The Charge Structure of Helix 1 in the Prion Protein Regulates Conversion to Pathogenic PrP^Sc

Abstract

The prion diseases are transmissible neurodegenerative disorders linked to a pathogenic conformer (PrP^Sc) of the normal prion protein (PrP^C). Accumulation of PrP^Sc occurs via a poorly defined process in which PrP^Sc complexes with and converts endogenous PrP^C to nascent PrP^Sc. Recent experiments have focused on the highly charged first alpha helix (H1) of PrP. It has been proposed that two putative asparagine-to-arginine intrahelical salt bridges stabilize H1 in PrP^C yet form intermolecular ionic bonds with adjacent PrP molecules during conversion of PrP^C to PrP^Sc (M. P. Morrissey and E. I. Shakhnovich, Proc. Natl. Acad. Sci. USA 96:11293-11298, 1999). Subsequent work (J. O. Speare et al., J. Biol. Chem. 278:12522-12529, 2003 using a cell-free assay of PrP^Sc conversion suggested that rather than promoting conversion, the salt bridges stabilize PrP^C against it. However, the role of individual H1 charges in PrP^Sc generation has not yet been investigated. To approach this question, we systematically reversed or neutralized each charged residue in H1 and tested the effect on conversion to PrP^Sc in scrapie-infected murine neuroblastoma (ScN2a) cells. We find that replacements of charged H1 residues with like charges permit conversion, while charge reversals hinder it. Neutralization of charges in the N-terminal (amino acids 143 to 146) but not the C-terminal (amino acids 147 to 151) half of H1 permits conversion, while complete reversal of charge orientation of the putative salt bridges produces a nonconvertible PrP. Circular dichroism spectroscopy studies and confocal microscopy immunofluorescence localization studies indicated that charge substitutions did not alter the secondary structure or cell surface expression of PrP^C. These data support the necessity of specific charge orientations in H1 for a productive PrP^Sc-PrP^C complex.

Prion diseases are infectious diseases of protein conformation that include bovine spongiform encephalopathy in cows, scrapie in sheep, and Creutzfeldt-Jakob disease in humans (19). The central event of prion disease is a change in the conformation of normal cellular prion protein (PrP^C) into a pathogenic conformer (PrP^Sc) capable of propagating itself by complexing with and converting endogenous PrP^C to additional PrP^Sc (22). While circular dichroism (CD) and Fourier transform infrared spectroscopy studies indicate that the conformational shift from PrP^C to PrP^Sc is marked by a loss of α-helical structure and a gain in β-sheet content, the specific segments of PrP that undergo these conformational rearrangements and the nature of the PrP^Sc-PrP^C association have not yet been clearly defined (7, 21, 25).

Nuclear magnetic resonance (NMR) studies of PrP^C reveal an unstructured N-terminal segment extending from residues 23 to 119 and a globular C-terminal domain that contains two N-linked glycosylation sites, three α-helices (H1, residues 143 to 153; H2, residues 171 to 192; H3, residues 199 to 226), and two short antiparallel β-strands (S1, residues 128 to 130; S2, residues 160 to 162) (Fig. 1) (17, 24). A recent model, based on electron crystallography performed on prion rods isolated from the brain of prion-infected rodents, suggests that H2 and H3 remain intact and linked by a disulfide bond in PrP^Sc, while segments N terminal to H2 rearrange into a left-handed β-helix structure with H1 residues protruding from this structure as a loop adjacent to neighboring PrP molecules, thereby placing the helix in an area that might influence prion replication (11, 31). Indeed, due to its intriguing properties, H1 has recently received attention as a candidate segment mediating the conversion of PrP^C to PrP^Sc. Morrissey and Shakhnovich (20) first predicted that H1 might be involved in the seeding of prion aggregates, noting that H1 is an unusually hydrophilic α-helix (Fig. 1A) with few contacts with the remainder of the molecule, whereas H2 and H3 are hydrophobic and linked by a disulfide bond. Interestingly, most of the charges on H1 lie on the outer face of the helix, away from the globular domain (Fig. 1B). Based on the charge structure of H1, they predicted that proposed intrahelical salt bridges that are thought to stabilize the helix might also rearrange to form intermolecular ionic bonds that promote aggregation of PrP into β-sheets. Consistent with this idea, PrP lacking H1 was resistant to conversion in a cell line chronically infected with prions (29), although the significance of this study is mitigated by a subsequent report in which deletion of H1 interfered with the trafficking and complex glycosylation of PrP^C, factors known to interfere with its conversion to PrP^Sc (32).

FIG. 1.

Architecture of charged residues in H1. A) The sequence of H1 with corresponding amino acid numbers (numbering pertains to mouse PrP). The helix is composed of six charged residues. Blue indicates basic residues, and red indicates acidic residues. B) The helix and its charged residues shown in reference to the full-length protein. Most of the charged residues of H1 are facing the outer surface of PrP, and the helix itself has few contacts with the rest of the C-terminal globular domain. The structure was derived from the NMR data of Riek et al. (26) (Protein Data Bank no. 1AG2). The figures were created using Swiss-Pdb Viewer software (27) and the Persistence of Vision Raytracer.

Other evidence supports H1 as a possible site of association of PrP^Sc with PrP^C during formation of the PrP^Sc-PrP^C complex. For instance, antibodies directed against H1 cured prion infections in two separate cell lines chronically infected with PrP^Sc (9), and transgenic mice that express the binding domain from one of these antibodies were resistant to prions injected into spleen and brain (13). Speare et al. (28) specifically addressed the stability of H1 in prion conversion by substituting neutral residues for the aspartates at positions 143 and 146, which form putative salt bridges with R147 and R150, respectively. Using an in vitro assay to study conversion of PrP^C to PrP^Sc, they concluded that salt bridges in H1 stabilize PrP^C against conversion and are unlikely to participate in PrP^Sc generation. Peptide studies using CD also indicate that H1 retains a high helical propensity over a wide range of pH and salt conditions, suggesting that the stability of the helix represents a barrier to PrP^Sc conversion (35). However, the role that individual charged residues within H1 play in the conversion of PrP^C to PrP^Sc has not been addressed. To do so, we systematically mutated each charged residue of H1 and tested the relative potential of these mutated PrPs to be converted to PrP^Sc in prion-infected mouse neuroblastoma cells (ScN2a), a well-proven cell-based PrP^Sc conversion assay. We find that disruption of the charge structure of H1 does not alter the secondary structure or the trafficking of PrP^C; however, its ability to convert to PrP^Sc is significantly affected, depending on the specific site disrupted and the type of mutation present. Thus, we find that, independent of perturbation of the proposed salt bridges, the charge structure of H1 is critical for conversion to proceed. These results not only reveal the involvement of novel residues in the conversion process but also place constraints on current structural models of PrP^Sc and provide additional sites to direct pharmacological intervention against prion diseases.

MATERIALS AND METHODS

Construction of PrP mutants.

The mouse Prnp coding sequence was amplified by PCR and cloned into the pSP72 cloning vector (Promega) using standard molecular biology techniques and the following primers: forward, CCCCCCGGATCCTATATGTCTAAAAAGCGGCCAAAG; reverse, GGGGGGAAGCTTTTATCAGGATCTTCTCCCGTCGTAATAGGCCT. Mutations were constructed from this template by PCR amplification with Pfu Turbo polymerase (Stratagene) using forward and reverse primers containing appropriate codon mutations. The following forward primers and reverse complement primers (not shown) were used: D143E, 5′-CATGATCCATTTTGGCAACGAGTGGGAGGACCGCTACTACC-3′; D143K, 5′-CATGATCCATTTTGGCAACAAGTGGGAGGACCGCTACTACC-3′; D143R, 5′-CATGATCCATTTTGGCAACAGGTGGGAGGACCGCTACTACC-3′; D143A, 5′-CATGATCCATTTTGGCAACGCCTGGGAGGACCGCTACTACC-3′; E145D, 5′-CATTTTGGCAACGACTGGGACGACCGCTACTACCGTGAAAAC-3′; E145K, 5′-CATTTTGGCAACGACTGGAAGGACCGCTACTACCGTGAAAAC-3′; E145R, 5′-CATTTTGGCAACGACTGGAGGGACCGCTACTACCGTGAAAAC-3′; E145A, 5′-CATTTTGGCAACGACTGGGCGGACCGCTACTACCGTGAAAAC-3′; D146E, 5′-TTTGGCAACGACTGGGAGGAGCGCTACTACCGTGAAAACATG-3′; D146K, 5′-TTTGGCAACGACTGGGAGAAACGCTACTACCGTGAAAACATG-3′; D146R, 5′-TTTGGCAACGACTGGGAGCGCCGCTACTACCGTGAAAACATG-3′; D146A 5′-TTTGGCAACGACTGGGAGGCCCGCTACTACCGTGAAAACATG-3′; R147D, 5′-GGCAACGACTGGGAGGACGACTACTACCGTGAAAACATGTACCGC-3′; R147E 5′-TGGCAACGACTGGGAGGACGAGTACTACCGTGAAAACATG-3′; R147K, 5′-GGCAACGACTGGGAGGACAAATACTACCGTGAAAACATGTACCGC-3′; R147A, 5′-GGCAACGACTGGGAGGACGCCTACTACCGTGAAAACATGTACCGC-3′; R150E, 5′-CTGGGAGGACCGCTACTACGATGAAAACATGTACCGC-3′; R150D 5′-CTGGGAGGACCGCTACTACGATGAAAACATGTACCGC-3′; R150K, 5′-CTGGGAGGACCGCTACTACAAAGAAAACATGTACCGCTACCGC-3′; R150A, 5′-CTGGGAGGACCGCTACTACGCTGAAAACATGTACCGCTACCGC-3′; E151D, 5′-GAGGACCGCTACTACCGTGACAACATGTACCGCTACCCTAACC-3′; E151K 5′-GAGGACCGCTACTACCGTAAAAACATGTACCGCTACCCTAACC-3′; E151R, 5′-GAGGACCGCTACTACCGTCGAAACATGTACCGCTACCC-3′; E151A, 5′-GAGGACCGCTACTACCGTGCAAACATGTACCGCTACCCTAACC-3′; D143R/R147D, 5′-CATTTTGGCAACCGCTGGGAGGACGACTACTACCGTGAA-3′; D146R/R150D, 5′-AACGACTGGGAGCGCCGCTACTACGACGAAAACATGTAC-3′; D143R/D146R/R147D/R150D, 5′-CATTTTGGCAACCGCTGGGAGCGCGACTACTACGATGAAAACATGTAC-3′.

The PCR was then digested with the restriction enzyme DpnI, and the remaining nonmethylated PCR product was transformed into XL-1 Blue supercompetent cells (Stratagene). All constructs to be expressed in mammalian cells were subcloned into the mammalian expression vector pCB6⁺ under control of the cytomegalovirus promoter. All constructs were confirmed by sequencing prior to use in experiments.

ScN2a cell line conversion assay.

ScN2a cells (kindly provided by S. B. Prusiner, University of California, San Francisco) are persistently infected with the Rocky Mountain Laborary (RML) strain of prions (5). Cells were cultured in 50% Optimem-50% Dulbecco's modified Eagle medium (Gibco) supplemented with 10% fetal bovine serum (Gibco) without antibiotics. Cells were maintained at 37^o with 5% CO₂. Mouse PrP constructs containing the hamster-specific 3F4 epitope (methionine substitutions at residues 108 and 111), which allows specific detection of transfected PrP in these cells, were used for transient expression in ScN2a cells. Cells were transiently transfected using Lipofectamine 2000 (Invitrogen) according the manufacturer's instructions. Forty-eight hours after transfection, cells from 60-mm dishes were lysed in 500 μl lysis buffer (20 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1 mM EDTA, 0.5% Triton X-100, 0.5% Na-deoxycholate) and cleared by centrifugation at 1,000 × g for 1 min, and 400 μl of the cleared lysate was digested with 10 μg/ml Proteinase K (PK) for 30 min at 37°C. Reactions were stopped with 2 mM phenylmethylsulfonyl fluoride.

Deglycosylation reactions.

Aliquots of cell lysate (50 μg protein) were subjected to either endoglycosidase H or PNGase F (both from New England Biolabs) per the manufacturer's instructions. Reactions were stopped with Laemmli sample buffer, boiled, and submitted to Western blotting.

Western blots.

Samples were prepared for Western blotting from equal fractions of total cell lysates so that relative comparisons of conversion efficiency could be made. PK-digested samples were prepared from 400 μl of the original cell lysates as described above, while nondigested samples were prepared from 80 μl of lysates. Protein was precipitated from lysate aliquots by the addition of 4 volumes of ice-cold methanol, followed by incubation at −20°C for 30 min and centrifugation at 16,000 × g for 30 min at 4°C. The resultant protein pellet was air dried and resuspended in 2× Laemmli sample buffer and submitted to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), using a 14% polyacrylamide gel. Separated proteins were then transferred onto polyvinylidene difluoride membranes, washed in TBST (100 mM Tris, pH 7.4, 0.9% NaCl, 0.1% Tween 20), blocked in TBST containing 5% milk, and incubated for 3 h at room temperature with 1% milk containing monoclonal antibody (MAb) 3F4 (Dako) at a 1:3,000 dilution in TBST. Following a 1-h incubation in anti-mouse horseradish peroxidase-conjugated secondary antibody (Amersham), protein signal was detected with Kodak Biomax film using ECL plus (Amersham Biosciences), as described by the manufacturer.

Prokaryotic expression of recombinant PrP.

Wild-type and mutant PrP constructs of the mature mouse PrP sequence (residues 23 to 230) were subcloned from the pSP72 vector into the pRSET vector (Invitrogen), which contains an N-terminal His tag for expression in Escherichia coli BL21(DE3)pLysS, and purified using the ProBond Purification system (Invitrogen). Bacteria selected for the presence of the PrP-containing pRSET vector were grown overnight at 37°C in SOB broth (2% Bacto tryptone, 0.5% Bacto yeast extract, 0.05% NaCl, 0.05% KCl, 10 mM MgCl₂, 10 mM MgSO₄) containing ampicillin and chloramphenicol, and 1 ml of this was used to inoculate 50 ml SOB broth that was grown to an optical density at 600 nm of 0.5, at which time 1 mM isopropyl-β-d-thiogalactopyranoside was added to induce protein expression. Expression proceeded for 2 h at 37°C, at which time the cells were harvested by centrifugation, lysed in guanidinium lysis buffer (6 M guanidine HCl, 20 mM sodium phosphate, pH 7.8, 500 mM NaCl), sonicated to solubilize inclusion bodies containing PrP, and cleared of cellular debris by centrifugation. The cleared protein solution was purified using Probond Resin (Invitrogen). The protein solution and resin were incubated together for 30 min, after which the resin was washed with binding buffer (8 M urea, 20 mM sodium phosphate, pH 7.8, 500 mM NaCl), followed by wash buffer (8 M urea, 20 mM sodium phosphate, pH 7.8, 500 mM NaCl) and native wash buffer (50 mM sodium phosphate, pH 8.0, 500 mM NaCl, 20 mM imidazole). Protein was then eluted using 250 mM imidazole in native wash buffer. Fractions were analyzed using SDS-PAGE and Western blotting to confirm the presence of PrP, and fractions containing only PrP were pooled, dialyzed against 10 mM sodium phosphate (pH 6.8), and concentrated to 0.5 mg/ml with a centrifugal filtration device with a 10-kDa-molecular-mass cutoff. Final protein purity was estimated at greater than 95%.

Confocal immunofluorescence microscopy.

N2a or COS-7 cells were grown in chamber-well slides and transiently transfected with PrP constructs carrying the 3F4 epitope using Lipofectamine 2000 reagent (Invitrogen) per the manufacturer's instructions. To detect surface PrP after 24 h of expression, cells were washed three times in cold phosphate-buffered saline (PBS) and incubated for 1 h at 4°C in PBS with 1% bovine serum albumin (BSA) and a 1:50 dilution of MAb 3F4. Cells were washed and fixed in 4% paraformaldehyde for 30 min. After washing, cells were blocked with 5% milk, 1% BSA in PBS for 30 min and then washed and incubated with Alexa 488-conjugated anti-mouse antibody at 1:100 dilution in PBS. Cells were then washed and covered in 80% glycerol in PBS. Alexa 488 signal was visualized using an Olympus IX70 with a Fluoview confocal module.

CD spectroscopy.

CD spectra were recorded on an Aviv Instruments spectrometer at 25°C using a quartz cell with a path length of 1 mm. Purified samples of PrP prepared from bacterial expression were diluted to 5 μM in 10 mM phosphate buffer, pH 6.8. Spectra were taken over the wavelength range of 190 to 260 nm with a 10-s integration time and a bandwidth of 1 nm. Data were plotted using Igor Pro software (Wavemetrics).

RESULTS

H1 mutations alter conversion to PrP^Sc.

To directly assess the role of H1 charges in the conversion of PrP^C to PrP^Sc, we employed the ScN2a mouse neuroblastoma cell line. This cell line has been used extensively and carries an advantage over cell-free conversion assays, in that they have proven infectivity in animal bioassays (1, 5, 27). When these cells are transfected with mouse PrP carrying the hamster-specific 3F4 antibody epitope, recombinant PrP^C is robustly expressed and available as a substrate for conversion to PrP^Sc by endogenously generated mouse PrP^Sc. Newly expressed PrP^C and nascent PrP^Sc can be specifically detected with the monoclonal antibody 3F4. PrP^Sc is specifically identified by its resistance to digestion by Proteinase K (PK). Thus, to investigate the effect of H1 charges on the conversion of PrP^C to PrP^Sc, we prepared several mutant PrP constructs for transfection into ScN2a cells. Each charged residue within the helix was replaced with an oppositely charged residue, a neutral alanine, and a similarly charged residue. The latter was done to ensure that the observed effects on conversion were unrelated to a simple sequence substitution. Transfection of mutant constructs was performed concurrently with wild-type (wt) constructs in order to make relative comparisons of conversion efficiencies for each residue. ScN2a cells were lysed and assayed for recombinant 3F4-tagged PrP^C and PrP^Sc 48 h after transfection. To detect newly converted PK-resistant PrP^Sc, equal volumes of lysate were digested with PK and submitted to Western blot analysis using the monoclonal antibody (MAb) 3F4. PrP^C is normally expressed as three glycoforms, consisting of unglycosylated, monoglycosylated, and diglycosylated PrP, which range in molecular mass from ∼25 to 37 kDa. In all cases, mutant PrP was fully glycosylated and expressed to similar levels as wt PrP (Fig. 2 and 3, odd-numbered lanes), suggesting that the individual mutations do not interfere with the normal processing of PrP in ScN2a cells. In addition to mature PrP and its full-length glycoforms, metabolic fragments of PrP^Sc migrating faster than 20 kDa are often present, likely the products of the neutral protease calpain (34).

FIG. 2.

Effect of charge substitutions within the N-terminal half of H1 on the conversion of PrP^C to PrP^Sc in ScN2a cells. H1 mutants containing the 3F4 epitope were expressed in ScN2a cells for 48 h, and lysates were incubated for 30 min at 37°C with (+) or without (−) 10 μg/ml Proteinase K. Western blot membranes were probed with the MAb 3F4 to detect recombinant PrP. PK-resistant PrP is evident for wild-type PrP in each blot (lanes 1 and 2). A) Replacement of Asp (D) with Glu (E) (lanes 3 and 4), Lys (K) (lanes 5 and 6), or Ala (A) (lanes 9 and 10) does not reduce the amount of PrP^Sc present after digestion of the samples with PK. Some reduction is evident in the Arg (R) mutant (lanes 7 to 9). B) Position 145 is less permissive to alterations in charge. Substituting similarly charged residues allows efficient conversion to PrP^Sc (lanes 3 and 4), as does neutralizing the charge (lanes 9 and 10). Opposite charges strongly impair conversion to PrP^Sc (lanes 5 to 8). C) Substitutions at residue 146 produced a trend similar to that for residue 145. Neutralization with Ala (lanes 9 and 10) resulted in a PrP^Sc with a distinctive pattern of PK resistance. Molecular mass markers, in kilodaltons, are indicated on left.

FIG. 3.

Effect of charge substitutions within the C-terminal half of H1 on the conversion of PrP^C to PrP^Sc in ScN2a cells. The C-terminal half of H1 is less tolerant of alterations in charge and residue than the N-terminal half. A) The R147 mutant converts efficiently to PrP^Sc when Lys (lanes 7 and 8) is substituted but not when acidic residues Asp or Glu (lanes 3 to 6) are substituted. Ala significantly reduces conversion (lanes 9 and 10). B) For position 150, conversion efficiency is highly sensitive to charge (lanes 3 to 6) or residue (lanes 7 and 8) substitution. Neutralization with Ala also inhibits conversion (lanes 9 and 10). C) Position 151 is also restrictive to changes in charge (lanes 5 to 8) and residue (lanes 3 and 4). Like other C-terminal residues, Ala at position 151 inhibits conversion to PrP^Sc.

N-terminal residues of H1.

The N-terminal charged residues of H1 consist of Asp (D) at 143, Glu (E) at 145, and Asp at 146. The most N-terminal of these, D143, was permissive to change in both residue and charge (Fig. 2A); that is, PrP constructs containing alterations in the residue (D143E) and charge character [D143Lys(K), D143Ala(A)] retained their ability to convert to PK-resistant PrP. Only the D143Arg(R) mutant displayed a reduction in conversion relative to that of wt PrP. In contrast, position 145 was more restrictive to change (Fig. 2B). In this instance, although replacement of Glu with the similarly charged Asp (E145D) was permissive to conversion, substituting oppositely charged Lys and Arg residues was not. However, when the charge at position 145 was neutralized with Ala (A), conversion was comparable to or slightly greater than that of wt PrP. Additionally, PrP^Sc generated from the E145A mutant displayed a prominent diglycosylated PrP fraction after PK digestion (Fig. 2B, lane 10). The effect of substitutions at position 146 paralleled those at 145, in that like charges permitted conversion while opposite charges impaired it (Fig. 2C). Furthermore, as with position 145, replacement of residue 146 with alanine yielded a PK-resistant species with a protein signature that differs from that generated by wt PrP (Fig. 2C, lanes 9 and 10). While PrP^C carrying either the E145A or D146A mutation was indistinguishable from wt PrP in its glycosylation pattern and expression levels, PK-resistant PrP from these constructs invariably showed different relative proportions of glycosylated fractions. For E145A, the diglycosylated form was overrepresented, while in the E146A mutant, the monoglycosylated form of PrP was most prominent. Such differences in glycosylation have been considered a feature for prion strain identification (30) and suggest H1 charge may influence strain properties. Thus, for charged N-terminal residues of H1, charge-preserving substitutions allow conversion while charge reversals significantly inhibit it. Replacement of these residues with alanine permitted conversion, but the PK-resistant PrP generated from these charge-neutralized constructs displayed altered glycoforms suggestive of different prion strains.

C-terminal residues of H1.

The effect of charge perturbations on the C-terminal triad of residues R147, R150, and E151 differs slightly from that of the N-terminal residues. For position 147, changing the Arg (R) to the oppositely charged Asp (R147D) or Glu (R147E) abrogated conversion, while maintaining charge character with a Lys substitution (R147K) permitted conversion, as was generally observed with the N-terminal residues (Fig. 3A, lanes 1 to 8). In contrast to its permissive effect in the N-terminal half of H1, Ala substitution significantly inhibited but did not completely prevent the ability of PrP to convert to PrP^Sc (Fig. 3A, lanes 9 and 10). The Arg at position 150 was the most restrictive site within the C-terminal half of H1 (Fig. 3B). Charge reversals at this position prevented conversion of PrP mutants, as seen with other positions; however, replacement with the similarly charged residue Lys (K) also prevented conversion to a significant degree, comparable to that produced by reversals in the N-terminal half of the helix (Fig. 3B, lanes 7 and 8). Here again, neutralization to Ala impaired conversion, resulting in barely detectable levels of PrP^Sc compared with that of wt PrP (Fig. 3B, lanes 9 and 10). A similar trend was evident at position 151, the most C-terminal charge of the helix (Fig. 3C). A similarly charged substitution at this site is tolerated (E151D) but does not convert with the efficiency of wt PrP, and an Ala at this position restricts conversion. Substitution with opposite charges resulted in a complete block of conversion. These results suggest that the C-terminal half of H1 is less receptive to change in both residue and charge, although charge reversal has the greatest impact. In further contrast with the N-terminal half of H1, neutralization of the charged residues impairs conversion of mutant PrP^C to PrP^Sc.

H1 mutations do not prevent cell surface expression of PrP.

PrP is a glycosylphosphatidylinositol-linked surface membrane glycoprotein that follows the secretory pathway. Following translocation to the endoplasmic reticulum lumen, it is core glycosylated at two asparagine-linked sites and then passed to the Golgi apparatus, where its sugars are trimmed and processed en route to its final destination at the external leaflet of the plasma membrane. Trafficking to the cell surface is obligatory for its conversion to PrP^Sc in the ScN2a cells, and the introduction of certain mutations may impact this normal trafficking (2, 6, 16). To ensure that the differences in relative conversion we observed with mutations of H1 were not the result of impaired trafficking to the plasma membrane, we transfected ScN2a cells with 3F4-tagged mutant and wt PrP and, after 24 h of expression, assessed the pattern of glycosylation and surface localization with MAb 3F4. Data from position 151 are presented, as substitution at this site had a pronounced effect on conversion efficiency in ScN2a cells, as shown above. The pattern of glycosylation of similarly charged and oppositely charged substitutions at this position revealed no significant differences from wt PrP, as demonstrated by Western blot analysis (Fig. 4A). PNGase F treatment, which cleaves asparagine-linked glycans, resulted in a single prominent band at ∼27 kDa, confirming a glycosylation profile similar to that of wt PrP. We also subjected the samples to endoglycosidase H, which removes high-mannose sugars from immature glycoproteins that have not trafficked beyond the mid-Golgi stack. The H1 mutants were completely resistant to endoglycosidase H. Similar results were obtained for all mutants used in this study. To confirm surface expression of the various mutants, we performed indirect immunofluorescence microscopy of nonpermeabilized cells, using MAb 3F4 to selectively label recombinant PrP. A representative experiment is presented (Fig. 4B). A single confocal slice of processed cells reveals surface staining of both mutant and wt PrP. These experiments were also performed in a cell line that does not express detectable levels of PrP (COS-7) with the same results, ensuring that trafficking of mutant PrP to the cell surface is not dependent on some interaction with endogenous PrP molecules. Results for position 151 are shown, and similar results were obtained for all mutants used in this study.

FIG. 4.

Mutations in H1 do not alter trafficking of PrP to the cell surface. N2a cells were transiently transfected with 3F4-tagged wt or mutant PrP carrying the 3F4 epitope and were assayed after 24 h of expression. A) Western blot of wild-type and mutant PrP displays similar glycosylation patterns of each (top and bottom). PNGase F treatment cleaves N-linked glycans of representative H1 mutants E151D and E151K (top), although they are resistant to cleavage by endoglycosidase H (Endo H), indicating the protein has trafficked beyond the mid-Golgi stack (bottom). B) Nonpermeabilized cells were stained with MAb 3F4 and an Alexa 488-conjugated secondary antibody. Images are single 1-μm-thick confocal slices. Rim staining indicates plasmalemma localization. No differences in staining were detected among cells expressing wt, E151D, or E151K mouse PrP. Similar results were obtained for all other constructs used. Scale bar, 20 μm.

Helix 1 charge mutations do not alter the secondary structure of PrP.

Because a change in PrP structure could theoretically alter its rate of conversion to PrP^Sc, we sought to confirm that the secondary structure of PrP is not affected by a change in the charge status of H1. Mutational analysis using Swiss-Pdb Viewer software (12) predicted that each charge position along the helix could accommodate any charged residue with energetically favorable rotamers. Thus, wild-type (wt) and mutant recombinant mouse PrPs excluding the signal sequences for endoplasmic reticulum entry and glycosylphosphatidylinositol anchor attachment (i.e., PrP23-230) were generated in a bacterial expression system, and secondary structure was analyzed by circular dichroism. Spectra of wt PrP were similar to that of others using bacterially expressed recombinant PrP (14, 28), as well as PrP purified from cultured cells (33) and brain tissue (15, 21). Minima at 208 and 222 nm, indicative of high α-helical content (Fig. 5), are consistent with the observed structure of monomeric PrP from NMR studies of full-length PrP (14). The same mutants as those used for comparison of glycosylation and surface staining were tested for their effect on structure, based on their profound effects in the ScN2a cell conversion assay. In the assay conditions used, the spectra of these mutant PrPs overlap completely with those of wt PrP, suggesting that the stability of H1 is not grossly dependent on the charges and that charge perturbations in H1 do not affect the overall structure of PrP molecules. These data, in combination with the lack of effect of charge alterations on the normal trafficking and processing of PrP (Fig. 2, 3, and 4), suggest that H1 is remarkably resistant to charge and residue perturbations, yet such alterations can potently affect conversion efficiency.

FIG. 5.

Secondary structure of recombinant wt and mutant PrPs was assessed by far-UV circular dichroism (CD) spectroscopy. Samples containing wild-type (solid line), E151D (dotted line), and E151K (dashed line) recombinant mouse PrP in 10 mM phosphate buffer at pH 6.8 were scanned from 260 nm to 190 nm at 25°C, and the resultant spectra were overlaid. Spectral minima at 222 and 208 nm are indicative of α-helical content, as previously reported for PrP in solution. Mutations in H1 do not alter the CD spectra. The protein concentration for each sample was 5 μM.

Reversal of proposed salt bridges does not restore conversion.

It has been proposed that H1 is stabilized by salt bridge interactions between its charged residues (17, 23) and that salt bridge integrity might play a role in conversion to PrP^Sc (20, 28). For instance, based on molecular dynamics simulations, Dima and Thirumalai (8) have suggested that H1 salt bridge instability could lead to a shift in the population of the intermediate between PrP^C and PrP^Sc, known as PrP*, from which conversion to PrP^Sc is thought to be most efficient. Several of the mutations described above perturb the proposed salt bridges yet are not converted to PrP^Sc, suggesting that it is the specific charge structure rather than helix stability that regulates conversion. We sought to confirm this idea and differentiate between individual charge effects on conversion and effects due to changes in the overall stability of the helix. To approach this question in living cells, PrP mutants were constructed in which the residues that make up each of the two pairs of putative salt bridges were exchanged, thereby reestablishing the salt bridge pair but with the charges reversed. A construct was prepared for each individual salt bridge as well as one that reversed both salt bridges (Fig. 6A). Analysis using Swiss-Pdb Viewer indicated that the charged residues in these double charge-reversal mutants contained rotamers that were capable of forming salt bridges. Transient transfection in ScN2a cells resulted in high expression levels and glycosylation profiles that were indistinguishable from that of wt PrP, yet PK-resistant PrP was not observed for any of these mutants even after overexposure of Western blots (Fig. 6B). As with other H1 mutants, salt bridge reversal did not prevent surface expression of PrP (data not shown). Thus, although H1 was receptive to salt bridge reversal, it was no longer competent for conversion to PrP^Sc. These results place an emphasis on the role of individual charges in the conversion process and suggest that the integrity of the salt bridges in H1 is not the critical determinant in conversion, but rather its charge structure is.

FIG. 6.

Salt bridge (SB) reversal does not restore ability to convert to PrP^Sc. A) PrP constructs were made with mutations in the residues proposed to be involved in salt bridge formation in H1 (dashed lines). Mutations restored proposed salt bridges by reversing the charge order of one or both bonds. Glycosylation and expression levels indicate normal processing and proper trafficking of mutant PrPs. However, none were capable of conversion to PK-resistant PrP^Sc. The Western blot is overexposed for emphasis.

DISCUSSION

We addressed the role of H1 charges in the process of PrP^Sc generation. Although disruptions of the charge structure did not detectably alter the conformation or general trafficking of PrP^C, the charged residues of H1 were shown to be critical for efficient conversion to PrP^Sc. While some positions were more tolerant of alterations to charge, as in the case of position 143, in most cases substitutions resulting in charge reversals prevented conversion to PrP^Sc, whereas substitutions for like charges permitted conversion. In particular, the C-terminal residues (147, 150, and 151) were most resistant to change. At these positions even replacement with a neutral residue inhibited conversion, in contrast to the N-terminal half of the helix in which neutral mutations at positions 143, 145, and 146 allowed conversion comparable to that of wt PrP. This is the first demonstration of the critical nature of the individual charges of H1 in prion propagation, and it highlights the importance of several novel residues whose identity is critical for the conversion of PrP^C to PrP^Sc.

In H1 of PrP, two pairs of charged residues, D143-R147 and D146-R150, have been proposed to form salt bridges (10, 17), as has R150-E151, which can form in place of the D146-R150 bond (17, 20). Speare et al. (28) investigated the role of the putative H1 salt bridges in the conversion of recombinant PrP to a PK-resistant state in a cell-free in vitro conversion assay. They observed that, in conversion buffers favorable to salt bridge formation, conversion of PrP^C to a PK-resistant form was enhanced when a neutral amino acid was substituted at D143 or D146 in comparison to wt PrP, concluding that salt bridges resident in wt PrP stabilize H1 against conversion to PrP^Sc. However, the bonding partners of these residues were not directly tested in their study. Using ScN2a cells as a cell-based conversion assay, we did not detect a relative change in the amount of nascent PrP^Sc derived from the D143A or D146A mutants compared to wt PrP. However, we did observe a difference when the salt bridge partners of these residues, namely, R147 and R150, were neutralized, although rather than enhancing conversion, this impaired it. Our results support a more complex role for H1 in PrP^Sc production than previously considered that implicates the specific charge orientation of the helix as the important determinant rather than a stabilization effect of the putative salt bridges. If the state of integrity of the salt bridges was the sole determinant, we would expect identical results regardless of whether D143 and D146 or their bonding partners R147 and R150 were neutralized. The finding that neutralization by alanine scanning of the N-terminal region did not impair conversion, yet neutralization of the C-terminal half did, strongly suggests a specific intermolecular orientation that is considerably more dependent on the C-terminal charges.

Based on the finding that specific charged residues of H1 are critical for PrP^C to be efficiently converted to PrP^Sc, one might interpret these data as supportive of the hypothesis of Morrissey and Shakhnovic that H1 could be involved in the fibrillization of PrP molecules by transfer of H1 intrahelical salt bridges to interhelical salt bridges between H1 residues on adjacent PrP molecules (20). However, a variety of evidence suggests that H1-H1 interactions are not a site of initial PrP-PrP associations that lead to aggregation and conversion. For instance, while many peptide segments of PrP form fibrils in solution, H1 does not (18), suggesting that it alone is not competent for fibrillization. Additionally, peptides spanning H1 retain their helical nature under a wide range of pH and salt conditions, suggesting that this region is unlikely to convert to a β-sheet conformation as one of the early steps in conversion (35). Thus, although inhibition of H1-H1 interactions is unlikely to be responsible for the loss of conversion of H1 charge mutants, it may be that charge reversals in H1 are serving to inhibit interaction between H1 and another area of an adjacent PrP molecule. In this regard, our findings are consistent with a recent structural model of PrP^Sc, based on electron crystallography data of prion rods, in which highly structured PrP^Sc was isolated from infected hamster brain (11). In this model, the N-terminal portion of PrP^C is refolded into a left-handed β-helix upon conversion to the pathogenic isoform, and H1 protrudes from the β-helix in a loop structure where it comes into contact with adjacent PrP^Sc molecules. We speculate that the charges on H1 are involved in the interaction between PrP molecules in the scrapie-associated conformation and that individual residues on the helix carry varying levels of contact importance. For instance, N-terminal H1 residues were more receptive to substitution and conversion persisted unaffected when these residues were neutralized, suggesting that they play a more passive role in the conversion process and that contacts between adjacent PrP^Sc molecules are less dependent on these residues. However, neutralization of C-terminal H1 residues significantly impaired conversion, supporting the participation of these charged residues in the formation of contacts between PrP molecules during the conversion process; i.e., when the C-terminal charges are neutralized, attractive forces are lost and conversion is reduced.

An intriguing implication from the results of charge neutralization at residues 145 and 146 is that H1 may contribute to mediating the strain properties of prions. Clinically, prion strains are characterized by differences in the incubation period and phenotype of disease (3, 4) and appear to be linked to conformational differences in PrP^Sc (26). Biochemically, different strains of prion disease can often be differentiated by the pattern of the PK-resistant fragments of PrP^Sc that remain after proteolytic digestion. These fragments may differ in the migration rate or pattern of glycosylation following separation by Western blot analysis (30). The D146A mutant clearly attains a PK-resistant protein signature different from that of wt PrP, in that the monoglycosylated band is much more prominent. The nature of this occurrence is unclear but may reflect an inability of this mutant to attain the exact conformation of endogenous wt PrP^Sc within the ScN2a cells or that it develops an altered alignment during its interaction with PrP^Sc, either of which suggests an important role for H1. These data present a unique opportunity to investigate the origins of this new strain profile and may assist in understanding the nature of strain emergence.

In summary, we find that the charge structure of H1 is a critical determinant in the generation of PrP^Sc. Our data indicate that while the charges of H1 do not grossly affect the folding or trafficking of PrP, they directly impact the propagation of PrP^Sc and possibly the emergent strain of prion. These results provide the first demonstration in living cells of the importance of specific H1 charges in prion generation. They also highlight a potential region of PrP to target for prevention of PrP^Sc generation and propagation.