Toxicity of novel C-terminal prion protein fragments and peptides harbouring disease-related C-terminal mutations Maki Daniels 1 , Grazia Maria Cereghetti 2 and David R. Brown 1 1 Department of Biochemistry, Cambridge University, UK; 2 Institute of Molecular Biology and Biophysics, ETH-Hoeneggerberg, Zu ¨ rich, Switzerland Mice expressing a C-terminal fragment of the prion protein instead of wild-type prion protein die from massive neuronal degeneration within weeks of birth. The C-terminal region of PrP c (PrP121–231) expressed in these mice has an intrinsic neurotoxicity to cultured neurones. Unlike PrP Sc , which is not neurotoxic to neurones lacking PrP c expression, PrP121–231 was more neurotoxic to PrP c -deficient cells. Human mutations E200K and F198S were found to enhance toxicity of PrP121–231 to PrP-knockout neurones and E200K enhanced toxicity to wild-type neurones. The normal metabolic cleavage point of PrP c is approximately amino- acid residue 113. A fragment of PrP c corresponding to the whole C-terminus of PrP c (PrP113–231), which is eight amino acids longer than PrP121–231, lacked any toxicity. This suggests the first eight amino residues of PrP113–121 suppress toxicity of the toxic domain in PrP121– 231. Addition to cultures of a peptide (PrP112–125) correspond- ing to this region, in parallel with PrP121–231, suppressed the toxicity of PrP121–231. These results suggest that the prion protein contains two domains that are toxic on their own but which neutralize each other’s toxicity in the intact protein. Point mutations in the inherited forms of disease might have their effects by diminishing this inhibition. Keywords: prion; neurotoxicity; circular dichroism; neuro- degeneration. PrP c is a normal cell surface glycoprotein expressed by many cells including neurones and astrocytes [1–3], microglia [4], oligodendroglia [2], leukocytes [5] and muscle cells [6]. PrP c is attached to the cell membrane via a GPI (glyco- phosphoinositol) anchor [7]. Predominantly expressed at synapses [8], it has been suggested that PrP c is important for neuronal activity [9]. More recently it has been shown that PrP c binds copper via an octameric repeat region [10]. PrP c has been shown to bind significant amounts of copper in vivo and this copper binding may be necessary for its normal form [11]. Recent work has suggested two functions for the protein. PrP c influences uptake of copper into neurones [12] where it can be utilized for synaptic release [12] or incorporation into enzymes such as Cu/Zn super- oxide dismutase [13]. Other data suggest that once PrP c has bound copper, the protein can act as a superoxide dismutase or superoxide scavenger [14]. During normal metabolism of PrP c , cleavage occurs in a region around amino-acid residues 112– 114. The metabolic C-terminal fragment of this protein can be detected nor- mally in brain [15] and the N-terminal fragment, retaining the copper binding region can be purified by metal affinity chromatography from brain [16]. The rate of cleavage is regulated by protein kinase C [17] and cleavage has been suggested to be brought about by the metalloprotease disintegrins ADAM10 and ADAM17 [18]. However, the cellular fate or function of these fragments after cleavage remains unknown. Conformational change in PrP c structure results in a higher percentage of b sheet structure and increased protease resistance, suggesting that the protein can no longer be cleaved at this point. Prion diseases are fatal neurodegenerative diseases. In these diseases, PrP c is converted to a protease resistant form (PrP Sc) that cannot be cleaved at the normal cleavage site. PrP Sc represents an altered isoform that differs markedly in conformation and accumulates to high levels in nervous tissue [19]. PrP Sc is either a major part or the sole constituent of the infectious agent of prion disease [20] and is also neurotoxic when applied to cultured cells [21]. PrP Sc is evidently the cause of neurodegeneration in vivo. However, induction of neuronal loss both in vivo and in vitro requires the expression of PrP c [21,22]. Mice lacking expression of PrP c are resistant to both the toxicity of PrP Sc and its neurodegenerative effects [22,23]. Attempts to understand the mechanism of PrP Sc neuro- toxicity have focussed on a single peptide known as PrP106–126 [24,25]. This peptide corresponds to the region of the human protein that is normally cleaved during cellular processing of PrP c but which becomes protease resistant when PrP c is converted to PrP Sc . This peptide has many features of PrP Sc including protease resistance, ability to form fibrils and high b sheet content. The mechanism of the action of this peptide has been studied in detail in culture by many groups [24,26,27,28,29,30,31] including our own [25,32,33,34,35,36,37] and has also been shown to be toxic in vivo [38]. The basic mechanism by which PrP106–126 kills neurones in cerebellar cell cultures has been shown to be the same as that by which PrP Sc acts. Both require neuronal expression of PrP c [21,25,32] and the involvement of a stress event such as superoxide production by activated microglia [21,25]. PrP106–126 kills the neurones, probably Correspondence to D. R. Brown, Department of Biology and Biochemistry, University of Bath, Bath BA2 7AY, UK. Fax: 1 44 1225 826 779, Tel.: 1 44 1225 323 133, E-mail: bssdrb@bath.ac.uk (Received 14 September 2001, accepted 1 October 2001) Abbreviations: MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl- tetrazolium bromide. Eur. J. Biochem. 268, 6155–6164 (2001) q FEBS 2001 as a result of reduction in neuronal resistance to oxidative stress [33]. Apoptosis induced by the peptide follows as a result of increased calcium influx through L-type channels and NMDA receptors [34]. Attempts to define the toxic region of the protein further have identified the region AGAAAAGA as being necessary for both fibril formation and PrP Sc -like toxicity [36]. The maximum toxicity was found for a peptide corresponding to amino-acid residues 113–126 of the human sequence containing this palindrome [36]. Mice have been generated that express truncated versions of PrP c . In particular, a mouse expressing a prion protein of 106 amino acids [39] has been used to identify those parts of the protein necessary for both infection and neurodegenera- tion in prion disease. The truncated version of PrP c can be converted into a truncated PrP Sc capable of infecting the same transgenic mice and inducing neurodegeneration. The 106 amino acids comprise the amino-acid residues 89 –140 and from 177 to the C-terminus. Although some mice expressing truncated forms of PrP c do not develop spontaneous disease [39,40] others do [41,42]. Of particular interest are mice with N-terminal deletions lacking amino-acid residues up to 121 or 134 of the mouse sequence. These mice express truncated protein at the cell surface and develop degeneration in the cerebellum shortly after birth. This disease is not prion disease but represents a novel form of cerebellum-specific degeneration. Some mice that have genetic ablations to prevent expression of all PrP c overexpress the prion protein homologue, Doppel [43]. Such mice show Purkinje cell degeneration. Doppel is homo- logous to the N-terminally truncated PrP that induces cerebellar cell loss. Co-expression of full length PrP c with either the N-terminally truncated PrP c in mice [42] or re- introduced into the Doppel expressing PrP c -knockout mice [44] prevents neurodegeneration. Therefore expression of Doppel or a C-terminal region of PrP c (PrP121–231) in the absence of the full length prion protein leads to neurodegeneration. The mechanism of this is unknown. The present investigation was carried out to determine if the C-terminal region of PrP c (PrP121–231) has intrinsic neurotoxicity and whether PrP c has an intrinsic mechanism to inhibit this. Unlike PrP Sc PrP121–231 is more neurotoxic to PrP c -deficient cells. Mutants of PrP121–231 carrying known human mutations in the C-terminal region (E200K, D178N and F198S) were also investigated. In particular, E200K and F198S were also found to enhance toxicity. However, PrP113–231 lacked any toxicity suggesting the normal N-terminus of the C-terminal metabolic cleavage product of the prion protein suppressed toxicity. MATERIALS AND METHODS Unless stated, pharmacological agents were purchased from Sigma. Animals Prion protein knockout mice (Npu-Prnp8 / 8) used in this study were those described by Manson et al. [45] or for some experiments a different strain was used (Zrk-Prnp8 / 8) and these were as described by Bu ¨ eler et al. [46]. The wild-type mice used were either 129Ola mice (as control for Npu- Prnp8 / 8) or descendants of an F1 generation mouse produced by interbreeding the original parental strains (C57BL/6 J and 129/Sv(ev) mice) used to generate Zrk-Prnp8 / 8 mice originally. Neuronal cell culture Preparation of cerebellar cells from 6-day-old mice (P6) or cortical cells from newborn mice (P0) was as previously described [25,36]. Briefly, the cerebella were dissociated in Hank’s Solution (Gibco) containing 0.5% trypsin (Sigma) and plated at 1–2 Â 10 6 cells : cm 22 in 24-well trays (Falcon) coated with poly D-lysine (50 mg : mL 21 , Sigma). Cultures were maintained in Dulbecco’s minimal essential medium (Gibco) supplemented with 10% fetal bovine serum, 2 m M glutamine and 1% antibiotics (penicillin, streptomycin, fungizone; Gibco). Cultures were maintained at 37 8C with 5% CO 2 for 10 days. The neuronal nature of these cultures was confirmed by immunostaining of parallell wells with neurofilament as previously described [3]. Peptides or proteins were added to cultures initially and on the third day. Cell survival was determined on day 5 or 7. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyl-tetrazo- lium bromide (MTT; Sigma) was diluted to 200 m M in Hanks’ solution (Gibco) and added to cultures for 1 h at 37 8C. The MTT formazan product was released from cells by addition of dimethylsulfoxide (Sigma) and measured at 570 nm in a Unicam Helios spectrophotometer (ATI Unicam). Relative survival in comparison to untreated control cultures could then be determined. For coculture experiments, cerebellar cells were plated as normal in 24-well trays. Microglia were plated at 10 000 cells per well in Falcon insets with 0.4-mm pore diameter. Peptides could then be applied directly to the lower wells and to the insert. After peptide treatments, the inserts were removed and MTT assays carried out on the cerebellar cells as before. Glia cell culture Microglia were isolated as described previously [25]. Briefly, cortices from newborn mice were dissociated with trypsin and seeded into tissue culture flasks (Falcon). When cultures were confluent, microglia were isolated by rapid shaking for 2 hours. Microglia were isolated by shaking mixed glial cultures for 2 h. The floating cells were collected and plated for 30 min. The nonadhering cells were discarded. The adhering cells were identified as pure cultures of micro- glia by immunostaining for ferritin as previously described [25]. PrP peptides and protein Mouse prion protein peptides (PrP121 – 231), both wild-type and those carrying human mutations, were generated as previously described [47]. The wild-type peptide (PrP121– 231) was mutated to carry amino-acid substitutions equivalent to the human mutations E200K, D178N and F198S as previously described [47]. The numbering refers to the human sequence but the equivalent amino-acid residue of the mouse sequence was altered (one codon proximal in each case). Purification of these mutants from inclusion bodies was as previously described [47] except that the bacteria were 6156 M. Daniels et al. (Eur. J. Biochem. 268) q FEBS 2001 grown at 26 8C to avoid major degradation of the protein and no gradient was applied to the DE52/CM52 column during purification. Full length protein (PrP23–231) and other recombinant proteins equivalent to the complete metabolic C-terminal cleavage product of PrP (PrP113–231) or dele- tion mutants, PrP23–112, PrP79 – 231 and PrP105– 231, were also purified and refolded as previously described [14]. PrP23–112, PrP113–231, PrP105–231 and PrP79–231 were generated by PCR-based mutagenesis using splint oligonucleotides to introduce a restriction site before codon 113, 105 or 79 or after codon 112. Following digestion of the pET-PrP construct with Nde I the N-terminus encoding region was removed and religation of the vector lead to a pET-PrP construct expressing the complete metabolic C-terminal fragment of PrP as determined by sequencing. The N-terminal fragment was prepared by digestion of the mutated plasmid with Xho I, removing the C-terminal encoding fragment and religating the plasmid. Fidelity of the N-terminus of these deletion mutants was determined by mass spectroscopy analysis. Shorter peptides were synthesized in house by the PNAC facility. These peptides based on the mouse prion protein sequence with corresponding amino-acid residues were PrP112–125(AGAAAAGAVVGGLG), 121–146(AVVGG LGYMLGSAMSRPIIHFGSDYED), PrP147– 171(RYYRE NMYRYPNQVYYRPVDQYSNQ), PrP163–184(RPVDQ YSNQNNFVHDCVNITIK), PrP180– 198(NITIKQHTVTT KGENFT) and PrP196–220(NFTETDVKMMERVVRQM CVTQYQKE). CD spectroscopy and analysis CD spectra were recorded for prion proteins and peptides using a CD6 spectropolarimeter (Jobin Yvon, Division d’Instrumente S.A.) or a Jasco J-810 spectropolarimeter, calibrated with ammonium (1)-camphor-10-sulfonate by a method similar to that described previously [36]. Peptide samples were diluted from high concentration dimethylsulf- oxide stocks to a concentration of 2 mg : mL 21 in 10 mM sodium phosphate (pH 7.4). These samples were measured in cuvettes of 1 mm or 0.5 mm pathlength (Hellma). The spectrum from 190 nm to 250 nm was analysed with step resolution of 0.5 nm at a temperature of 23 8C. Five scans were averaged and the background from buffer was subtracted. Spectra are presented as molar ellipticity (u). RESULTS PrP121–231 is toxic to neurones Prion proteins of different lengths were tested for toxicity by application to cultures of cerebellar neurones. Cerebellar cultures were prepared from 6-day-old mice. The proteins were applied to the cultures and the cultures were maintained in serum-free medium for 7-days. The medium was changed every 3 days and replenished with prion proteins. At the end of that time, survival was determined using an MTT assay. Full length recombinant PrP c , PrP79– 231, PrP105– 231 and PrP113 –231 showed no toxicity to wild-type cultures of cerebellar cells (Fig. 1A). However, PrP121–231 was toxic to these cells in a dose-dependent manner. The proteins were also applied to cultures of Npu- Prnp8 / 8 cerebellar cells and the same results were obtained (Fig. 1B). However, PrP121–231 was more toxic to Prnp8 / 8 cerebellar cells than to wild-type cells (Student’s t-test, P , 0.05). Additionally, PrP121–231 was more toxic to Zrk-Prnp8 / 8 cerebellar cells than wild-type cerebellar cells Fig. 1. Toxicity of prion proteins. Cerebellar cell cultures from wild- type (A) and Npu-Prnp8 / 8 (B) cerebellar cells were grown in serum-free medium and treated with prion proteins at different concentrations for 7 days. After that time the survival of the cultures was measured with an MTT assay. The values were compared to those for untreated cultures as a percentage. Shown are PrP23–231 (W) PrP79–231 (O), PrP105–231 (A), PrP113–231 (K) and PrP121–231 (X). Bovine serum albumin (B) was used as a negative control. (C) Wild-type and Zrk Prnp8 / 8 cerebellar cells were either treated with L-leucine methyl ester (grey bars) or cocultured with additional microglia (black bars). Cerebellar cells were treated with 1 mg : mL 21 PrP23–231 or PrP121–231 for 7 days. After this time, the survival was measured using an MTTassay and comparing values to those of untreated cultures (open bars). Shown are the mean ^ SEM of four experiments (different cultures) with three determinations (separate wells) each. q FEBS 2001 Novel PrP peptide toxicity (Eur. J. Biochem. 268) 6157 (Fig. 1C). This confirms that the increased toxicity to Prnp8 / 8 cerebellar cells is a result of the lack of PrP c expression and not the genetic background of the mice. These results suggest that the toxicity of PrP121–231 does not require the expression of native PrP c . Previously, it has been shown that the synthetic peptide fragment PrP106–126 is toxic and that this toxicity requires microglia. L-Leucine methyl ester selectively destroys microglia [25]. Cultures of wild-type and Zrk- Prnp8 / 8 cerebellar cells were treated with 50 m ML-leucine methyl ester for 2 h before application of PrP121 – 231 or were cocultured continuously with wild-type microglia during application of PrP121– 231 for 7 days. At the end of this time an MTT assay was carried out. Treatment with L-leucine methyl ester had only a minor effect on the toxicity of the protein. Addition of microglia had no signifi- cant effect on the toxicity of PrP121– 231. These results suggest that the toxicity of PrP121– 231 does not require involvement of microglia. In order to verify that PrP121–231 is toxic to neurones, staining with Hoechst 33342 reagent was carried out during treatment of cerebellar cell cultures. Dying neurones could be determined by the presence of condensed and/or frag- mented nuclei (Fig. 2). Treatment with PrP121–231 caused a great increase in the number of fragmented nuclei present after 4 days of treatment as compared to treatment with full length protein (PrP23–231). The quantitation of this is shown in Fig. 2C. Peptide dissection of PrP121–231 toxicity Five synthetic peptides corresponding to the majority of PrP121–231 were synthesized. These peptides were PrP121–146, PrP147–171, PrP163–184, PrP180–198 and PrP196–220. Initially we attempted to synthesize a set of four nonoverlapping peptides. However, one of these peptides self terminated during synthesis. Therefore two overlapping peptides were produced to cover the region 172–180. As for the recombinant proteins, the toxicity of these peptides was tested on cultures of wild-type and Npu-Prnp8 / 8 cerebellar cells (Fig. 3). Peptides PrP121–146, PrP147–171 and PrP180–198 showed little or no toxicity to wild-type cells whereas PrP196–220 and PrP163–184 were both toxic to wild-type cerebellar cells (Fig. 3A). However, when applied to Npu-Prnp8 / 8 cerebellar cells PrP147–171 showed toxicity. This suggests that this peptide is toxic in the absence of PrP c expression. PrP163–184 showed reduced toxicity. However, PrP196–220 also remained similarly toxic (Fig. 3B). These results suggest that the toxicity of the C-terminal domain of PrP c in particular is associated with amino-acid residues 163–184 but that amino-acid residues further N-terminal to this may also participate in the toxicity to PrP c deficient cells. Additionally residues of PrP196–220 might also participate in this toxicity. The peptides used in this analysis were studied using the CD spectroscopic technique. The analysis of the struc- tural content was determined by the curve fitting program CNNR. The CD spectra of the five peptides appears in Fig. 4. The analysis of the structural content of the peptides appears in Table 1. As can be seen in Fig. 4, PrP180–198 has a noticeably different spectrum with an unusual minimum at 230 nm. It is currently unclear what this minimum represents. PrP196–200 has a high percentage of b sheet structure. This peptide readily forms fibrils (data not shown). Mutants of PrP121 –231 Further recombinant mouse proteins based on PrP121–231 were analysed for toxicity. These new proteins contained the equivalent of one of three human point mutations associated with human prion disease. The mutations analysed were E200K, D178N and F198S. For simplicity the mutants of PrP121–231 shall be referred to by these mutation assign- ments even though in the mouse sequence the location of the residue is numerically one place closer to the N-terminus in the sequence (e.g. E200K in human but in mouse the E is at 199). PrP121–231 will be referred to as wild-type protein. Wild-type cerebellar cells were treated with the four Fig. 2. PrP121–231 causes apoptotic cell death. Cerebellar cells from wild-type mice were exposed to PrP23–231 (A) or PrP121–231 (B) for 2 days. The cells were then stained with the Hoechst reagent to detect fragment or condensed nuclei. An increase in aberrant nuclei (bright condensed, fragmented) was only detected in cultures treated with PrP121–231. Scale bar, 50 m M (C) Quantitation of aberrant nuclei in culture treated with either PrP23 –231, PrP121–231 or the control. Ten fields were counted on three coverslips for four separate experiments. Shown and mean and SEM. 6158 M. Daniels et al. (Eur. J. Biochem. 268) q FEBS 2001 different proteins for 7 days, after which time an MTT assay was used to quantitate survival. The E200K mutation was the only mutation to increase the toxicity of PrP121–231 significantly above that of the wild-type protein (P , 0.05) (Fig. 5A). Zrk-Prnp8 / 8 cerebellar cells were also treated with the proteins and assayed for survival in the same manner (Fig. 5B). Both E200k and F198S showed significantly (P , 0.05) more toxicity to Zrk-Prnp8 / 8 cerebellar cells than the wild-type PrP121 –231. The implication of these results is that the mutations in the region of amino-acid residues 198–200 enhance the toxicity of the peptide PrP121–231. Peptide inhibition of PrP121 –231 toxicity In the paper by Shmerling et al. [42] it was suggested that the in vivo toxicity of PrP121–231 was inhibited by the coexpression of full length PrP c (PrP23–231). Therefore we tested whether PrP23–231, PrP23–112, PrP105 – 231 or PrP113–231 would inhibit the toxicity of PrP121 –231 to Npu-Prnp8 / 8 cerebellar cells. Of these proteins only PrP23 – 231 inhibited toxicity completely as determined by MTT assay (Fig. 5C). However, PrP105– 231 showed a low level of inhibition. This result suggests that PrP23–231 contains a domain inhibitory to the toxicity of PrP121–231, but this domain does not lie in the unstructured N-terminus; possibly it lies in the hydrophobic domain of PrP c . A recent study on optimization of the neurotoxicity of the well characterized prion protein peptide PrP106–126 [36] indicated that optimal toxicity of this region is located in amino-acid residues 113–126 of the human sequence. This region is identical to amino-acid residues 112–125 of the Fig. 4. Circular dichroism study of peptides. Peptides were freshly prepared in 10 m M phosphate buffer pH 7.4 at 2 mg : mL 21 . The peptides shown are PrP112–125, PrP121–146, PrP147– 171, PrP163– 184, PrP180–198 and PrP196–220. Five sweeps were collected at 0.5-nm intervals. Spectra are presented as molar ellipticity (u). Fig. 3. Toxicity of peptides to cerebellar neurones. Cerebellar cell cultures from wild-type (A) and Npu-Prnp8 / 8 (B) cerebellar cells were grown in serum free medium and treated with prion protein peptides at different concentrations for 5 days. After that time the survival of the cultures was measured with an MTT assay. The values were compared to those for untreated cultures as a percentage. Shown are PrP121–146 (W) PrP147–171 (X), PrP163–184 (A), PrP180–198 (K) and PrP196– 220 (O). Shown are the mean ^ SEM of four experiments with three determinations each. Table 1. Analysis of CD of peptides. Values determined using the CNNR program. Analysis based on the spectra between 190 and 250 nm. Helix b Sheet b Turn Random coil PrP121–146 12 14 8 56 PrP147–171 21 34 20 21 PrP163–184 33 24 17 26 PrP180–198 20 32 19 40 PrP196–220 20 41 20 20 q FEBS 2001 Novel PrP peptide toxicity (Eur. J. Biochem. 268) 6159 mouse sequence. This peptide is highly toxic to wild-type cerebellar cells but is not at all toxic to Zrk-Prnp8 / 8 cerebellar cells [37]. The ability of this peptide to inhibit the toxicity of PrP121–231 was analysed. Increasing amounts of PrP112– 125 were mixed with fixed amounts of E200K, D178N, F198S and wild-type PrP121–231 and applied to Npu-Prnp8 / 8 cerebellar cells for 5 days. After this time, an MTT assay was used to measure survival (Fig. 5D). PrP112–125 suppressed the toxicity of wild- type PrP in a dose-dependent manner reaching saturation above a 1 : 1 molar ratio. PrP112–125 was not able to completely suppress the toxicity of the mutant proteins. There was no further effect of toxicity above a 1: 1 molar ratio. The relative order of effective suppression was wild-type . F198S E200K . D178N. Indeed, for D178N there was no significant inhibition of toxicity (P . 0.05). CD analysis of peptide–protein interaction The CD spectra of PrP121–231 and its mutants (in the presence or absence of an equimolar amount of PrP112– 125) were measured (Fig. 6). In general, and as previously described [47], the three point mutations did not alter the CD spectra markedly (Fig. 6A). PrP112–125 has previously been shown to have a predominantly random coiled structure with some amount of b sheet (Fig. 4). The mixture of wild-type PrP121–231 and PrP112–125 resulted in a unique CD spectrum that did not resemble that of PrP121–231, PrP112 –125 or an arithmetic sum of the two. Fig. 5. Toxicity of PrP121–231 mutants. Cerebellar cell cultures from wild-type (A) and Zrk-Prnp8 / 8 (B) cerebellar cells were grown in serum free medium and treated with mutants of PrP121–231 at different concentrations for 7 days. After that time the survival of the cultures was measured with an MTT assay. The values were compared to those for untreated cultures as a percentage. Shown are wild-type (W) D178N (K), F198S (O) and E200K (X). (C) Npu-Prnp8 / 8 cerebellar cells were either treated with 50 m M of PrP121–231 and with increasing concentrations of PrP23– 231 (X), PrP23–112 (W), PrP105– 231 (K) or PrP112–231 (O). After 5 days treatment the survival was measured using an MTT assay and comparing values to those of untreated cultures. Shown are the mean ^ SEM of four experiments with three determinations each. (D) Npu-Prnp8 / 8 cerebellar cells were either treated with PrP112–125 at varying concentrations alone (B)or with the addition of 50 m M of PrP121–231 (W) or PrP121–231 mutants, D178N (K), F198S (O) or E200K (X). After 5 days treatment the survival was measured using an MTT assay and comparing values to those of untreated cultures. Shown are the mean ^ SEM of four experiments with three determinations each. Fig. 6. Circular dichroism study of prion proteins. Proteins were freshly prepared in 10 m M phosphate buffer pH 7.4 at 0.1 mg : mL 21 . Wild-type PrP121–231 (WT) and three mutants (E200K, F198S, D178N) were measured either on their own (compared in A) or mixed with an equimolar amount of PrP112–125 (B– E). The spectra of the wild-type form of PrP121–231 and the mutants on their own (thin line) are compared to spectra of the same proteins mixed with PrP112–125 (thick line). In addition the spectra for PrP105– 231 (F) and PrP113– 231 (G) were also determine on their own (thin line) or mixed with PrP112–125 (thick line). The increased levels of structural elements were determined by a CD spectra analysis program. Five sweeps were collected at 0.5-nm intervals. Spectra are presented as molar ellipticity [u]. The increased helical content in all the proteins induced by PrP112–125 was determined and plotted in (H). 6160 M. Daniels et al. (Eur. J. Biochem. 268) q FEBS 2001 PrP112–125 had similar but lesser effect on the mutant proteins E200K and F198S. There was little effect on the spectrum of D178N. In particular, the mixtures showed relative increases in helical content (Fig. 6H). This change in spectrum possibly represents a change in the structure of the protein as a result of an interaction between the protein and the peptide. In order to determine if the structure produced by the interaction of PrP112–125 and PrP121–231 represents a ‘true’ PrP structure, the CD spectra of PrP105–231 and PrP113–231 was determined (Fig. 6). The CD spectra of both proteins appear similar but dissimilar to the mixtures between PrP121–231 and PrP112–125. The spectra of PrP113–231 and PrP105–231 were similar to that of PrP121–231. However, both spectra show a loss of a strong minima at 222 nm and an increase at 230 nm. However, mixing these proteins with PrP112–125 did not cause any change in the spectrum except a slight increase at 230 nm. There was no increase in helical content (Fig. 6H). DISCUSSION The precise mechanism of neurodegeneration in prion disease is unknown. Much research has concentrated on a peptide PrP106–126 to understand neuronal death in models of these diseases. Until this report there has been no complete investigation of the prion protein for toxic domains. It has been known for some time that the N-terminus (amino-acid residues 23 –90) is not associated with disease. We aimed to investigate whether the C-terminal domain has inherent toxicity to neurones because a recent publication by Shmerling et al. [42] has shown that mice expressing a C-terminal fragment of PrP c (either deletion of amino-acid residues 32–121 or 32– 134) die within several weeks of birth due to massive neuro- degeneration in the granule cell layer of the cerebellum. A different C-terminal fragment of PrP c (PrP112–231) is generated by normal cleavage of the protein [15]. It has been suggested that this fragment can be cycled back to the cell surface where it may have a function independent of the full length molecule [48]. We found that the fragment PrP121–231 was highly neurotoxic. This seemed to us to be paradoxical as one would not expect a normal cleavage product to be potently neurotoxic. However, slightly longer peptides (PrP105–231 and PrP113–231) showed no toxicity. The region of the mouse protein containing amino-acid residues 105 –125 is almost identical to the sequence of human PrP, which is known as PrP106–126, the neurotoxic peptide used by many laboratories as a model of the neurotoxicity of PrP Sc . However, this region of the protein, as part of a C-terminal fragment of PrP c , has no demonstrable toxicity. Furthermore, addition of a peptide with amino-acid residues 112–125 of the mouse PrP c sequence to cerebellar cultures in parallel with PrP121–231 neutralized the toxicity of PrP121 –231. The same result was obtained with mouse PrP121–231 and the human PrP106–126 (data not shown). Under these conditions, the PrP106–126 toxicity was fully inhibited. Additionally we also showed that PrP23–231 inhibits toxicity but that neither the N-terminal fragment (PrP23–112) nor the full C-terminal fragment (PrP113–231) inhibit the toxicity. Studies of the N-terminus have shown that it is unstructured and therefore it is unlikely to undergo a conformational change when PrP23–231 is cleaved. However, PrP113–231 contains a very hydrophobic N-terminus (amino-acid residues 113–135) and it is quite possible that this region would be masked if it interacted with other hydrophobic residues in the rest of the protein. The normal cleavage product of PrP c metabolism is a protein cleaved at around amino-residue 112–114 and not 121–122 [15,17,18]. Therefore separation of the hydro- phobic, palindromic region of 112 –120 from the rest of the C-terminus is artefactual. Such a fragment of the protein has never been shown to exist in vivo. Furthermore, the paper by Shmerling et al. [42] demonstrates that expression of a fuller C-terminal cleavage product with a deletion only of residues 32–106 does not develop pathology. The toxicity of the C-terminal part of PrP c has never been apparent before the publication by Shmerling et al. [42] and most interest has focussed on the known neurotoxic region PrP106–126 whose toxicity came to attention because of study of the toxicity of PrP Sc [24]. If deletion of PrP c up to residue 121 creates a toxic protein then one would expect that deletion of the C-terminus of the protein should similarly cause prob- lems. Muramato et al. [41] showed that deletion of amino- acid residues 177–200 or 201–217 results in neurodegen- erative disease in mice expressing such modified PrPs. However, the disease was not as severe as that described by Shmerling et al. [42] and the cell death does not appear to be due to the toxicity of the hydrophobic domain (amino-acid residues 112–135). Our findings regarding PrP121–231 mediated neurotoxicity might explain the cerebellar degeneration described by Shmerling et al. [42]. However, although our results suggest that the neurodegeneration in this model could be the result of direct neurotoxic effects of PrP121–231, the full mechanism of this effect needs to be elucidated. The neurodegeneration seen in the 32–121 deleted mice [42] was prevented by the presence of coexpressed full- length PrP c . The previous interpretation of this result was that PrP c has a ligand that interacts with it at two domains. When the protein only interacts with PrP c at the C-terminal domain, the resulting signal transduction leads to death. However, the problem with this model is that the metabolic C-terminal domain of PrP c (PrP112–231) exists in neurones all the time. Even if it has a much lower affinity for the ‘ligand’ than full length PrP c then one would nevertheless be causing the death of cerebellar granular cells. A better hypothesis according to our data is that the C-terminal domain (PrP121–231) as expressed by Shmerling et al. [42] can interact with full length PrP c . This interaction via binding to the hydrophobic domain of PrP c then inhibits the toxicity of the C-terminus. The normal N-terminal domain of the C-terminal cleavage product of PrP c is conserved so highly as to be unchanged from reptile to man [49,50]. This is very unusual, especially for a region of the protein associated with neurotoxicity. Therefore it is clearly essential to the normal metabolism of PrP c . We have shown previously [11,37] that this region of the protein is the site of interaction between PrP c with PrP Sc . Therefore it is possible, under normal metabolic conditions, that the hydrophobic-palindromic region of PrP c interacts directly with the globular domain of the C-terminus of another molecule of PrP c . However, we consider it more likely that each molecule folds in such a way that the very N-terminus of PrP’s C-terminal cleavage q FEBS 2001 Novel PrP peptide toxicity (Eur. J. Biochem. 268) 6161 product folds into the globular domain. Due to the very hydrophobic nature of this region, this may occurs directly after cleavage of the protein. Our studies of toxicity used wild-type neurones and neurones deficient in PrP c expression in parallel. This allowed us to distinguish fragments or peptides that were toxic in absence of PrP c . This also allowed us to demonstrate that the toxicity of PrP121 – 231 is different to that of PrP Sc , as PrP Sc is nontoxic to cells lacking PrP c expression [21] but the toxicity of PrP121–231 is increased to Prnp8 / 8 cerebellar cells. In particular, the region of amino-acids 147–171 shows toxicity only to Prnp8 / 8 cerebellar cells and might explain the increased toxicity of PrP121–231 to Prnp8 / 8 cerebellar cells. PrP peptides based on the hydrophobic domain of PrP c have also been suggest to have effects not mediated directly through PrP c [28,30] but this is the first time the peptides based other regions of PrP have been shown to be toxic to Prnp8 / 8 neurones. Of importance are the studies with peptides that highlight the region of the C-terminus involved in the toxicity of PrP121–231 to cerebellar neurones. Two peptides, PrP163– 184 and PrP196–220, appeared to be the most toxic. This is of considerable interest as the majority of point mutations associated within inherited human prion disease are found within these regions (e.g. D178N, V180I, T183A, F198S, E200K, R208H, V201I, Q217R). Indeed, the two point mutations studied in this paper, E200K and F198S, both map to PrP196 –220 and both enhanced the toxicity of PrP121– 231. It is quite possible that these point mutations alter the ability of the globular domain to interact with the conserved hydrophobic region. This was demonstrated here by a decreased ability of PrP112–125 to inhibit the toxicity of the PrP121–231 mutants. Many of these point mutations lie along a hydrophobic cleft in the protein [47,51]. It is quite possible that the N-terminal domain of the C-terminal cleavage product will fold into this cleft. However, this report does not give any information on this aspect of the structure of the protein. Previously, NMR studies of PrP c have used either PrP121– 231 or PrP23–231 as the protein for analysis [47,52–55]. The general finding of these reports was that the C-terminal of the protein contains three helices that fold together to make a globular domain while the N-terminus has little structure. For both the full length protein and the C-terminal fragment the globular domain was very similar. However, the protein PrP121–231 does not represent the true C-terminal fragment of PrP as it does not include the perfectly conserved amino-acid residues 112– 120 which constitute the true N-terminus. Therefore it is probably of vital importance that the C-terminal fragment containing these additional amino acids be studied further, as the true C-terminal fragment of PrP c may form a novel conformation as yet unknown. Support for this comes from the analysis by CD of two proteins PrP105–231 and PrP113–231 that have moderately different spectra to that of PrP121–231. It has previously been suggested that such proteins prepared from bacteria are sensitive to protease digestion and the amino-acid residues up to 121 are cleaved off [53]. We took particular care to ensure that our proteins were not degraded in this way. Analysis of the effect of point mutations on the structure of PrP c have focussed on their effect on the protein PrP121– 231. We have observed that two mutations alter the toxicity of this protein but this is not likely to have direct relevance to prion disease as the protein PrP121–231 does not normally exist. However, if the amino-acid residues where these mutations occur are central to refolding of the protein after cleavage then it is possible that C-terminus could be toxic in the inherited prion diseases. Alternatively, or in addition, a misfolded C-terminal protein could somehow influence formation of PrP Sc . Another alternative is that interaction between the globular domain and the hydro- phobic-palindromic region influences cleavage of the protein. Then, a large amount of uncleaved mutant PrP c might remain and be sufficient to eventually aggregate and begin conversion to PrP Sc . In summary, we have shown that a C-terminal fragment of PrP c commonly used in the investigation of PrP c structure and known to induce neurodegeneration in mice lacking expression of the full length PrP c is neurotoxic. This neuro- toxicity does not require expression of PrP c . The neuro- toxicity can be inhibited by the presence of the hydrophobic- palindromic region (amino-acid residues 112–125) of the protein either applied as a separate peptide or attached to the protein. Known human point mutations enhance the toxicity of this PrP c fragment. However, PrP121–231 might not represent the true C-terminal cleavage produce of PrP c metabolism and re-investigation of the biology of the C-terminus of PrP c is necessary. ACKNOWLEDGEMENTS The authors than Charles Weissmann for the Zrk1 prion protein knockout mice and Jean Manson for the Npu prion protein knockout mice. Thanks also to Rudi Glockshuber for PrP121 –231 mutant proteins. In addition we thank Bill Broadhurst and Tim Daffron for help with the CD analysis. REFERENCES 1. Kretzschmar, H.A., Prusiner, S.B., Stowring, L.E. & DeArmond, S.J. (1986) Scrapie prion proteins are synthesized in neurons. Am. J. Pathol. 122, 1–5. 2. Moser, M., Colello, R.J., Pott, U. & Oesch, B. (1995) Develop- mental expression of the prion protein gene in glial cells. Neuron 14, 509–517. 3. Brown, D.R. (1999) Prion protein peptide neurotoxicity can be mediated by astrocytes. J. Neurochem. 73, 1105 –1113. 4. Brown, D.R., Besinger, A., Herms, J.W. & Kretzschmar, H.A. (1998) Microglial expression of the prion protein. Neuroreport 9, 1425–1429. 5. Diomede, L., Sozzani, S., Luini, W., Algeri, M., De Gioio, L., Chiese, R., Lievans, P.M., Bugiani, O., Forloni, G., Tagliavini, F. & Salmona, M. (1996) Activation effects of a prion protein fragment [PrP-(106–126)] on human leucocytes. Biochem. J. 320, 563–570. 6. Brown, D.R., Schmidt, B., Groschup, M.H. & Kretzschmar, H.A. (1998) Prion protein expression in muscle cells and toxicity of a prion protein fragment. Eur. J. Cell Biol. 75, 29–37. 7. Stahl, N., Borchelt, D.R. & Prusiner, S.B. (1990) Differential release of cellular and scrapie prion proteins from cellular membranes by phosphatidylinositol-specific phospholipase C. Biochemistry 29, 5405–5412. 8. Sale ` s, N., Rodolfo, K., Ha ¨ ssig, R., Faucheux, B., Di Giamberdino, L. & Moya, K.L. (1998) Cellular prion protein localization in rodent and primate brain. Eur. J. Neurosci. 10, 2464–2471. 9. Collinge, J., Whittington, M.A., Sidle, K.C.L., Smith, C.J., Palmer, M.S., Clarke, A.R. & Jefferys, J.G.R. (1994) Prion protein is necessary for normal synaptic function. Nature 370, 295–297. 6162 M. Daniels et al. (Eur. J. Biochem. 268) q FEBS 2001 10. Brown, D.R., Qin, K., Herms, J.W., Madlung, A., Manson, J., Strome, R., Fraser, P.E., Kruck, T.A., von Bohlen, A., Schulz- Schaeffer, W., Giese, A., Westaway, D. & Kretzschmar, H. (1997) The cellular prion protein binds copper in vivo. Nature 390, 684–687. 11. Brown, D.R., Clive, C. & Haswell, S.J. (2001) Anti-oxidant activity related to copper binding of native prion protein. J. Neurochem. 76, 69–76. 12. Brown, D.R. (1999) Prion protein expression aids cellular uptake and veratridine-induced release of copper. J. Neurosci. Res. 58, 717–725. 13. Brown, D.R. & Besinger, A. (1998) Prion protein expression and superoxide dismutase activity. Biochem. J. 334, 423 –429. 14. Brown, D.R., Wong, B S., Hafiz, F., Clive, C., Haswell, S.J. & Jones, I.M. (1999) Normal prion protein has an activity like that of superoxide dismutase. Biochem. J. 244, 1–5. 15. Chen, S.G., Teplow, D.B., Parchi, P., Teller, J.K., Gambetti, P. & Autilio-Gambetti, L. (1995) Truncated forms of the human prion protein in prion diseases. J. Biol. Chem. 270, 19173–19180. 16. Pan, K M., Stahl, N. & Prusiner, S.B. (1992) Purification and properties of the cellular prion protein from Syrian hamster brain. Protein Sci. 1, 1343– 1352. 17. Vincent, B., Paitel, E., Frobert, Y., Lehmann, S., Grassi, J. & Checler, F. (2000) Phorbol ester-regulated cleavage of normal prion protein in HEK293 human cells and murine neurons. J. Biol. Chem. 275, 35612– 35616. 18. Vincent, B., Paitel, E., Saftig, P., Frobert, Y., Hartmann, D., De Strooper, B., Grassi, J., Lopez-Perez, E. & Checler, F. (2001) The disintegrins ADAM10 and TACE contribute to the constitutive and phorbol-esters-regulated normal cleavage of the cellular prion protein. J. Biol. Chem. 276, 37743–37746. 19. Prusiner, S.B. (1998) Prions. Proc. Natl Acad. Sci. USA 95, 13363–13383. 20. Prusiner, S.B. (1982) Novel proteinaceous infectious particles cause scrapie. Science 216, 136–144. 21. Giese, A., Brown, D.R., Groschup, M.H., Feldmann, C., Haist, I. & Kretzschmar, H.A. (1998) Role of microglia in neuronal cell death in prion disease. Brain Pathol. 8, 449–457. 22. Brandner, S., Isenmann, S., Raeber, A., Fischer, M., Sailer, A., Kobayashi, Y., Marino, S., Weissmann, C. & Aguzzi, A. (1996) Normal host prion protein necessary for scrapie-induced neuro- toxicity. Nature 379, 339–343. 23. Bu ¨ eler, H., Aguzzi, A., Sailer, A., Greiner, R A.P.A., Auget, M. & Weissmann, C. (1993) Mice devoid of PrP are resistant to scrapie. Cell 73, 1339–1347. 24. Forloni, G., Angeretti, N., Chiesa, R., Monzani, E., Salmona, M., Bugiani, O. & Tagliavini, F. (1993) Neurotoxicity of a prion protein fragment. Nature 362, 543– 546. 25. Brown, D.R., Schmidt, B. & Kretzschmar, H.A. (1996) Role of microglia and host prion protein in neurotoxicity of a prion protein fragment. Nature 380, 345– 347. 26. Tagliavini, F., Prelli, F., Verga, L., Giaccone, G., Sarma, R., Gorevic, P., Ghetti, B., Passerini, F., Forloni, G., Salmona, M., Bugiani, O. & Frangione, B. (1993) Synthetic peptides homologous to prion protein residues 106 –147 form amyloid-like fibrils in vitro. Proc. Natl Acad. Sci. USA 90, 9678–9682. 27. Perovic, S., Schro ¨ der, H.C., Pergande, G., Ushijima, H. & Mu ¨ ller, W.E.G. (1997) Effect of flupirtine on Bcl-2 and glutathione level in neuronal cells treated in vitro with the prion protein fragment (PrP106–126). Exp. Neurol. 147, 518– 524. 28. Haı ¨ k, S., Peyrin, J.M., Lins, L., Rosseneu, M.Y., Brasseur, R., Langeveld, J.P., Tagliavini, F., Deslys, J.P., Lasme ´ zas, C. & Dormont, D. (2000) Neurotoxicity of the putative transmembrane domain of the prion protein. Neurobiol. Dis. 7, 644–656. 29. Fabrizi, C., Silei, V., Menegazzi, M., Salmona, M., Bugiani, O., Tagliavini, F., Suzuki, H. & Lauro, G.M. (2001) The stimulation of inducible nitric-oxide synthase by the prion protein fragment 106–126 in human microglia is tumor necrosis factor-alpha- dependent and involves p38 mitogen-activated protein kinase. J. Biol. Chem. 276, 25692– 25696. 30. Pillot, T., Drouet, B., Pinc¸on-Raymond, M., Vanderkerckhove, J., Rosseneu, M. & Chambaz, J. (2000) A nonfibrillar form of the fusogenic prion protein fragment [118–135] induces apoptotic cell death in rat cortical neurons. J. Neurochem. 75, 2298–2308. 31. White, A.R., Guirguis, R., Brazier, M.W., Jobling, M.F., Hill, A.F., Beyreuther, K., Barrow, C.J., Masters, C.L., Collins, S.J. & Cappai, R. (2001) Sublethal concentrations of prion peptide PrP106– 126 or the amyloid beta peptide of Alzheimer’s disease activates expres- sion of proapoptotic markers in primary cortical neurons. Neuro- biol. Dis. 8, 299–316. 32. Brown, D.R., Herms, J. & Kretzschmar, H.A. (1994) Mouse cortical cells lacking cellular PrP survive in culture with a neuro- toxic PrP fragment. Neuroreport 5, 2057 –2060. 33. Brown, D.R., Schulz-Schaeffer, W.J., Schmidt, B. & Kretzschmar, H.A. (1997) Prion protein-deficient cells show altered response to oxidative stress due to decreased SOD-1 activity. Exp. Neurol. 146, 104–112. 34. Brown, D.R., Herms, J.W., Schmidt, B. & Kretzschmar, H.A. (1997) Different requirements for the neurotoxicity of fragments of PrP and b-amyloid. Euro. J. Neurosci. 9, 1162 –1169. 35. Brown, D.R., Pitschke, M., Riesner, D. & Kretzschmar, H.A. (1998) Cellular effects of a neurotoxic prion protein peptide are related to its b-sheet content. Neurosci. Res. Comm. 23, 119– 128. 36. Brown, D.R. (2000) Prion protein peptides: optimal toxicity and peptide blockade of toxicity. Mol. Cell Neurosci. 15, 66–78. 37. Brown, D.R. (2000) PrP Sc -like prion protein peptide inhibits the function of cellular prion protein. Biochem. J. 252, 511 –518. 38. Ettaiche, M., Pichot, R., Vincent, J P. & Chabry, J. (2000) In vivo cytotoxicity of prion protein fragment PrP106– 126. J. Biol. Chem. 275, 36487– 36490. 39. Supattapone, S., Bosque, P., Muramoto, T., Wille, H., Aagaard, C., Peretz, D., Nguyen, H O.B., Heinrich, C., Torchia, M., Safar, J., De Cohen, F.E.A., Rmond, S.J., Prusiner, S.B. & Scott, M. (1999) Prion protein of 106 residues creates an artificial transmission barrier for prion replication in transgenic mice. Cell 96, 869– 878. 40. Fischer, M., Ru ¨ licke, T., Raeber, A., Sailer, A., Moser, M., Oesch, B., Brandner, S., Aguzzi, A. & Weissmann, C. (1996) Prion protein (PrP) with amino-proximal deletions restoring susceptibility of PrP knockout mice to scrapie. EMBO J. 15, 1255–1264. 41. Muramato, T., DeArmond, S.J., Scott, M., Telling, G., Cohen, F.E. & Prusiner, S.B. (1997) Heritable disorder resembling neuronal storage disease in mice. Nat. Med. 3, 750–756. 42. Shmerling, D., Hegy, I., Fischer, M., Bla ¨ ttler, T., Brandner, S., Go ¨ tz, J., Ru ¨ licke, T., Flechsig, E., Cozzio, A., von Mering, C., Hangartner, C., Aguzzi, A. & Weissmann, C. (1998) Expression of amino-terminally truncated PrP in the mouse leading to ataxia and specific cerebellar lesions. Cell 93, 203– 214. 43. Moore, R.C., Lee, I.Y., Silverman, G.L., Harrison, P.M., Strome, R., Heinrich, C., Karunaratne, A., Pasternak, S.H., Chishti, M.A., Liang, Y., Mastrangelo, P., Wang, K., Smit, A.F., Katamine, S., Carlson, G.A., Cohen, F.E., Prusiner, S.B., Melton, D.W., Tremblay, P., Hood, L.E. & Westaway, D. (1999) Ataxia in prion protein (PrP) -deficient mice is associated with upregulation of the novel PrP-like protein doppel. J. Mol. Biol. 292, 797–817. 44. Nishida, N., Tremblay, P., Sugimoto, T., Shigematsu, K., Petromilli, C., Erpel, S.P., Nakaoke, R., Atarashi, R., Hountani, T., Sakaguchi, S., DeArmond, S.J., Prusiner, S.B. & Katamine, S. (1999) A mouse prion protein transgene rescues mice deficient for the prion protein gene from purkinje cell degeneration and demyelination. Lab. Invest. 79, 689 –697. 45. Manson, J.C., Clarke, A.R., Hooper, M.L., Aitchison, L., McConnell, I. & Hope, J. (1994a) 129/Ola mice carring a null q FEBS 2001 Novel PrP peptide toxicity (Eur. J. Biochem. 268) 6163 mutation in PrP that abolishes mRNA production are develop- mentally normal. Mol. Neurobiol. 8, 121 –127. 46. Bu ¨ eler, H., Fischer, M., Lang, Y., Bluethmann, H., Lipp, H P., DeArmond, S.J., Prusiner, S.B., Aguet, M. & Weissmann, C. (1992) Normal development and behaviour of mice lacking the neuronal cell-surface PrP protein. Nature 356, 577–582. 47. Liemann, S. & Glockshuber, R. (1999) Influence of amino acid substitutions related to inherited human prion diseases on the thermodynamic stability of the cellular prion protein. Biochemistry 38, 3258– 3267. 48. Harris, D.A., Tuber, M.T., van Dijken, P., Shyng, S L., Chait, B.T. & Wang (1993) Processing of a cellular prion protein: identification of N- and C- terminal cleavage sites. Biochemistry 32, 1009– 1016. 49. Simonic, T., Duga, S., Strumbo, B., Asselta, R., Ceciliani, F. & Ronchi, S. (2000) cDNA cloning of turtle prion protein. FEBS Lett 469, 33–38. 50. Wopfner, F., Wiedenho ¨ fer, G., Schneider, R., von Bunn, A., Gilch, S., Schwarz, T.F., Werner, T. & Scha ¨ tzl, H.M. (1999) Analysis of 27 mammalian and 9 avian PrPs reveals high conservation of flexible regions of the prion protein. J. Mol. Biol. 289, 1163– 1178. 51. Riek, R., Wider, G., Billeter, M., Hornemann, S., Glockshuber, R. &Wu ¨ thrich, K. (1998) Prion protein NMR structure and familial human spongiform encephalopathies. Proc. Natl Acad. Sci. USA 95, 11667– 11672. 52. Riek, R., Hornemann, S., Wider, G., Billeter, M., Glockshuber, R. &Wu ¨ thrich, K. (1996) NMR structure of the mouse prion protein domain PrP (121 –321). Nature 382, 180– 182. 53. Hornemann, S. & Glockshuber, R. (1996) Autonomous and reversible folding of a soluble amino-terminally truncated segment of the mouse prion protein. J. Mol. Biol. 261, 614 –619. 54. Zahn, R., Liu, A., Lu ¨ hrs, T., Riek, R., von Schroetter, C., Lo ´ pez, F., Mailleter, M., Calzolai, L., Wider, G. & Wu ¨ thrich, K. (2000) NMR solution structure of the human prion protein. Proc. Natl Acad. Sci. USA 97, 145 –150. 55. Hornemann, S., Korth, C., Oesch, B., Riek, R., Wider, G., Wu ¨ thrich, K. & Glockshuber, R. (1997) Recombinant full-length murine prion protein mPrP (23 –231): purification and spectro- scopic characterization. FEBS Lett. 413, 277–281. 6164 M. Daniels et al. (Eur. J. Biochem. 268) q FEBS 2001 . Toxicity of novel C-terminal prion protein fragments and peptides harbouring disease-related C-terminal mutations Maki Daniels 1 , Grazia Maria Cereghetti 2 and David R. Brown 1 1 Department. Neurotoxicity of a prion protein fragment. Nature 362, 543– 546. 25. Brown, D.R., Schmidt, B. & Kretzschmar, H.A. (1996) Role of microglia and host prion protein in neurotoxicity of a prion protein fragment methyl ester had only a minor effect on the toxicity of the protein. Addition of microglia had no signifi- cant effect on the toxicity of PrP121– 231. These results suggest that the toxicity of