Prion Proteins with Insertion Mutations Have Altered N-terminal Conformation and Increased Ligand Binding Activity and Are More Susceptible to Oxidative Attack*

We compared the biochemical properties of a wild type recombinant normal human cellular prion protein, rPrPc, with a recombinant mutant human prion protein that has three additional octapeptide repeats, rPrP8OR. Monoclonal antibodies that are specific for the N terminus of rPrPc react much better with rPrP8OR than rPrPc, suggesting that the N terminus of rPrP8OR is more exposed and hence more available for antibody binding. The N terminus of PrPc contains a glycosaminoglycan binding motif. Accordingly, rPrP8OR also binds more glycosaminoglycan than rPrPc. In addition, the divalent cation copper modulates the conformations of rPrPc and rPrP8OR differently. When compared with rPrPc, rPrP8OR is also more susceptible to oxidative damage. Furthermore, the abnormalities associated with rPrP8OR are recapitulated, but even more profoundly, in another insertion mutant, which has five extra octapeptide repeats, rPrP10OR. Therefore, insertion mutants appear to share common features, and the degree of abnormality is proportional to the number of insertions. Any of these anomalies may contribute to the pathogenesis of inherited human prion disease.

PrP Sc . The accumulation of PrP Sc in the central nervous system is then thought to impair function, induce structural damage, and cause disease.
At least 20 different pathogenic mutations in the human PRNP gene have been identified (7). These are either insertional or point mutations. Insertion mutation occurs solely in the octapeptide repeat region, which is involved in the binding of divalent cations such as copper and zinc. Human PrP c has five octapeptide repeats. In pathogenic mutations, the number of additional octapeptide repeats ranges from one to nine (8,9). A transgenic mouse line with nine additional octapeptide repeats spontaneously develops neurodegeneration but does not produce infectious PrP Sc (10,11).
Although it is clear that mutation in the prion gene causes human prion diseases, very little is known about the mechanisms by which the mutant protein causes disease. Prion proteins with pathogenic mutations may cause disease because of gain of toxic functions, loss of normal physiologic functions, or both.
Recently, we identified a novel insertion mutation in a Chinese family, whose PRNP has three additional octapeptide repeats (12). Subsequently, another family with the identical mutation was identified in Europe (13). In this study, we report studies comparing the conformation of insertion mutant recombinant protein with wild type recombinant protein using a panel of well characterized anti-PrP monoclonal antibodies. We also studied the effects of the mutation on binding to glycosaminoglycan (GAG) and copper, two known ligands of PrP c , as well as susceptibility to oxidative insults. Finally, we investigated whether our findings with the three extra octapeptide repeat mutant proteins are applicable to another insertion mutant protein with five extra octapeptide repeats.

EXPERIMENTAL PROCEDURES
Plasmid Constructions-The human wild type rPrP c cDNA gene corresponding to the putative mature fragment  and the human pathogenic mutant rPrP 8OR with three extra octapeptide repeats encoding residues  were amplified by polymerase chain reaction (PCR) using human genomic DNA templates (12). The primers were 5Ј-ATCCATATGAAGAAGCGCCCGAAGCCTG-3Ј (forward sequence) and 5Ј-ACCGGAATTCCTAGCTCGATCCTCTCTGG-3Ј (reverse sequence). The PCR product was cloned between restriction sites NdeI and EcoRI of the vector pET42(ϩ) (Novagen) termed pET-rPrP c and pET-rPrP 8OR . A similar procedure was also carried out to obtain pET-rPrP 10OR , which contains five extra octapeptide repeats, using the human prion insertion mutant DNA template (14). A deletion mutant of rPrP c , which lacks the N-terminal GAG binding motif (resi-* This work was supported in part by National Institutes of Health Grant NS-045981-01 and by United States Department of Army Award DAMD17-03-1-0286 (to M.-S. S.).The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1  dues 23-35) (15), designated as rPrP ⌬23-35 was constructed from pET-rPrP c using QuikChange site-directed mutagenesis kit. Primers were 5Ј-TACATATGGGCAGCCGATACCC-3Ј (forward sequence) and 5Ј-CGGGTATCGGCTGCCCATATGTATATCTCC-3Ј (reverse sequence). The insertion sequences were verified by using an Applied Biosystems 373A automated sequencer. Generation and Purification of Recombinant Human rPrP C , rPrP 8OR , and rPrP 10OR -The purification and refolding process was carried out on a nickel ion-charged Sepharose column (Amersham Biosciences) as described previously (16). Briefly, freshly transformed Escherichia coli BL21 (DE3) (Novagen) containing the plasmid pET-rPrP c or pET-rPrP 8OR was transferred to 1 liter of Luria broth (LB) medium supplemented with 50 g/ml kanamycin at 37°C until the A 600 reached 1.0 and induced overnight with 1 mM isopropyl ␤-D-thiogalactoside. Bacteria were harvested by centrifugation at 4,000 ϫ g for 15 min at 4°C, resuspended in 20 mM Tris/HCl (pH 8.0), 150 mM NaCl, 0.1 mM phenylmethylsulfonyl fluoride, 1 mg/ml lysozyme, and incubated at 21°C for 2 h before further lysis by sonication. Samples were centrifuged at 13,000 ϫ g for 20 min, and the protein pellets were extensively washed using 1% Triton X-100 and then resuspended in 20 mM Tris/HCl (pH 8.0), 8 M urea, and 10 mM 2-mercaptoethanol. The inclusion bodies were purified and refolded on a nickel ion-charged Sepharose column by decreasing urea gradient concentrations. rPrP was eluted by 50 mM sodium acetate (pH 4.0) followed by dialysis against PBS (pH 7.4). A similar procedure was used to obtain rPrP 10OR and rPrP ⌬23-35 . Protein concentration was determined with a Bio-Rad protein assay kit.
Capture ELISA-Capture antibody (8H4 or 11G5) was coated at 50 ng/well in flat-bottomed, 96-well Costar plates (Corning) overnight at 4°C. Excess antibody was removed, and wells were blocked with phosphate-buffered saline (PBS) containing 3% BSA (Sigma) for 3 h at room temperature. Plates were washed three times with PBS (pH 7.4) containing 0.05% Tween 20 (PBS-T), and recombinant human rPrP c , rPrP 8OR , or rPrP 10OR (1 g/ml) in PBS was added in triplicate to respective wells overnight at 4°C. Plates were washed three times with PBS-T, and appropriate dilutions of biotinylated mAbs were added to designated wells for further 6 h at room temperature. Plates were washed three times with PBS-T, and HRP-conjugated streptavidin at 1:10,000 was added for 1 h. ELISA plates were washed twice with PBS-T before adding the substrate p-nitrophenyl phosphate at 0.5 mg/ml. ELISA plates were read at 405 nm on a Beckman Coulter AD340 micro-ELISA plate reader.
To study the effects of divalent copper on the mAb binding profiles to rPrP c or rPrP 8OR , ELISA plates were coated overnight with 50 ng/well mAb 8H4, and unbound sites were blocked using 3% BSA. After washing three times, rPrP c or rPrP 8OR (1 g/ml) was added for 6 h. Plates were washed three times, and different concentrations of CuSO 4 were added for overnight incubation at room temperature with gentle agitation (21). The plates were washed with PBS-T, and biotinylated SAF32, 11G5, 2C2, or 7A12 (0.1 g/ml) was added for 6 h. After extensive washing, HRP-conjugated streptavidin was added for reading at 405 nm. Antibody binding is given as a percentage of the increased binding signal compared with untreated sample.
Detection of rPrP Binding to GAG-GAG binding experiments were performed as described previously (15). Briefly, chondroitin sulfate (from bovine trachea), heparin sulfate (from porcine intestinal mucosa), and heparin (from porcine intestinal mucosa) (all from Sigma) were coated on plates at 10 g/well at 4°C overnight and blocked with 3% BSA in PBS at room temperature for 3 h. Plates coated with BSA were used as controls. Appropriate dilutions of rPrP c , rPrP 8OR , and rPrP 10OR were incubated with the coated plates for 2 h. After three washes with PBS-T, bound rPrP was detected with mAb 8H4. All experiments were carried out in triplicate and repeated at least three times.
Detection of Carbonyl Groups on rPrP-Protein oxidation was measured by assaying carbonyl groups on proteins using the OxyBlot protein oxidation detection kit (Intergen, Purchase, NY) as described (22). To produce oxidative damages, equal amounts of rPrP c , rPrP 8OR , or rPrP 10OR (1 g) was incubated with different concentrations of H 2 O 2 at room temperature for 30 min. Samples were then derivatized to 2,4dinitrophenyl (DNP) by reaction with 2,4-dinitrophenyl hydrazine (DNPH) at room temperature for 30 min before loading onto SDSpolyacrylamide gels. The proteins were electrotransferred, probed with a rabbit antiserum against DNP-modified carbonyl groups, and visualized using the chemiluminescence blotting system (Roche Applied Science). No immunoreactivity was detected in non-DNPH-modified samples.

Characterization of Recombinant Wild Type Prion Protein (rPrP C )
and Mutant Prion Protein (rPrP 8OR )-We analyzed the two recombinant proteins by SDS-PAGE under either non-reducing or reducing conditions. In both conditions, rPrP c appears as a single band with a molecular mass of approximately 23 kDa, which is the expected molecular mass of full-length rPrP c protein. rPrP 8OR also appears as a single band but migrates a bit slower, with a molecular mass of 25 kDa, reflecting the addition of 24 amino acids in the octapeptide repeat region (Fig. 1A). Neither rPrP c nor rPrP 8OR preparation contains larger molecular species. To further characterize the two recombinant proteins, we immunoblotted them with three different anti-PrP mAbs. The binding epitopes of these mAbs are presented diagrammatically in Fig. 1B. The three mAbs, 8B4 (residues [35][36][37][38][39][40][41][42][43][44][45], 8H4 (residues 175-185), and 8F9 (residues 220 -231), reacted equally with rPrP c and rPrP 8OR (Fig. 1C).
Conformational Differences between rPrP C and rPrP 8OR ; N terminus of rPrP 8OR Is More Available for mAb Binding-We next studied the conformations of rPrP c and rPrP 8OR by comparing their mAb binding profiles, using a panel of mAbs in a capture ELISA format (Fig. 2). mAb 8H4 (Fig. 2, A-G) or mAb 11G5 (Fig. 2H) was immobilized on ELISA plates to capture either rPrP c or rPrP 8OR . Biotinylated anti-PrP mAbs with dissimilar specificities were then added in different concentrations to react with the captured proteins. Two biotinylated mAbs, 8B4 and 5B2, with epitopes at the N terminus reacted much stronger with rPrP 8OR than rPrP c (Fig. 2, A and B). mAb SAF32 reacts with an epitope within the octapeptide repeat region. At lower concentrations, biotinylated SAF32 reacted stronger with rPrP c than rPrP 8OR (Fig. 2C). Five other biotinylated mAbs, 7A12, 7H6, 11G5, 2C2, and 8H4, which react with epitopes in either the central region or the C terminus, reacted similarly with rPrP c and rPrP 8OR (Fig. 2, D, E, F, G, and H, respectively). Overall, these results FIGURE 2. Antibody binding profiles of rPrP c and rPrP 8OR . One hundred ng/well rPrP c or rPrP 8OR was captured on ELISA plates, which had been coated with 50 ng/well affinity purified mAb 8H4. After washing, various concentrations of biotinylated mAbs 8B4 (A), 5B2 (B), SAF32 (C), 7A12 (D), 7H6 (E), 11G5 (F), and 2C2 (G) were added to the respective wells. The binding of each mAb to rPrP c (f) and rPrP 8OR (F) is presented as A 405 values. H, ELISA plates were precoated with mAb 11G5, and biotinylated 8H4 was used to detect the captured rPrP. BSA was used as the negative control for each panel. The data presented are the means Ϯ S.E., and all experiments were performed at least three times. suggest that the N terminus of rPrP 8OR is more exposed and thus more available for binding of N terminus-specific mAbs. The addition of three octapeptide repeats also altered the conformation of the octapeptide repeat region but in a more subtle way.
Binding of rPrP c and rPrP 8OR to GAGs-All mammalian PrP c contains a GAG binding motif, KKRPK, the first five amino acids at the N terminus (15). If the N terminus of rPrP 8OR is more exposed, it should bind better to GAG. To test this hypothesis, ELISA plates were coated with three different GAGs: chondroitin sulfate B (Fig. 3A), heparin sulfate (Fig. 3B), and heparin (Fig. 3C). After extensive washing, different amounts of rPrP c , rPrP 8OR , or a deletion mutant protein, rPrP ⌬23-35 lacking the N-terminal GAG binding motif, were added. mAb 8H4 was then used to detect bound proteins. It is clear that rPrP 8OR binds the three GAGs much better than rPrP c . On the other hand, rPrP ⌬23-35 , which lacks the KKRPK motif, failed to bind any of the GAGs (Fig. 3). These results provide additional support for our conclusion that the N terminus of rPrP 8OR is more exposed and thus more available for ligand binding. Furthermore, these results provide conclusive evidence that the N terminus of PrP c contains the predominant GAG binding motif.
Cu 2ϩ Has Different Effects on the Conformations of rPrP C and rPrP 8OR -Addition of Cu 2ϩ increases the binding of mAbs 11G5 and 2C2 to brain-derived mouse PrP c suggesting that binding of Cu 2ϩ results in conformational changes in PrP c (21). We therefore investigated whether Cu 2ϩ also alters the binding profiles of these mAbs to rPrP c and rPrP 8OR (Fig. 4). Similar to our findings with native, brain-derived mouse PrP c , the addition of Cu 2ϩ also significantly increased the binding of mAbs 11G5 and mAb 2C2 (Fig. 4, B and C) but not the binding of mAb SAF32 or mAb 7A12 to rPrP c (Fig. 4, A  and D). In contrast, adding Cu 2ϩ did not increase the binding of mAb 11G5 or mAb 2C2 to rPrP 8OR (Fig. 4, B and C).
We next investigated whether Cu 2ϩ also influences the conformation of the N terminus. In the presence of Cu 2ϩ , rPrP c but not rPrP 8OR binds more mAb 8B4 and GAG in a Cu 2ϩ concentration-dependent manner (Fig. 4, E and F). These results suggest that Cu 2ϩ affects the conformation of rPrP c and rPrP 8OR differently. rPrP 8OR Is More Susceptible to Oxidative Damage-Protein oxidation results in the generation of carbonyl groups that can be chemically modified with DNPH. The degree of DNPH modification correlates with the level of protein oxidation (23). Histidine residues in rPrP c are highly susceptible to oxidation (24). Therefore, the additional histidine residues in the octapeptide repeat region of rPrP 8OR may render the mutant protein more susceptible to oxidation. We determined whether rPrP 8OR is indeed more prone to oxidation. Equal amounts of rPrP c and rPrP 8OR were modified with DNPH, separated by SDS-PAGE, and then immunoblotted with an anti-DNP antibody (22). Anti-DNP immunoreactivity was detected only on rPrP 8OR but not on rPrP c (Fig. 5A, lanes  1 and 5). Therefore, rPrP 8OR contains carbonyl groups, which are signatures of prior protein oxidation.
We next investigated whether rPrP 8OR is also more susceptible to the oxidation agent, H 2 O 2 . Equal amounts of rPrP c or rPrP 8OR was incubated with different concentrations of H 2 O 2 ranging from 1 to 100 M and subjected to DNPH modification. Subsequently, each sample was divided into three aliquots: one was immunoblotted with anti-DNP antibody; one was immunoblotted with anti-PrP mAb 8B4; and the third one was immunoblotted with anti-PrP mAb 7A12. DNP immunoreactivity was detected only on rPrP 8OR and in an H 2 O 2 concentrationdependent manner (Fig. 5A, lanes 6 -8). All samples reacted equally with anti-PrP mAbs 7A12 and 8B4 indicating that they had comparable amounts of protein (Fig. 5, B and C).
Selective Modification of the N terminus by High Concentrations of H 2 O 2 -In subsequent experiments, we found that mAb 8B4 was unable to bind rPrP 8OR when rPrP 8OR was exposed to very high concentrations of H 2 O 2 (such as 1 M) (Fig. 6A). Under identical conditions, the epitopes of mAbs 7A12 and 8F9, which were located at the central region and the C terminus respectively, remained intact (Fig.  6, B and C). Similar treatment did not alter the binding of mAb 8B4 to rPrP c (Fig. 6A). These results are consistent with our interpretation that the N terminus of rPrP 8OR is more exposed and thus is more susceptible to oxidation.
Findings with rPrP 8OR Are Applicable to rPrP 10OR -We next sought to determine whether some of the aberrant features found in rPrP 8OR are applicable to another insertion mutant protein that has five extra octapeptide repeats, rPrP 10OR . When immunoblotted with mAb 8B4, rPrP 10OR migrates a bit slower than rPrP 8OR , reflecting the addition of more octapeptide repeats (Fig. 7A). Similar to rPrP 8OR , rPrP 10OR also reacts more strongly with N terminus specific mAb 8B4 in captured ELISA (Fig. 7B), binds more GAG (Fig. 7C), and is more susceptible to oxidative damage (Fig. 7D). In all cases, the levels of abnormality are more profound in rPrP 10OR than in rPrP 8OR . At 1 mM H 2 O 2, aggregates with higher molecular masses were detected in rPrP 10OR but not in rPrP 8OR suggesting that aggregate formation may depend on the num- . rPrP 8OR binds more GAG than rPrP c . The binding of rPrP c , rPrP 8OR , and a recombinant human PrP, rPrP ⌬23-35 lacking the KKRPK GAG binding motif, to GAG was compared by ELISA (15). ELISA plates were coated with 10 g/ml different GAGs: chondroitin sulfate B (A); heparin sulfate (B); and heparin (C). Various concentrations of rPrP c (f), rPrP 8OR (F), and rPrP ⌬23-35 (OE) were added to respective wells. After washing, bound rPrPs were detected using mAb 8H4 followed by HRP-conjugated goat anti-mouse IgG Fc-specific antiserum. The data presented are the means Ϯ S.E. of three experiments each performed at least three times independently.
ber of octapeptide inserts. Collectively, these findings suggest that all prion proteins with insertion mutations share common features, and the degrees of abnormality are proportional to the number of insertions.

DISCUSSION
In this study, we describe four new findings on the biochemical properties of a recombinant human prion protein with three extra octapep-tide repeats, rPrP 8OR . The mutant protein shows the following aberrant features: 1) its N terminus is more exposed; 2) it binds better to GAG; 3) it behaves differently after binding to Cu 2ϩ ; and 4) it is more susceptible to oxidative attack. Importantly, some of the aberrant properties associated with rPrP 8OR are also observed in another insertion mutant prion protein with five extra repeats, rPrP 10OR , and the aberrations are even more profound in rPrP 10OR . . rPrP 8OR is more susceptible to oxidative damage than rPrP c . A, 1 g of rPrP c and rPrP 8OR was incubated with 0, 1, 10, and 100 M H 2 O 2 at room temperature for 30 min. The oxidative marker carbonyl groups in rPrP samples were derivatized to DNP by reaction with DNPH and subjected to Western blot using anti-DNP antibody. Aliquots of the above samples were also subjected to Western blot using mAbs 8B4 (B) and 7A12 (C). FIGURE 6. The mAb 8B4 reactive epitope on rPrP 8OR is selectively eliminated in the presence of 1 M H 2 O 2 . 1 g of rPrP c or rPrP 8OR was incubated with 0, 10, 100, and 1,000 mM H 2 O 2 at room temperature for 30 min. Samples were separated by SDS-PAGE and immunoblotted with mAb 8B4 (A), 7A12 (B), and 8F9 (C). In A, at 1 M H 2 O 2 , the mAbs 8B4 reactive epitope was no longer available for antibody binding for rPrP 8OR but remained intact for rPrP c . Under the same conditions, there were no effects on 7A12 (B) and 8F9 (C) binding epitopes for both rPrP c and rPrP 8OR . FIGURE 4. Copper has differential effects on the mAb binding profiles rPrP c and rPrP 8OR . One hundred ng/well either rPrP c or rPrP 8OR was captured on ELISA plates coated with 50 ng/well mAb 8H4. After washing, the bound rPrPs were incubated with PBS or various concentrations of CuSO 4 in PBS, and plates were incubated overnight at room temperature. Bound rPrPs were detected using different biotinylated mAbs: SAF32 (A); 11G5 (B); 2C2 (C); 7A12 (D); and 8B4 (E). For binding to GAG (F), plates were coated with heparin (10 g/ml), and bound rPrPs were detected using mAb 8H4 followed by HRP-conjugated goat anti-mouse IgG Fc-specific antiserum. Antibody binding is given as an increased percentage of optical density values. Under PBS-treated conditions, the mean optical density values for rPrP c and rPrP 8OR were 2. NMR analyses of rPrP c at pH 4.5 and 5.5 show that the N terminus is highly flexible and unstructured; in contrast, the C-terminal region has a well structured globular domain (25,26). The finding that half of PrP c lacks considerable secondary and tertiary structure is unusual and intriguing. Earlier transgenic mouse studies suggested that the N terminus of PrP c is not required for pathogenesis (27). However, more recent studies suggest that the N terminus of PrP c is important in the pathogenesis of prion diseases. The N terminus modifies disease phenotypes (28) and influences the conformations of protease-resistant PrP Sc generated in vivo and in vitro (29,30) and PrP c aggregation (31). Furthermore, it has been reported that the N terminus of PrP is structured if the studies are carried out at pH values between 6.5 and 7.8, i.e. the pH at cell membrane (32). Therefore, the roles the N terminus plays in the pathogenesis of prion diseases are complex and not completely understood.
Our finding that the N termini of rPrP 8OR and rPrP 10OR are more available for mAb 8B4 binding suggests that the octapeptide repeat region of PrP c also influences the conformation of the N terminus. Furthermore, the effect is proportional to the number of insertions; the longer the insertion, the greater the effect.
Insertion of three extra octapeptide repeats definitely alters the conformation of the octapeptide repeat region. The epitope for mAb SAF32 resides between amino acid residues 63 and 94 (19). When the concentration of SAF32 is limiting, SAF32 binds much less rPrP 8OR than rPrP c . This result suggests that binding of SAF32 is a cooperative phenomenon. There is more than one SAF32 binding site within the octapeptide repeat region; binding of one SAF32 may alter the conformation of the PrP molecule, allowing additional SAF32 to bind more readily.
By using synthetic peptides, we reported that the first five amino acids, KKRPK, at the N terminus of PrP c are essential for binding of GAG (15). However, other investigators using recombinant protein fragments have reported that there are additional GAG binding motifs; fragments containing residues 53-93 or 110 -128 also bind GAG (33,34). We found that when the first 12 amino acids of the N terminus are deleted, the recombinant protein is no longer able to bind GAG. Therefore, the N-terminal end that contains the KKRPK motif is the predominant GAG binding site on PrP c . The reason that rPrP 8OR and rPrP 10OR bind GAG better is because the N termini of these two proteins are more exposed and thus more available for GAG binding. We propose a model in which the octapeptide repeat region is the "neck" of PrP c . As the length of this region increases, the N-terminal "head" increasingly protracts from the C-terminal globular domain, rendering it more available for ligand binding. Whether binding of PrP c to GAG is important in prion disease is not known. PrP Sc particles in vivo contain GAG (35). In vitro, GAG facilitates the conversion of PrP c to PrP Sc (36). Cell surface GAG has also been reported to be the receptor for PrP Sc (37). GAG may function as a scaffold for concentrating PrP c , creating a reservoir of PrP c for conversion. Because rPrP 8OR and rPrP 10OR bind GAG better, they will be more prone to be concentrated in the scaffold. Our findings with recombinant prion proteins may have implications for pathogenesis. It has been reported that patients with more octapeptide repeats have earlier disease onset and shorter disease duration (9). Interestingly, by using the N terminus-specific mAb 8B4, we found that the amounts of full-length PrP species are greatly increased in the brain of patients as well as in mice with prion diseases (38). Presumably, these full-length PrP species are able to bind GAG.
The octapeptide repeat region of PrP c binds divalent cations such as Cu 2ϩ and Zn 2ϩ (39,40). However, whether such interactions are important in the pathogenesis of prion disease is not clear. We reported earlier that binding of Cu 2ϩ to native mouse rPrP c changes the binding profiles of two anti-PrP mAbs, 11G5 and 2C2, which react with epitopes that are downstream of the octapeptide repeat region (21). In this study, we provided new evidence that binding of Cu 2ϩ to rPrP c also alters the conformation of the N terminus, rendering the N terminus more available for antibody and GAG binding. However, these changes do not occur when Cu 2ϩ binds to rPrP 8OR . Although the physiologic significance of these findings is not clear, it is obvious that this modulation is non-operational in rPrP 8OR . The failure to modulate the conformation of the PrP molecule is not because rPrP 8OR binds more Cu 2ϩ than rPrP c . Irrespective of the number of octapeptide repeats, the maximum number of Cu 2ϩ -ions a PrP molecule binds is five . 3 Oxidative stress is caused by an imbalance in the levels of reactive oxygen species, which are the byproducts of normal cellular metabolism (41). Oxidative stress causes modification of amino acids and fragmentation and/or aggregation of proteins, eventually leading to cell death. The brain is very sensitive to reactive oxygen species because of the amount of oxygen it consumes. Oxidative stress has been speculated to play an important role in neurodegenerative diseases such as prion diseases and Alzheimer disease (42). Whether PrP c is directly involved in regulating reactive oxygen species remains controversial (43). However, it is clear that both rPrP 8OR and rPrP 10OR are more prone to oxidative damages. An increase in oxidative damage can be observed with 1 M H 2 O 2 , a level that is physiologically relevant (44).
Oxidative stress cross-links amino acids, resulting in aggregation of proteins (45). When incubated with 1 mM H 2 O 2 , rPrP 10OR but not rPrP 8OR begins to form aggregates of various sizes. Therefore, a minimum number of octapeptide repeats is required for aggregate formation under these conditions. The increase in the number of histidine residues within the octapeptide repeat region is the most likely explanation why rPrP 8OR and rPrP 10OR are more susceptible to oxidative attack.
The epitope of mAb 8B4 is no longer available for antibody binding when rPrP 8OR is incubated in 1 M H 2 O 2 , a level that is most likely unfeasible in vivo. Nevertheless, this phenomenon occurs only with rPrP 8OR and mAb 8B4 epitope. Identical treatment does not alter the binding of either mAb 7A12 or 8F9 to rPrP 8OR or the binding of mAb 8B4 to rPrP c . The reasons that mAb 8B4 fails to bind rPrP 8OR under this condition may be: the N terminus of rPrP 8OR is truncated; the amino acid residues within the mAb 8B4 epitope have been altered; or the mAb 8B4 epitope is buried because of conformational changes in other regions of the molecule. It is unlikely that the phenomenon is caused by N-terminal truncation because we did not observe a reduction in the molecular mass of rPrP 8OR after incubation with 1 M H 2 O 2 and immunoblotting with mAb 7A12 or 8F9. On the other hand, the mAb 8B4 epitope includes amino acid residues RYPGQGSPG (18). Other than cysteine and methionine, the amino acids that are most susceptible to oxidation include histidine, arginine, tyrosine, phenylalanine, and glutamine (46). We speculated that because the N terminus of rPrP 8OR is more exposed, it would be more susceptible to oxidative modification. However, we cannot rule out the possibility that the failure of mAb 8B4 to bind to rPrP 8OR is caused by conformational changes in the other regions of the molecule, resulting in sequestration of the epitope.
Based on these findings, we hypothesize that an increase in the number of octapeptide repeats renders the PrP molecule more susceptible to oxidative attack and that conformational changes at the N terminus enhance the binding of mutant PrP to GAG, which further promotes PrP aggregation. Because these aberrant features are proportional to the number of the insertions, our findings provide a biochemical explanation for the observation that patients with more octapeptide repeat insertions have earlier disease onset and shorter disease duration (9).