Biological and Biochemical Characterization of Mice Expressing Prion Protein Devoid of the Octapeptide Repeat Region after Infection with Prions

Accumulating lines of evidence indicate that the N-terminal domain of prion protein (PrP) is involved in prion susceptibility in mice. In this study, to investigate the role of the octapeptide repeat (OR) region alone in the N-terminal domain for the susceptibility and pathogenesis of prion disease, we intracerebrally inoculated RML scrapie prions into tg(PrPΔOR)/Prnp0/0 mice, which express mouse PrP missing only the OR region on the PrP-null background. Incubation times of these mice were not extended. Protease-resistant PrPΔOR, or PrPScΔOR, was easily detectable but lower in the brains of these mice, compared to that in control wild-type mice. Consistently, prion titers were slightly lower and astrogliosis was milder in their brains. However, in their spinal cords, PrPScΔOR and prion titers were abundant and astrogliosis was as strong as in control wild-type mice. These results indicate that the role of the OR region in prion susceptibility and pathogenesis of the disease is limited. We also found that the PrPScΔOR, including the pre-OR residues 23–50, was unusually protease-resistant, indicating that deletion of the OR region could cause structural changes to the pre-OR region upon prion infection, leading to formation of a protease-resistant structure for the pre-OR region.


Introduction
Transmissible spongiform encephalopathies or prion diseases, which include Creutzfeldt-Jakob disease in humans and scrapie and bovine spongiform encephalopathy in animals, are neurodegenerative disorders caused by prions [1,2]. Prions consist mainly of the abnormally folded, proteinase K (PK)-resistant isoform of prion protein, designated PrP Sc [3]. Structural conversion of the normal cellular isoform, designated PrP C , into PrP Sc is a key event in prion propagation. Indeed, mice devoid of PrP C (Prnp 0/0 ) are resistant to the disease without PrP Sc accumulation and prion propagation in the brain, even after inoculation with prions [4,5,6,7]. However, the exact conversion mechanism remains largely unknown.
PrP C is a glycoprotein tethered to the outer cell membrane via a glycosylphosphatidylinositol anchor moiety and expressed most abundantly in the brain, particularly by neurons [8]. Reduction in susceptibility to RML scrapie prions was reported in tg(PrPD32-93)/Prnp 0/0 and tg(PrPD23-88)/Prnp 0/0 mice, which express mouse (mo) PrP lacking residues 32-93 or 23-88 on the Prnp 0/ 0 background, respectively [9,10]. The incubation times of these mice were accordingly extended [9,10]. The incubation times of experimental prion diseases in mice are usually inversely correlated to the expression level of PrP C in the brain. Indeed, tg(moPrP)/Prnp 0/0 mice, which express mouse wild-type PrP C in the brains at 8 fold higher levels than control wild-type mice, showed a shorter incubation time of 5062 days post-inoculation (dpi) with RML prions, while the wild-type mice became sick at 12761 dpi [10,11]. Tg(PrPD23-88)/Prnp 0/0 mice were shown to express PrPD23-88 in their brains two fold higher than moPrP C in tg(moPrP)/Prnp 0/0 mice [10]. However, tg(PrPD23-88)/Prnp 0/ 0 mice developed the disease with a longer incubation time of 16164 dpi than tg(moPrP)/Prnp 0/0 mice with 5062 dpi [10]. Tg(PrPD32-93)/Prnp 0/0 mice also developed the disease with longer incubation times of 232 to 313 dpi than control wild-type mice with 158611 dpi, although tg(PrPD32-93)/Prnp 0/0 mice expressed PrPD32-93 in the brains 4 fold higher than PrP C in the control mice [9]. These results indicate that the N-terminal residues of PrP affect susceptibility to RML prions in mice. It was also reported that the MHM2(D23-88) molecule, a mousehamster chimeric PrP deletion mutant carrying hamster PrPderived methionine residues at 108 and 111 substituted for leucine and valine residues in mouse PrPD23-88, completely failed to restore susceptibility to RML prions in Prnp 0/0 mice [10,11]. These results indicate that the chimeric region, corresponding to residues 108 through 111, also influences the susceptibility to RML prions in mice.
The so-called octapeptide repeat (OR) region, which comprises 5 copies of an octapeptide sequence, is located in the unstructured N-terminal domain of PrP. PrPD32-93 lacks the entire OR region (residues 51-90) and most of the OR region is missing in PrPD23-88. It is thus suggested that the OR region might be involved in the susceptibility to RML prions in mice. However, PrPD32-93 and PrPD23-88 lack not only the OR region but also other regions. Therefore, it still remains unclear whether the decreased susceptibility in tg(PrPD32-93)/Prnp 0/0 and tg(PrPD23-88)/Prnp 0/ 0 mice could be due to the deletion of the OR region either alone or together with other regions.
We previously established a tg mouse line, designated tg(PrPDOR)/Prnp 0/0 , which expresses mouse PrP with a deletion of the OR region alone on the Prnp 0/0 background [12]. In the present study, to investigate the role of the OR region alone in prion susceptibility and the pathogenesis of prion disease, we intracerebrally inoculated RML prions into tg(PrPDOR)/Prnp 0/ 0 mice.

Ethics Statement
The Ethics Committee of Animal Care and Experimentation of University of Occupational and Environmental Health, Kitakyushu, Japan approved this study (approval number AE-080-13). Animals were cared for in accordance with The Guiding Principle for Animal Care and Experimentation of University of Occupational and Environmental Health and Japanese Law for Animal Welfare and Care. Animals C57BL/6 mice were purchased from CLEA Japan, Tokyo, Japan and ddY mice were from Kyudo, Tosu, Japan. ddY mice are outbred albino mice maintained in a closed colony. Tg(PrPDOR)/Prnp 0/0 mice with the C57BL/66129Sv6FVB mixed background were produced elsewhere [12]. In this study, tg(PrPDOR)/Prnp 0/0 mice (C57BL/66129Sv6FVB) were crossed at least more than twice with Zrch I Prnp 0/0 mice, which had been backcrossed to C57BL/6 mice more than 9 times.

Prion Inoculation
Brains were removed from terminally ill wild-type C57BL/6 mice infected with RML prions. A single brain was homogenized (10%, w/v) in phosphate-buffered saline (PBS) by passing it through 18 to 26 gauge needles and then diluted to 1% with PBS. Four to five week-old mice were intracerebrally inoculated with a 20 ml-aliquot of the homogenates.

Immunohistochemistry
Paraffin embedded samples were sectioned, deparaffinized, rehydrated and treated with L.A.B. Solution (Polysciences, Inc., U.S.A.) for 10 min. Nonspecific endogenous peroxidase activity was quenched by incubating the specimens with 3% H 2 O 2 for 10 min and then the specimens were blocked with 5% normal rabbit serum for 10 min at RT. For detection of PrP Sc or PrP Sc DOR, the specimens were treated with formic acid for 1 min before the blocking step. The specimens were then incubated with 1/500 polyclonal rabbit anti-GFAP antibodies (DAKO Cytomation, Denmark) or 1/100 polyclonal rabbit IBL-N anti-PrP antibodies (Immuno Biological Laboratories) for 2 h at RT. After washing in PBS, the specimens were incubated with HRP-labeled polymer anti-rabbit (EnVision TM System, DAKO Cytomation, Denmark) for 1 h at RT, washed in PBS, and then visualized using the avidin-biotin complex method (Vector Labs, U.S.A.). The nuclei were counterstained with Mayer's hematoxylin.

PNGase F Digestion
PNGase F digestion was performed according to the manufacturer's protocol (New England Biolabs, Inc., Ipswich, MA). Briefly, the PK-treated homogenates were denatured by boiling for 10 min in the presence of 0.5% SDS and 1% mercaptoethanol and then treated with PNGase F (500 units/L) in 1% Nonidet P-40 and 0.05 M sodium phosphate (pH 7.5) for 60 min at 37uC.

Standard Curve and Prion Titer Determination
To create a standard curve between prion titers and incubation times, 10% (w/v) brain homogenate of RML-infected ddY mice were serially diluted 10-fold with PBS, ranging from 10 21 to 10 210 in PBS, and a 20 ml-aliquot of each dilution was intracerebrally inoculated into ddY mice aged 4-5 weeks. The mice were observed until 1 year after inoculation. The ID 50 /gram of the tissue was determined according to the method of Reed and Muench and then a standard curve was created. Prions titers (ID 50 /g) in tissues of interest were determined as follows: A 20 mlaliquot of the tissue homogenates was intracerebrally inoculated into 5 or 6 ddY mice aged 4-5 weeks and their incubation times were determined. Thereafter, prion titers in the homogenates were calculated using the standard curve.

Expression Vectors
To construct an expression vector encoding mouse PrP tagged with the 3F4 epitope designated moPrP(3F4), the 59 fragment of mouse PrP cDNA was amplified by polymerase chain reaction (PCR) using a mouse PrP cDNA as a template with a BamHI-PrP(ATG)-S sense primer (Table S1) and a moPrP-3F4 anti-sense primer (Table S1). Then, full-length PrP cDNA was amplified by PCR using a mouse PrP cDNA as a template with the amplified 59 fragment as a sense primer and a PrP(stop)-XbaI-AS anti-sense primer (Table S1). After sequence confirmation, the amplified fragment was inserted into BamH I/Xba I-digested pcDNA3.1(+) (Invitrogen, Carlsbad, CA), resulting in pcDNA3.1-moPrP(3F4).

Statistical Analysis
Log-rank test was used for analysis of the incubation times of infected mice.

Incubation Times and Foreleg Paresis in tg(PrPDOR)/ Prnp 0/0 Mice after Infection with RML Prions
We intracerebrally inoculated RML prions into tg(PrPDOR)/ Prnp 0/0 mice and control C57BL/6 wild-type mice. Uninfected tg(PrPDOR)/Prnp 0/0 mice remained healthy for more than 500 days. Wild-type mice developed disease-specific symptoms, such as weight loss, decreased locomotive activity, ruffled hair coat and hunched back, at 16564 days post-inoculation (dpi) ( Table 1). Tg(PrPDOR)/Prnp 0/0 mice succumbed to the disease with slightly shorter incubation times of 14769 dpi (Table 1). This is probably due to higher expression of PrPDOR in the brains of tg(PrPDOR)/ Prnp 0/0 mice than in that of PrP C in wild-type mice. PrPDOR was detected in the brain and spinal cord about 2-3 fold more than PrP C in wild-type mice on Western blotting using SAF61 anti-PrP antibodies, which recognize residues 142-160 (human PrP numbering) (Fig. 1). Lack of the OR region in PrPDOR was confirmed by Western blotting using SAF32 anti-OR region antibody ( Fig. 1). Tg(PrPDOR)/Prnp 0/0 mice also displayed the additional unusual symptom of foreleg paresis at early stages of the disease.

Astrogliosis in tg(PrPDOR)/Prnp 0/0 Mice Infected with RML Prions
We investigated brain and cervical cord sections from terminally ill tg(PrPDOR)/Prnp 0/0 and wild-type mice for astrogliosis, a pathological hallmark of prion diseases, by immunohistochemical analysis using anti-GFAP antibodies. Astrogliosis was stronger in the brain and cervical cord sections from infected tg(PrPDOR)/ Prnp 0/0 and wild-type mice, compared to that in uninfected tg(PrPDOR)/Prnp 0/0 and wild-type mice (Fig. 2, A and B). However, brain astrogliosis in infected tg(PrPDOR)/Prnp 0/0 mice was slightly milder than in infected wild-type mice ( Fig. 2A). In contrast, in the cervical cord sections, astrogliosis was as strong in infected tg(PrPDOR)/Prnp 0/0 mice, as in infected wild-type mice (Fig. 2B). Western blotting showed consistent results. Compared to the GFAP expression in infected wild-type mice, it was mildly decreased in the brains of infected tg(PrPDOR)/Prnp 0/0 mice, but not in their spinal cords (Fig. 2, C and D). We also investigated the brain sections of terminally ill tg(PrPDOR)/Prnp 0/0 and wild-type mice for spongiosis. Vacuoles were similarly observed in the brains of both types of mice, scant in the cerebral cortex ( Fig. S1A) but common in the hippocampus (Fig. S1B) and cerebellum (Fig. S1C). The brain sections were further immunohistochemically stained for abnormal PrP isoforms, PrP Sc and PrP Sc DOR, with IBL-N anti-PrP antibodies, which were raised against the Nterminal residues 24-37, after treatment with formic acid. The immunoreactive signals were strong in the brains of both types of infected mice, compared to those in control uninfected mice, and were similarly distributed in the brains of both types of infected mice (Fig. S2).
PK-resistant PrP, or PrP Sc DOR, in tg(PrPDOR)/Prnp 0/0 Mice Infected with RML Prions We investigated the brains and spinal cords of terminally ill tg(PrPDOR)/Prnp 0/0 and wild-type mice for PK-resistant isoforms, PrP Sc DOR and wild-type PrP Sc , respectively, using Western blotting with SAF61 antibodies. PrP Sc DOR was easily detectable in the brains and spinal cords of two individual tg(PrPDOR)/ Prnp 0/0 mice (Fig. 3, A and B). However, compared to wild-type PrP Sc in infected wild-type mice, a reduced amount of PrP Sc DOR was detected in the brains of infected tg(PrPDOR)/Prnp 0/0 mice (Fig. 3A). Western blotting of the brains showed that total PrPs were more abundant in infected wild-type mice than in infected tg(PrPDOR)/Prnp 0/0 mice (Fig. 3A), despite PrPDOR being expressed in the brains of uninfected tg(PrPDOR)/Prnp 0/0 mice more than PrP C in uninfected wild-type mice (Fig. 1A). This is probably due to different amounts of wild-type PrP Sc and PrP Sc DOR accumulating in the brains. In the spinal cords, the amount of PrP Sc DOR in infected tg(PrPDOR)/Prnp 0/0 mice was similar to that of wild-type PrP Sc in infected wild-type mice (Fig. 3B).
Prion Propagation in tg(PrPDOR)/Prnp 0/0 Mice Infected with RML Prions We also determined prion titers (LD 50 /gram of tissue) in the brains and spinal cords of terminally ill tg(PrPDOR)/Prnp 0/0 and wild-type mice. To do this, we first created a standard curve between prion titers and incubation times by intracerebral inoculation of serially diluted brain homogenates of RML prionaffected mice into indicator mice. The mortalities and incubation times of the indicator mice are shown in Table 2. According to the method of Reed and Muench [14], prion titers of the homogenate were calculated as 10 8.5 LD 50 /g. The standard curve was given by Log 10 (LD 50 /g) = 14.0820.05X, where X is incubation time (days), 131,6,215. We thereafter intracerebrally inoculated the homogenates of 2 pooled brains and 2 pooled spinal cords from the terminally ill wild-type and tg(PrPDOR)/Prnp 0/0 mice into indicator mice. The brain and spinal cord used were from the same mouse. The inoculation of wild-type brain homogenate caused the disease in indicator mice at 11261 dpi, whereupon prion titers in the homogenate were calculated as .7.5 Log(LD 50 /g) (Table 3). However, after inoculation with tg(PrPDOR)/Prnp 0/0 brain homogenate, the indicator mice succumbed to the disease with significantly longer incubation times of 15068 dpi (Log-rank test, p = 0.0455), indicating that prion titers in the brains of terminally ill tg(PrPDOR)/Prnp 0/0 mice were 6.7 Log(LD 50 /g) ( Table 2). In contrast, in the spinal cords of infected tg(PrPDOR)/Prnp 0/0 mice, prion titers were not reduced ( Table 3). The spinal cord homogenates from terminally ill wild-type and tg(PrPDOR)/ Prnp 0/0 mice rendered the indicator mice ill at 15968 and 142613 dpi (Log-rank test, p = 0.3321), with prion titers in the homogenates being calculated as 6.3 and 7.0 Log(LD 50 /g), respectively ( Table 3).

The Pre-OR Region is Unusually PK-resistant in PrP Sc DOR
We recognized that the PK-resistant fragments of PrP Sc DOR in the brains and spinal cords appeared to migrate slightly slower than those of wild-type PrP Sc on Western blotting (Fig. 3,  A and B). This suggests that the PK-resistant core of PrP Sc DOR is higher in molecular weight than that of wild-type PrP Sc . To confirm this, we treated the PK-digested brain homogenates from terminally ill tg(PrPDOR)/Prnp 0/0 and wild-type mice with PNGase F before subjecting them to Western blotting. The molecular size of the deglycosylated PK-resistant fragment of PrP Sc DOR was clearly higher than that of wild-type PrP Sc (Fig. 4A). We also performed Western blotting of the brain homogenates with IBL-N antibodies raised against the N-  terminal residues 24-37 of PrP. The antibody reacted with the PK-resistant fragments from PrP Sc DOR but not from wild-type PrP Sc (Fig. 4B). We detected no PK-resistant fragments with molecular size .2 kDa in the brains of terminally ill wild-type mice on Western blotting using IBL-N antibodies (data not shown). Taken together, these results suggest that the entire PrP Sc DOR, including the pre-OR residues is PK-resistant, while only the C-terminal part is PK-resistant in wild-type PrP Sc .

Deletion of OR Residues 51-88 Renders the Pre-OR Residues PK-resistant in Prion-infected N2a Cells
Since PrPD32-93 in tg(PrPD32-93)/Prnp 0/0 mice lacks the entire OR region [9], we asked whether or not the remaining pre-OR residues could become PK-resistant upon conversion. In addition, since PrPD23-88 in tg(PrPD23-88)/Prnp 0/0 mice has 2 residues intact in the OR region [10], we also asked whether the 2 remaining OR residues in PrPD23-88 could potentially block the   pre-OR residues from becoming PK-resistant upon conversion.
We also replaced all of the lysine residues with positively charged arginine residues in moPrP(3F4) D32-88(3K3R) (Fig. 8A). This mutant protein was converted into moPrP(3F4) Sc D32-88(3K3R) in N2aC24L1-3 cells and the isoform gave rise to doublet non-glycosylated and mono-glycosylated bands with similar molecular size to those of moPrP(3F4) Sc D32-88 with the PK-resistant pre-OR residues (Fig. 8B). The upper band of the doublet was weakly detected by IBL-N antibodies (Fig. 8C). These results indicate that positive charges might play an important role for the pre-OR region of moPrP(3F4) D32-88 to become PKresistant in N2aC24L1-3 cells.

Discussion
Lines of evidence indicate that the N-terminal region of PrP is involved in the susceptibility of mice to prions. Tg(PrPD32-93)/ Prnp 0/0 and tg(PrPD23-88)/Prnp 0/0 mice, which lack the Nterminal residues 32-93 or 23-88, respectively, developed the disease with markedly elongated incubation times after infection with RML prions [9,10]. Moreover, Prnp 0/0 mice expressing PrP with further deletion in the N-terminal domain up to residue 106 from residue 32, or PrPD32-106, were free of the disease even after inoculation with RML prions [15]. In contrast, no extended incubation times were observed in tg(PrPD32-80)/Prnp 0/0 mice infected with RML prions [16]. PrPD32-93 and PrPD23-88 lack all or most of the OR region, respectively. However, PrPD32-80 still contains one intact octapeptide sequence in the OR region. This suggested that lack of the OR region from PrP could result in the decreased susceptibility to RML prions in the mice. However, in the present study, we observed no extended incubation times in tg(PrPDOR)/Prnp 0/0 mice, which express PrP lacking only the OR region, after infection with RML prions. The expression level of PrPDOR in the brain was lower than the reported level of PrPD32-93 or PrPD23-88 [9,10]. Taken together, these results indicate that, although deletion of the OR region alone from PrP barely affects the susceptibility to RML prions, a large deletion including the OR region in the N-terminal domain could result in remarkable reduction in the susceptibility of mice to RML prions.
We observed different pathogenesis between the brains and spinal cords of terminally ill tg(PrPDOR)/Prnp 0/0 mice. PrP Sc DOR and prion infectivity in the brains were lower than those in control wild-type mice. Astrogliosis in the brains was also milder than that in control wild-type mice. However, in the spinal cords, PrP Sc DOR, prion infectivity and astrogliosis were observed similarly to control wild-type mice. These results clearly indicate that, while the OR region is not essential for conversion; its deletion affects conversion taking place in the brain. Moreover, infected tg(PrPDOR)/Prnp 0/0 mice developed an unusual symp- Figure 4. The pre-OR region of PrP Sc DOR is PK-resistant. (A) The brain homogenates of terminally ill wild-type and tg(PrPDOR)/Prnp 0/ 0 mice were treated with PNGase F after digestion with PK, and subjected to immunoblotting with M-20 anti-PrP antibodies. The deglycosylated PK-resistant band of PrP Sc DOR was higher in molecular size than that of full-length PrP Sc . Arrows indicates PK-resistant deglycosylated PrPs. (B) The brain homogenates from terminally ill wild-type and tg(PrPDOR)/Prnp 0/0 mice were digested with PK, and subjected to immunoblotting with N-terminus-specific IBL-N anti-PrP antibody. The IBL-N antibodies recognized the PK-resistant PrPs from PrP Sc DOR but not from full-length PrP Sc . doi:10.1371/journal.pone.0043540.g004 tom of foreleg paresis, indicating that deletion of the OR region also modifies clinical signs. These unusual phenotypes were also reported in infected tg(PrPD32-93)/Prnp 0/0 mice. This indicates that lack of the OR region from PrP induces such unusual phenotypes in mice after infection with RML prions, as observed in infected tg(PrPDOR)/Prnp 0/0 and tg(PrPD32-93)/Prnp 0/0 mice. However, compared to the levels of PrP Sc DOR and prion infectivity in the brains of tg(PrPDOR)/Prnp 0/0 mice, the reported . The cell lysates were treated with PK at 5 mg/ml and then subjected to Western blotting. Both moPrP(3F4) and moPrP(3F4) D32-88 were converted to the PK-resistant isoforms, moPrP Sc (3F4) and moPrP Sc (3F4) D32-88, respectively. However, IBL-N anti-PrP antibody reacted only with the PK-resistant fragments of moPrP Sc (3F4) D32-88. doi:10.1371/journal.pone.0043540.g005 levels of PrP Sc D32-93 and prion infectivity are lower in tg(PrPD32-93)/Prnp 0/0 mice [9]. Astrogliosis was easily detectable in the brains of tg(PrPDOR)/Prnp 0/0 mice, but undetectable in tg(PrPD32-93)/Prnp 0/0 mice [9]. The foreleg paresis was developed at early stages in tg(PrPDOR)/Prnp 0/0 mice, but only at late stages in tg(PrPD32-93)/Prnp 0/0 mice [9]. These results The cell lysates were treated with PK at 5 mg/ml. All mutant proteins were converted into PK-resistant isoforms in N2aC24L1-3 cells. The PK treatment revealed doublet non-glycosylated and mono-glycosylated bands in moPrP(3F4) Sc D32-88 (arrows), indicating that the pre-OR region of some moPrP(3F4) Sc D32-88 molecules is PK-resistant. Similar doublet bands were observed in moPrP(3F4) Sc D32-88(2P2A) (arrows). However, moPrP(3F4) Sc D32-88(3K3A) gave rise to doublet bands with the upper band migrating very closely to the lower band (arrowheads). (C) Since substitution of proline residues into alanine residues disrupted the IBL-N epitope, the PK-resistant pre-OR residues in moPrP(3F4) Sc D32-88(2P2A) failed to be visualized by IBL-N anti-PrP antibodies. doi:10.1371/journal.pone.0043540.g006 suggest that the effects of deletion of the OR region alone on conversion are limited, compared to those of deletion of residues 32-93 including the OR region. Therefore, the levels of PKresistant PrP and prion infectivity were higher in the brains of tg(PrPDOR)/Prnp 0/0 mice than in tg(PrPD32-93)/Prnp 0/0 mice.
Consequently, astrogliosis was detectable in tg(PrPDOR)/Prnp 0/ 0 mice but not in tg(PrPD32-93)/Prnp 0/0 mice. The onset of foreleg paresis could be associated with the amounts of PrP Sc DOR or PrP Sc D32-93 in the brain or in the spinal cord. However, the exact mechanism underlying the foreleg paresis remains unknown. Figure 7. The pre-OR residues 23-31 with a substitution of the proline residues by tryptophan or glycine residues form a PKresistant structure in prion-infected N2a cells. (A) Amino acid sequences of the pre-OR residues 23-31 in mutant proteins. Bold residues indicate substituted residues. (B) Western blotting of N2aC24L1-3 cells transfected with control pcDNA3.1(+) and expression vectors encoding each mutant protein using 3F4 anti-PrP antibodies. The cell lysates were treated with PK at 5 mg/ml. All of the mutant proteins were converted into PKresistant isoforms in N2aC24L1-3 cells, and all of the mutant isoforms, moPrP(3F4) Sc D32-88, moPrP(3F4) Sc D32-88(2P2A), moPrP(3F4) Sc D32-88(2P2W) and moPrP(3F4) Sc D32-88(2P2G), gave rise to similar doublet non-glycosylated and mono-glycosylated bands. (C) Since substitution of proline residues into alanine, tryptophan or glycine residues disrupted the IBL-N epitope, the PK-resistant pre-OR residues in these mutant proteins failed to be visualized by IBL-N anti-PrP antibodies. doi:10.1371/journal.pone.0043540.g007 Upon the conversion of PrP C into PrP Sc , only the 2/3 Cterminal part of PrP C undergoes profound conformational changes to form the PK-resistant core of PrP Sc [1]. In contrast, the Nterminal part of PrP Sc remains PK-sensitive [1]. Consistent with this, we failed to detect any PK-resistant fragments with molecular size .2 kDa in the brains of terminally ill wild-type mice on Western blotting with IBL-N anti-N-terminus antibodies (data not shown). However, we found here that the entire PrP Sc DOR, including the pre-OR residues 23-50, appeared unusually PKresistant. No data were available whether the pre-OR region of PrP Sc D32-93 were PK-resistant [9]. However, it is very likely that the region could be PK-resistant in PrP Sc D32-93 because the OR region was completely deleted in PrPD32-93. Indeed, we found that the entire PrPD32-88, including the pre-OR residues 23-31, was converted to be PK-resistant in infected N2a cells. These results indicate that the pre-OR region has a potential to undergo conformational changes to become PK-resistant upon conversion, and that the OR region usually prevents the pre-OR region from undergoing such conformational changes. We also showed that the conversion activity of the pre-OR region in PrPD32-88 was diminished by substitution of either all of the positively charged lysine residues or of lysine residues 24 and 27 with uncharged alanine residues, but not affected by a substitution of all the lysine residues with positively charged arginine residues. These results suggest that the positive charge at 24 and 27 residues might be important for the pre-OR region to form a PK-resistant structure when the OR region is deleted. Deletion of the authentic PK cleavage site located within the OR region might be relevant to the unusual folding of the pre-OR region. Alternatively, length of the intervening sequence between the pre-OR region and the PKresistant C-terminal core might be a key factor to induce the unusual folding in the pre-OR region. It is also possible that, since the OR region binds to Cu 2+ via a histidine residue [17], loss of the binding activity to Cu 2+ might be responsible for formation of the PK-resistant pre-OR region.
The mechanism underlying the unusual phenotypes, such as brain-preferential reduction of the PrP conversion and foreleg paresis, in infected tg(PrPDOR)/Prnp 0/0 and tg(PrPD32-93)/ Prnp 0/0 mice remains unknown. RML prions are a strain mixture. It is thus possible that a certain specific prion strain(s) might be selected in tg(PrPDOR)/Prnp 0/0 or tg(PrPD32-93)/Prnp 0/0 mice because of lack of the OR region, causing the unusual phenotypes in these mice after infection with RML prions. However, inoculation of the brain or spinal cord homogenates from terminally ill tg(PrPDOR)/Prnp 0/0 mice did not induce such unusual phenotypes in wild-type mice (data not shown). Moreover, tg(PrPD23-88)/Prnp 0/0 mice were reported to show no such unusual phenotypes despite PrPD23-88 lacking most of the OR region [10]. These results suggest the unlikelihood of this possibility. Another possibility is that the overexpression of PrPDOR or PrPD32-93 in the spinal cords might cause high accumulation of PrP Sc DOR in the spinal cords, resulting in development of the unusual phenotypes in these infected mice. However, PrP Sc DOR was accumulated in the spinal cords of infected tg(PrPDOR)/Prnp 0/0 mice at a similar level to that of wildtype PrP Sc in infected wild-type mice. Moreover, no foreleg paresis was reported in infected tg(PrPD23-88)/Prnp 0/0 and tg(moPrP)/ Prnp 0/0 mice [10]. Tg(PrPD23-88) and tg(moPrP) mice were generated using the same cos.SHaTet expression vector system as in tg(PrPDOR) mice [10,12], indicating that, similarly to PrPDOR in tg(PrPDOR)/Prnp 0/0 mice, PrPD23-88 and moPrP C are overexpressed in the brains and spinal cords of these mice. Therefore, this possibility is also unlikely. Alternatively, the unusual phenotypes in infected tg(PrPDOR)/Prnp 0/0 or tg(PrPD32-93)/Prnp 0/0 mice might be due to indirect effects caused by deletion of the OR region, but not due to direct effects of deletion of the OR region. PrPDOR and PrPD32-93 include the pre-OR residues 23-31 intact, whereas PrPD23-88 does not, suggesting that the pre-OR residues in PrP Sc DOR and PrP Sc D32-93 might be associated with the unusual phenotypes in tg(PrPDOR)/Prnp 0/0 and tg(PrPD32-93)/Prnp 0/0 mice. Indeed, we showed here that the pre-OR residues of PrP Sc DOR or possibly PrP Sc D32-93 were unusually PK-resistant. The pre-OR region binds to glycosaminoglycans, the polysaccharide chains of proteoglycans, via the positively charged lysine-rich region [18], and the binding of PrP to glycosaminoglycans is important for conversion [19]. Therefore, the structurally changed pre-OR region in PrP Sc DOR and PrP Sc D32-93 might alter the binding affinity to a yet unidentified proteoglycan(s) important for conversion in the brain, resulting in disturbance of conversion in the brain. The structurally changed pre-OR region also might induce a new neurotoxic signal, causing foreleg paresis as in infected tg(PrPDOR)/Prnp 0/0 or tg(PrPD32-93)/Prnp 0/0 mice. However, further studies are required to elucidate the mechanism of the unusual phenotypes in infected tg(PrPDOR)/Prnp 0/0 or tg(PrPD32-93)/Prnp 0/0 mice.
There are many different groups of prion strains with strainspecific pathogenic properties [17,20]. It is postulated that the prion strain-specific properties are enciphered in the strain-specific conformation of PrP Sc [21,22]. Since changes in the protein conformation would cause changes in PK-accessibility, PrP Sc s with different conformations would have different PK cleavage sites. Indeed, it was reported that two different prion strains of transmissible mink encephalopathy, HY and DY strains, produced PrP Sc s with different PK cleavage sites [23]. HY strain produced PrP Sc with a longer PK-resistant fragment than that of PrP Sc produced by DY strain [24]. We showed here that the entire PrP Sc DOR, including the pre-OR residues 23-50, appeared PKresistant, while only the C-terminal part is PK-resistant in wildtype PrP Sc , suggesting that PrP Sc DOR and PrP Sc might form different conformations. It might be thus interesting to characterize the biological properties of prions associated with PrP Sc DOR. Figure 8. Positively charged lysine residues, particularly located at codons 24 and 27, are importnat for the pre-OR residues 23-31 to form a PK-resistant structure in prion-infected N2a cells. (A) Amino acid sequences of the pre-OR residues 23-31 in mutant proteins. Bold residues indicate substituted residues. (B) Western blotting of N2aC24L1-3 cells transfected with control pcDNA3.1(+) and expression vectors encoding each mutant protein using 3F4 anti-PrP antibodies. The cell lysates were treated with PK at 5 mg/ml. All of the mutant proteins were converted into PK-resistant isoforms in N2aC24L1-3 cells, and all of the mutant isoforms gave rise to doublet non-glycosylated and monoglycosylated bands. The doublet bands of moPrP(3F4) Sc D32-88(K23A), moPrP(3F4) Sc D32-88(K24A) and moPrP(3F4) Sc D32-88(K27A) were similar in molecular size to those of moPrP(3F4) Sc D32-88. However, moPrP(3F4) Sc D32-88(K24,27A) gave rise to the doublet band with the upper band migrating very closely to the lower band, similarly to moPrP(3F4) Sc D32-88(3K3A). MoPrP(3F4) Sc D32-88(K23,24A) and moPrP(3F4) Sc D32-88(K23,27A) showed the upper band with an intermediate molecular size. MoPrP(3F4) Sc D32-88(3K3R) giving rise to doublet bands with similar molecular size to those of moPrP(3F4) Sc D32-88. (C) IBL-N antibodies recognized all of the PK-resistant isoforms except for moPrP(3F4) Sc D32-88(3K3A) and moPrP(3F4) Sc D32-88(K24,27A). doi:10.1371/journal.pone.0043540.g008

Supporting Information
Figure S1 Similar spongiform change in the brains of infected wild-type and tg(PrPDOR)/Prnp 0/0 mice. The brains of uninfected or terminally ill wild-type and tg(PrPDOR)/ Prnp 0/0 mice were subjected to HE staining. Vacuoles were scant in the cerebral cortex (A) but common in the hippocampus (B), and cerebellum (C). No specific vacuoles were observed in the brains of uninfected mice. (TIF) Figure S2 Similar distribution of PrP Sc and PrP Sc DOR in the brains of infected wild-type and tg(PrPDOR)/ Prnp 0/0 mice. The brains of uninfected or terminally ill wildtype and tg(PrPDOR)/Prnp 0/0 mice were subjected to immunohistochemistry with IBL-N anti-PrP antibodies after treatment with formic acid. The immunoreactive signals were similarly observed in the brains of both types of infected mice, but not in control uninfected mice. (A), cerebral cortex; (B), hippocampus; (C), cerebellum. (TIF)