Solid-state NMR studies of the prion protein H1 fragment

Protein Science (1996), 5: 1655- 1661. Cambridge University Press. Show me full text with graphs and everything JONATHAN HELLER, 1 ANDREW C. KOLBERT, 2, 3, 6 RUSSELL LARSEN, 2, 4, 7 MATTHIAS ERNST, 2 TATIANA BEKKER, 3 MICHAEL BALDWIN, 3 STANLEY B. PRUSINER, 3 ALEXANDER PINES, 2, 4 and DAVID E. WEMMER 2, 5 Abstract Conformational changes in the prion protein (PrP) seem to be responsible for prion diseases. We have used conformation-dependent chemical-shift measurements and rotational-resonance distance measurements to analyze the conformation of solid-state peptides lacking long-range order, corresponding to a region of PrP designated H1. This region is predicted to undergo a transformation of secondary structure in generating the infectious form of the protein. Solid-state NMR spectra of specifically 13C-enriched samples of H1, residues 109-122 (MKHMAGAAAAGAVV) of Syrian hamster PrP, have been acquired under cross-polarization and magic-angle spinning conditions. Samples lyophilized from 50% acetonitrile/50% water show chemical shifts characteristic of a beta-sheet conformation in the region corresponding to residues 112-121, whereas samples lyophilized from hexafluoroisopropanol display shifts indicative of alpha-helical secondary structure in the region corresponding to residues 113-117. Complete conversion to the helical conformation was not observed and conversion from alpha-helix back to beta-sheet, as inferred from the solid-state NMR spectra, occurred when samples were exposed to water. Rotational-resonance experiments were performed on seven doubly 13C-labeled H1 samples dried from water. Measured distances suggest that the peptide is in an extended, possibly beta-strand, conformation. These results are consistent with the experimental observation that PrP can exist in different conformational states and with structural predictions based on biological data and theoretical modeling that suggest that H1 may play a key role in the conformational transition involved in the development of prion diseases. Introduction Prion diseases arise through a posttranslational change to the prion protein, PrP (Borchelt et al., 1990; Prusiner, 1992). These neurodegenerative illnesses are novel in that they can be transmitted by prions, proteinaceous agents apparently devoid of nucleic acids (Meyer et al., 1991; Kellings et al., 1992), and because they are manifest in sporadic, inherited, and infectious illnesses. The function of the normal cellular protein (PrPC), which is expressed ubiquitously on the surface of neurons, is unknown. Gene-targeted mice in which the PrP gene has been disrupted appear to develop and live normally (B¸eler et al., 1992), except that, unlike normal mice, they are resistant to infection by prions (B¸eler et al., 1993; Prusiner et al., 1993). In contrast, transgenic mice expressing a mutant PrP transgene encoding a proline to leucine substitution known to cause Gerstmann-Str”ussler-Scheinker disease in humans spontaneously develop a transmissible prion disease (Hsiao et al., 1990, 1994). PrPC is a glycosylphosphatidylinositol anchored protein found mostly at the cell surface, from which it can be released by phosphatidylinositol-phospholipase C (Stahl et al., 1987), unlike the pathogenic isoform (PrPSc), which accumulates within the cell (Taraboulos et al., 1990). Many lines of evidence have converged to suggest that the conversion of PrPC into PrPSc does not involve a covalent change (Stahl et al., 1993), but is conformational in nature (Pan et al., 1993; Cohen et al., 1994). This posttranslational event gives rise to different physical properties (Borchelt et al., 1990). PrPC is soluble in nonionic detergents and is protease sensitive. PrPSc is insoluble and proteolysis cleaves only the N-terminal third of the sequence, leaving a protease-resistant core termed PrP 27-30, which retains infectivity; in the presence of Sarkosyl, this rearranges into amyloid rods that stain with Congo red and show green-gold birefringence (Prusiner et al., 1983; McKinley et al., 1991). Fourier transform infrared spectroscopy (FTIR) and CD have demonstrated that PrPC is rich in alpha-helices and virtually devoid of beta-sheet (Pan et al., 1993), unlike PrPSc and PrP 27-30, which have a high beta-sheet content (Caughey et al., 1991; Pan et al., 1993; Safar et al., 1994). Secondary structure analysis based on sequence homology and molecular modeling predicted that PrPC contains four alpha-helices, designated H1-H4. Biological data suggest that it is the first two of these helices that convert to beta-sheet in PrPSc (Huang et al., 1994, 1996). When peptides corresponding to these four regions were synthesized, three of them were found to have very low solubility in H2O, and FTIR, CD, and electron microscopy showed that they formed beta-sheets and polymerized into fibrils (Gasset et al., 1992; Nguyen et al., 1995a). However, CD and solution NMR studies in organic solvents, such as hexafluoroisopropanol (HFIP), or detergents, such as SDS, have shown that H1 and H2, as well as peptides corresponding to longer segments of PrP containing these regions, can form alpha-helices (Zhang et al., 1995). Thus, these synthetic peptides seem to be able to model some aspects of the conformational pluralism that are exhibited by PrP. To date, both the cellular and scrapie isoforms of PrP have proven intractable to high-resolution spectroscopic or crystallographic study. PrPSc is particularly problematic because it is insoluble and forms aggregates lacking long-range order (Nguyen et al., 1995b). Solid-state NMR is one of the few techniques able to answer specific structural questions about peptides or proteins in immobile states, such as aggregated peptides and membrane proteins, through the use of chemical-shift information and specific distance measurements. We have used solid-state NMR to gain structural information about an aggregated form of the first of the predicted structural regions, H1 (residues 109-122 of the Syrian hamster PrP sequence). It has been shown that 13C chemical shifts are highly correlated with peptide secondary structure in the solid state (Kricheldorf & Muller, 1983; Saito, 1986). We have employed cross-polarization/magic-angle spinning (CPMAS) techniques (Pines et al., 1973; Schaefer & Stejskal, 1976) to determine chemical shifts of specifically 13C-labeled H1 peptides, and used this information to gain insight into the overall secondary structure of these peptides. 13C CPMAS spectra can yield isotropic chemical shifts with relatively high accuracy. These chemical shifts predominately reflect the local conformations of the peptides and are largely independent of the identity of neighboring residues. Conformations of alphaR-helix, alphaL-helix, omega-helix, 310 helix, and beta-sheet can be distinguished on the basis of chemical shift. Chemical shifts of amino acid carbons in the solid state in a beta-sheet conformation differ by as much as 8 ppm from those in an alpha-helix (Saito, 1986). Similarly strong correlations have been seen in solution (Spera & Bax, 1991; Wishart & Sykes, 1994), and reproduced in recent theoretical work (de Dios et al., 1993). By using CPMAS to determine isotropic shifts, meaningful information about secondary structure of aggregated proteins can be gained. We have also used internuclear distance-measurement techniques with doubly 13C-labeled peptides to determine specific distances. Types of secondary structure can be distinguished and structural details discovered best through the measurement of a large number of distances. Strong homonuclear dipolar couplings in solids prevent the use of solution-state proton NMR experiments, such as NOESY (Jeener et al., 1979) and TOCSY (Braunschweiler & Ernst, 1983; Davis & Bax, 1985). Thus, alternative techniques must be employed to determine distances in solids, and recently many such techniques have been designed (Raleigh et al., 1988; Gullion & Schaefer, 1989; Tycko & Dabbagh, 1990; Ishii & Terao, 1995). Rotational-resonance (R2) magnetization exchange (Raleigh et al., 1988; Levitt et al., 1990) is a homonuclear distance-measurement technique that has been applied to several biological systems (Creuzet et al., 1991; Smith et al., 1994b), including one amyloid system (Lansbury et al., 1995). For carbon labels, the technique can be used to measure distances of about 7 ‰, with no R2 effects indicating that the distance between a pair of 13C labels is greater than about 7 ‰. Conclusions From our distance measurements and chemical-shift data, it appears that H1 forms an extended, primarily beta-sheet-like conformation when lyophilized from AcN/H2O or dried from pure H2O. When lyophilized from HFIP, incomplete conversion to the second conformation and overlapping residues make distance measurement more difficult. However, the chemical-shift data from the HFIP form are consistent with the presence of an alpha-helix. This alpha-helical conformation appears to be only metastable, and reverts to an extended conformation when exposed to water vapor. The observed conversion is consistent with earlier FTIR data for the peptides (Gasset et al., 1992). Additionally, the present study defines the specific residues involved in secondary structure, and, interestingly, shows that the same sequence of residues is involved in the alpha-helical or beta-sheet conformation. Structural predictions for the protein from molecular modeling (Huang et al., 1994, 1996), and the hypothesis that the infectivity of PrP is the result of a conformational change in H1 and surrounding regions of the protein (Cohen et al., 1994), are also consistent with the present study. Clearly, caution is required when comparing results from this isolated peptide to the corresponding region of the entire PrP molecule. It may also be of interest to note that our results are similar to those of a solid-state NMR study of a portion of the Abeta peptide found in amyloid deposits in the brains of patients of Alzheimer's disease. Data on the Abeta peptide residues 34-42 suggested a beta-sheet or bent beta-sheet conformation around a Gly-Gly bond (Lansbury et al., 1995). The physical properties of the beta-sheet-rich PrPSc make it unlikely that high-resolution structural data will be obtained by conventional crystallographic or solution NMR approaches. Thus, there is considerable promise for a method such as solid-state NMR, capable of defining the local environment of individual atoms in a solid lacking long-range order and measuring interatomic distances. At present, the major disadvantage of this approach to defining the structure of a complex entity such as large peptide or protein is the substantial number of specifically labeled peptides required. Despite this limitation, further studies of longer PrP peptides incorporating multiple putative sites of secondary structure, such as peptide 90-145, containing both H1 and H2 (Zhang et al., 1995), are in progress. Eventually, it should be possible to extend these investigations to the entire PrPSc molecule.