1.5 ml of methionine/cysteine-deficient medium followed by a 90-120-min incubation with 1.4 mCi of [³⁵S]methionine/cysteine/25-cm² flask. Then the cells were washed twice with phosphate-buffered saline (20 mM sodium phosphate, pH 7.4, 130 mM NaCl) and lysed in 1.5 ml of lysis buffer containing 5 mM Tris-HCl, pH 7.4, 140 mM NaCl, 5 mM EDTA, 0.5% sodium deoxycholate, and 0.5% Triton X-100. The proteins were precipitated by the addition of 5 volumes of methanol, resuspended in detergent-lipid complexes, and immunoprecipitated by overnight incubation at 4 °C with anti-PrP 3F4 monoclonal antibody (20). Immune complexes were collected by binding to protein A-Sepharose beads. Radiolabeled PrPsen was eluted from Sepharose beads using 0.1 M acetic acid and stored at 4 °C until use. The ³⁵S-labeled hamster PrPsen used in this study was the glycophosphatidylinositol-negative form (³⁵S-HaPrPsen) described previously (9).

Purification and Analysis of PrPres-- PrPres was purified from brains of scrapie-infected Syrian golden hamsters by detergent lysis and differential centrifugation (21). Hamsters were infected 70 days before with the hamster scrapie strain 263K. The yield of PrPres was determined by Western blotting using a rabbit polyclonal PrP antiserum (R27) raised against a synthetic hamster PrP peptide (residues 89-103) (22). The purity of the preparations was estimated at 50-60% by silver-staining of SDS-PAGE gels. The hamster PrPres preparation (HaPrPres) was then diluted to 1 mg/ml in phosphate-buffered saline containing 1% sulfobetaine 3-14 and stored at 20 °C.

Cell-free Conversion Assay-- The cell-free conversion reaction was performed as described previously (9, 13). Briefly, purified PrPres was partially denatured with 2.5 M guanidine hydrochloride for 30-60 min at 37 °C. An aliquot of 200 ng of HaPrPres, typically 8 µl, was then incubated for 40 h at 37 °C with ~1 ng of immunopurified ³⁵S-PrPsen (~12,000 cpm/reaction) in a final volume of 20 µl of conversion buffer (50 mM sodium citrate, pH 6, 5 mM cetylpyridinium chloride, 0.625% N-lauryl sarcosinate) in the presence or absence of peptides. At the end of the incubation time, each reaction was split 1:10, the major fraction was digested with 100 µg/ml proteinase K (PK) in Tris-saline buffer (50 mM Tris, pH 8, 130 mM NaCl) for 1 h at 37 °C, and the minor part (PK) was reserved as an undigested control. The PK reaction was stopped by the addition of 10 µl of a mixture of 4 mg/ml thyroglobulin, 20 mM Pefabloc (Boehringer-Mannheim) to each fraction (+ and PK). Samples were then precipitated in 5 volumes of methanol and centrifuged for 20 min at 14,000 × g. The pellets were resuspended in sample buffer (65 mM Tris-HCl, pH 6.8, 5% glycerol, 5% SDS, 4 M urea, 5% -mercaptoethanol, 0.5% bromphenol blue), boiled 5 min, and analyzed by SDS-PAGE on NOVEX pre-cast gels. The percent of the conversion was calculated by the ratio between the PK-resistant ³⁵S-labeled bands of approximate molecular masses 28-30 kDa and the nondigested ³⁵S-PrPsen as quantified by phosphor autoradiographic imager analysis. At the concentration used, none of the peptides affected the PK digestion of PrPres (data not shown).

Sedimentation Experiments-- Immunopurified ³⁵S-HaPrPsen (12,000 cpm/reaction) was incubated for 24 h at 37 °C with 200 ng of 2.5 M guanidine hydrochloride-pretreated HaPrPres in the absence or presence of indicated concentrations of peptides P-(109-141), P-(113-141), P-(121-141), and A-(1-40). At the end of the incubation time, the samples were centrifuged at room temperature for 20 min at 14,000 × g. The supernatants and the pellets were separately collected and methanol-precipitated before SDS-PAGE analysis and phosphor autoradiographic imager quantification.

Fourier Transform Infrared Spectroscopy-- The dried peptides were rehydrated by mixing with conversion buffer and incubated for 4-60 h at room temperature before Fourier transform infrared spectroscopy analysis. Samples (6 µl) were loaded into variable path-length cells with CaF₂ plates adjusted to a path length of 6 µm. After purging the sample chamber to reduce the water vapor contribution, spectra were collected with a Perkin-Elmer system 2000 spectrometer. Spectral parameters were as follows: 254 scans; 4-cm¹ resolution; 1 cm/sec optical path difference velocity; Kaiser-Bessel apodization. Buffer spectra were subtracted from peptide spectra to give a flat base line in the 1800-2200-cm¹ region. A water vapor spectrum was then subtracted to minimize the water vapor bands in the 1750-1850-cm¹ region.

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

Inhibition of Conversion by Hamster PrP Peptides-- Various PrP synthetic peptides, whose localization on the primary hamster PrP sequence is shown in Fig. 1A, were tested for effects on the in vitro conversion of metabolically labeled PrPsen molecules into PrPres. The peptides were used in the cell-free conversion assay at a final concentration of 0.8 mM. As shown in Fig. 1B, the non-PK-treated ³⁵S-HaPrPsen appeared as a double band with molecular masses of 30 and 28.5 kDa, corresponding to mono- or unglycosylated forms of the molecule, respectively. In the absence of HaPrPres, no PK-resistant bands were seen in the control experiment. In those conditions, none of the peptides was able by itself to promote the conversion of ³⁵S-HaPrPsen to a PK-resistant form (Fig. 1B, lanes marked ). In the presence of 200 ng of HaPrPres, two PK-resistant ³⁵S-labeled bands were obtained with molecular masses of 24 and 20 kDa (Fig. 1B, lanes marked +). However, two synthetic hamster peptides, P-(106-128) and P-(109-141), were able to inhibit the formation of ³⁵S-labeled PK-resistant PrP bands. A third peptide, P-(89-103), partially inhibited the conversion, and six other peptides, P-(23-37), P-(57-64), P-(57-72), P-(57-80), P-(139-170), and P-(218-232) failed to inhibit the conversion.

View larger version (42K): [in this window] [in a new window]   Fig. 1.   Effect of PrP peptides on cell-free conversion. A, schematic representation of hamster PrP sequence and localization of the synthetic hamster PrP peptides. The PrP sequence is shown as a horizontal box, sites for the addition of Asn-linked oligosaccharides are shown as Vs, the disulfide bond is shown as S-S, the glycophosphatidylinositol anchor attachment signal and the signal peptide are shown as the carboxyl-terminal (term) and amino-terminal black boxes, respectively. The hatched boxes depict the four octapeptide repeat regions. The methionine residues at positions 109 and 112 allowing the protein recognition by the anti-hamster PrP monoclonal antibody 3F4 are designated as M. The location of the peptides (P) used and their designation are indicated below the PrP sequence. B, 35S-HaPrPsen (12,000 cpm/reaction, ~1 ng) was incubated for 40 h at 37 °C in conversion buffer without peptide (lane marked control) or with a final concentration of 0.8 mM each of synthetic peptide in the presence (lanes marked +) or in the absence (lanes marked ) of 200 ng of HaPrPres. At the end of the incubation time, samples were analyzed by SDS-PAGE and phosphor image autoradiography as described under "Experimental Procedure." The first lane shows the 35S-HaPrPsen without the PK digestion. Molecular mass markers in kilodaltons are indicated on the left.

View larger version (39K): [in this window] [in a new window]   Fig. 2.   Dose response of the inhibition of the conversion reaction induced by hamster P-(106-128), P-(109-141), and amyloid--protein fragment 1-40 (A-(1-40)) peptides. A, SDS-PAGE phosphor image of representative conversion reactions obtained after PK digestion (PK+) in the absence () or presence of indicated concentration of peptides, P-(106-128), P-(109-141), and A-(1-40). The first lane on the left panel represents the PK-untreated sample (PK). The second lane on the same panel represents the conversion reaction performed in the absence of hamster PrPres (HaPrPres). Molecular mass markers in kilodaltons are indicated on the left. B, dose response curves of the inhibition of the hamster conversion reactions induced by hamster P-(106-128), P-(109-141), and A-(1-40) peptides. The experiments were performed as described under "Experimental Procedures." Each point represents the mean of three independent experiments ± S.D. (bars). The black symbols represent the conversion assay done with 200 ng of HaPrPres mixed with 12,000 cpm of immunopurified 35S-labeled HaPrPsen in the presence of hamster peptide sequences, and the open symbols represent the experiments done in the presence of A-(1-40) peptide. The data are plotted as the percentage of the ratio between the conversion in the presence of peptide and the control conversion reaction in the absence of peptide.

Importance of the 8-Amino Acid Sequence 113-120-- The PrP synthetic peptide AGAAAAGA from position 113 to 120 has been described as the most highly amyloidogenic peptide in the protein (14). To assess the influence of the AGAAAAGA sequence contained in both P-(106-128) and P-(109-141), we synthesized two peptides differing only in this hydrophobic sequence, P-(113-141) and P-(121-141). The P-(113-141) peptide sequence efficiently inhibited the in vitro conversion with an IC₅₀ value of 40 µM (Figs. 3, A and B). In contrast, the P-(121-141) peptide failed to inhibit the conversion at all of the concentrations used (Figs. 3, A and B). When incubated with the labeled PrPsen in the absence of PrPres, none of these peptides conferred PK resistance to PrPsen molecules (Fig. 3A). Thus, the hydrophobic sequence from 113 to 120 appeared to be essential to the inhibitory effect of peptides from this region of PrP.

View larger version (28K): [in this window] [in a new window]   Fig. 3.   Assessment of the importance of the PrP-(113-120) sequence in the inhibition of the conversion. A, SDS-PAGE phosphor image of representative conversion reactions performed in the absence or presence of increasing concentration of peptides P-(113-141) (left panel) and P-(121-141) (right panel) or with a concentration of 800 µM each of peptides in the absence of hamster PrPres (HaPrPres). The samples were PK-treated as described under "Experimental Procedures" before SDS-PAGE analysis except the first lane on the left panel, which represents one-tenth of a conversion reaction and was nondigested with PK. B, hamster P-(113-141), P-(115-141), P-(117-141), P-(119-141), and P-(121-141) primary sequence peptides and dose response curves of the inhibition of the conversion reactions. The data represent the mean of three independent experiments and are plotted as the percentage of the ratio between the conversion in the presence of the peptide and the control conversion reaction. The sequences of the peptides used are indicated above the graph.

Conformational Analysis of Inhibitory Versus Noninhibitory Peptides-- Fourier transform infrared spectroscopy was used to compare the conformations of the PrP and A-(1-40) peptides to see if there was a correlation between inhibitory efficacy and conformation. For this purpose, peptides were incubated for 4-60 h in conversion buffer without any PrPsen or PrPres. Fourier transform infrared spectra of peptides incubated for 24 h at a concentration of 2 mM are shown in Fig. 4. Five of the peptides (P-(106-128), P-(109-141), P-(113-141), P-(121-141), and A-(1-40)) showed prominent absorbance maxima at ~1630 cm¹ that were indicative of a high -sheet content. For P-(109-141), this predominant absorbance at 1630 cm¹ was maintained down to concentrations of 0.2 mM. These same five peptides were at least partially insoluble as each formed a visible particulate suspension. Two other PrP peptides (P-(89-103) and P-(218-232)) maintained clear solutions and had absorbance maxima near 1667 cm¹, with no evidence of the -sheet absorbance at ~1630 cm¹. Although all the inhibitory peptides showed evidence of a high -sheet content, other peptides, such as P-(121-141) and A-(1-40), showed a similar -sheet content but gave no inhibition of conversion. Thus, the potential to form a high amount of -sheet structure might be necessary, but not sufficient, for the specific inhibitory activity of any given peptide.

View larger version (25K): [in this window] [in a new window]   Fig. 4.   Fourier transform infrared spectroscopy. The peptides were incubated in conversion buffer for 24 h at room temperature as described under "Experimental Procedures." Buffer spectra were subtracted from peptide spectra to give a flat base line in the 1800-2200 cm1 region. A water spectrum was then subtracted to minimize the water vapor bands in the 1750-1850 cm1 region.

Peptide Sedimentation Studies-- Considering the propensity of hydrophobic PrP peptides to aggregate in vitro, we investigated whether the inhibition of the conversion could be explained by peptide-induced aggregation of ³⁵S-HaPrPsen, which might prevent conversion of this precursor protein. For this purpose, conversion reaction mixtures were centrifuged in the presence or absence of various peptides, and pellet and supernatant fractions were analyzed for ³⁵S-HaPrPsen (Fig. 5). Whereas ³⁵S-HaPrPsen remained soluble in the absence of peptides, incubation with either inhibitory peptides (P-(109-141), P-(113-141)) or noninhibitory peptides (P-(121-141), A-(1-40)) caused aggregation and pelleting of ³⁵S-HaPrPsen. Thus, although peptide-induced aggregation of PrPsen occurred, this sedimentation was not sufficient for the inhibition of the conversion reaction.

View larger version (40K): [in this window] [in a new window]   Fig. 5.   Sedimentation properties of the 35S-labeled PrPsen in conversion reaction mixtures with or without peptides. SDS-PAGE phosphor image of supernatant (S) and pellet (P) fractions obtained by centrifugation at 14,000 × g for 20 min of conversion reactions performed by mixing 35S-HaPrPsen with 200 ng of hamster PrPres in the absence (control) or presence of 100 µM P-(109-141), 100 µM P-(113-141), 400 µM P-(121-141), or 400 µM A-(1-40). Molecular mass markers are indicated on the left.

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

In the present report, we used PrP synthetic peptides to map portions of the molecule that may be involved in PrPres-PrPsen interactions. Peptides from the central part of PrP sequence, P-(106-128), P-(109-141), and P-(113-141), were very efficient in inhibiting the in vitro cell-free conversion (30 µM < IC₅₀ < 230 µM). The inhibition of the cell-free conversion assay induced by such peptides suggests that specific interactions between PrPsen and PrPres may involve the central region of the molecule. Recently, Hölscher et al. (23) showed that a mutant mouse PrP lacking the sequence from 114 to 121 (AGAAAAGA) is not converted into a proteinase K-resistant isoform after expression in scrapie-infected mouse neuroblastoma cells. In our current studies, the presence of the corresponding hamster sequence from 113 to 120 correlated with the inhibition of the conversion reaction. Indeed, all the peptides having this sequence such as P-(106-128), P-(109-1410, and P-(113-141) had an inhibitory effect, whereas P-(121-141) did not exhibit this property (Figs. 3, A and B). Interestingly, residues 119 and 120 appeared to be most critical to the inhibition as P-(119-141) was nearly as effective as P-(113-141) (Fig. 3B). Peptides from the amino-terminal region such as P-(23-37), P-(57-64), P-(57-72), and P-(57-80) and from the carboxyl-terminal region such as P-(139-170) and P-(218-232) did not inhibit the conversion reaction. Only P-(89-103) was able to slightly decrease the conversion when incubated at the high concentration of 0.8 mM (Fig. 1B). P-(89-103) is contained in the PK-resistant core of PrPres, and its location is adjacent to the most inhibitory peptide sequences. These data suggested that, unlike the central portion of PrPres, the amino- and the carboxyl-terminal regions of the molecule are less involved in the intermolecular interactions resulting in PrPres formation. This explanation is consistent with the fact that the carboxyl- and amino-terminal portions of PrPres are more sensitive to the PK digestion than the central part (21, 24).

In our experiments, the presence of PrP amino acid residues from 113 to 120 were required for maximal inhibition, but other amino acid residues are likely also to be involved in the specific intermolecular interaction. The adjacent sequence from residues 129 to 141 seemed to play an important modulatory role in the inhibition of conversion, since P-(106-128) was significantly less inhibitory compared with P-(109-141) (Figs. 2, A and B). This interpretation is consistent with a previous report showing that a point mutation at position 138 could block the accumulation of PrPres in the mouse neuroblastoma cell system (8). Together these results suggest that the region of residue 138 in the PrP polypeptide is important in the intermolecular interactions leading to PrPres formation.

The mechanism of action of the inhibitory peptides is still unclear. Possibly the inhibitory peptides mimic the structure of PrPres and competitively bind to PrPsen molecules to block the conversion process. Alternatively, peptides could bind to PrPres itself, blocking further interaction with PrPsen. In the conversion reaction buffer, the peptides P-(106-128), P-(109-141), and P-(113-141) possessed a high -sheet content reminiscent to brain-derived PrPres (3-5). Whether the active state of the inhibitory peptides is the soluble or aggregated form is not known. However, their mechanism of action cannot be explained merely by their tendency to form -sheet structure and sedimentable aggregates with PrPsen, since noninhibitory peptides such as P-(121-141) and A-(1-40) displayed the same biochemical properties.

In our hands, none of the tested peptides were capable of conferring PK resistance to ³⁵S-PrPsen (Fig. 1B). Using slightly different conditions, other authors have shown that incubation of hamster PrPsen with hamster peptide P-(90-145) resulted in the formation of a PK-resistant, sedimentable PrPsen·peptide complex (18, 19). Upon PK digestion, no shift of the molecular mass of this PrPsen·peptide complex was reported. Thus, this complex may be a large aggregate unaccessible to PK, in contrast to brain-derived PrPres and PrPres formed in our cell-free conversion experiments, which loses ~67 amino-terminal amino acid residues (6-7 kDa) with PK treatment.

Studies using PrP may suggest new approaches for possible therapeutic intervention. The ability of PrP peptides to inhibit the accumulation of PrPres in vivo remains to be established. Nevertheless, PrP peptides dispensed by direct injection or delivered by gene therapy might provide specific therapeutic treatment for TSE diseases. Furthermore, structural analysis of the minimal peptide sequence required for inhibition of conversion might lead to development of nonpeptide compounds with therapeutic potential.