Two distinct conformers of PrPD type 1 of sporadic Creutzfeldt–Jakob disease with codon 129VV genotype faithfully propagate in vivo

Current classifications of sporadic Creutzfeldt–Jakob disease (sCJD) identify five subtypes associated with different disease phenotypes. Most of these histopathological phenotypes (histotypes) co-distribute with distinct pairings of methionine (M)/valine (V) genotypes at codon 129 of the prion protein (PrP) gene and the type (1 or 2) of the disease-associated PrP (PrPD). Types 1 and 2 are defined by the molecular mass (~ 21 kDa and ~ 19 kDa, respectively) of the unglycosylated isoform of the proteinase K-resistant PrPD (resPrPD). We recently reported that the sCJDVV1 subtype (129VV homozygosity paired with PrPD type 1, T1) shows an electrophoretic profile where the resPrPD unglycosylated isoform is characterized by either one of two single bands of ~ 20 kDa (T120) and ~ 21 kDa (T121), or a doublet of ~ 21–20 kDa (T121−20). We also showed that T120 and T121 in sCJDVV have different conformational features but are associated with indistinguishable histotypes. The presence of three distinct molecular profiles of T1 is unique and raises the issue as to whether T120 and T121 represent distinct prion strains. To answer this question, brain homogenates from sCJDVV cases harboring each of the three resPrPD profiles, were inoculated to transgenic (Tg) mice expressing the human PrP-129M or PrP-129V genotypes. We found that T120 and T121 were faithfully replicated in Tg129V mice. Electrophoretic profile and incubation period of mice challenged with T121−20 resembled those of mice inoculated with T121 and T120, respectively. As in sCJDVV1, Tg129V mice challenged with T121 and T120 generated virtually undistinguishable histotypes. In Tg129M mice, T121 was not replicated while T120 and T121−20 generated a ~ 21–20 kDa doublet after lengthier incubation periods. On second passage, Tg129M mice incubation periods and regional PrP accumulation significantly differed in T120 and T121−20 challenged mice. Combined, these data indicate that T121 and T120 resPrPD represent distinct human prion strains associated with partially overlapping histotypes.


Introduction
For several years sporadic Creutzfeldt-Jakob disease (sCJD) has been grouped into five distinct subtypes, denoted as sCJDMM(MV)1, -MM2, -MV2, -VV1 and -VV2 [18,38]. This grouping is based on the combination of two major molecular determinants of the disease phenotype: the methionine (M)/valine (V) polymorphic genotype at codon 129 of the prion protein (PrP), which dictates the MM, MV and VV 129 genotypes, and the type 1 or 2 of the disease-associated PrP (PrP D ) [7,8,31,38]. PrP D types 1 (T1) and 2 (T2) are distinguished by their respective ~ 21 kDa and ~ 19 kDa electrophoretic mobilities following treatment with proteinase K (PK), which are commonly monitored (for practical reasons) with the unglycosylated isoform (lower band) of the PK-resistant PrP D (resPrP D ) [17]. The distinct mobility reflects the different sizes of the PK-resistant region and, therefore, the distinct conformations of the T1 and T2 PrP D isoforms [31,36,40].
In this study, which is part of a body of research on sCJD subtypes, we focus on sCJDVV1, the least investigated subtype, especially with regard to the characteristics of its PrP D [11,37].
Sporadic CJDVV1 is also the rarest of the five subtypes, accounting for 2-3% of sCJD [11,39,46]; it presents at a younger age on average, with clinical onset often in the 3th or 4th decade of life, and has a relatively long course that often exceeds one year [11,18,41]. Phenotypically, sCJDVV1 is easily distinguishable from the other sCJD subtypes by the type and distribution of the histological lesions (histotype), which include severe spongiform degeneration (SD) with medium size vacuoles throughout the cerebral cortex, presence of ballooned neurons and widespread but light PrP deposition [11]. The electrophoretic profile of sCJDVV1 resPrP D T1 is complex, as shown by the heterogeneity of the unglycosylated isoform. We recently identified three alternative electrophoretic profiles or variants of T1: the T1 20 and T1 21 variants, where the two resPrP D fragments of ~ 21 and ~ 20 kDa occur separately, and the T1 21−20 variant where the two resPrP D fragments coexist in different ratios [11]. We also observed that T1 21 and T1 20 have distinct conformational characteristics suggesting that they represent distinct strains. Nonetheless, the histotypes associated with the T1 20 , T1 21 and T1 21−20 variants are similar violating the tenet that distinct prion strains are associated with distinct phenotypes [6,17,42].
To further investigate this issue, transgenic (Tg) mice expressing normal or cellular human PrP (PrP C ) with the codon 129 residue V (Tg129V) or M (Tg129M), were inoculated with sCJDVV1 brain isolates containing T1 20 , T1 21 or T1 21−20 * (the last isolate was obtained from a sCJDVV1-2 case harboring tiny amounts of T2, denoted by asterisk). Brain extracts from sCJDVV2, a different sCJD subtype that harbors resPrP D T2 (with a ~ 19 kDa unglycosylated fragment), were inoculated as controls. Both T1 20 and T1 21 were faithfully replicated in Tg129V mice with T1 21 showing a longer incubation period, whereas T1 21−20 * was reproduced as T1 21 . Replication was longer and less faithful in the Tg129M mice: the ~ 21-20 kDa resPrP D doublet was generated following inoculations with T1 20 , and T1 21 was not transmitted. The histotype in T1 20 and T1 21−20 *-inoculated Tg129M mice was characterized by the overlapping lesion profiles and the lack, in T1 21−20 *-inoculated mice only, of PrP deposits in cerebral cortex and cerebellum. Second passage in Tg129M mice recapitulated the results of the first passage except for the significantly shorter and different incubation periods of T1 20 and T1 21−20 *-infected mice.
The transmission in Tg129V mice of both T1 21 and T1 20 with the accurate replication of their electrophoretic profiles, along with the lack of replication of T1 21 only following serial transmissions in Tg129M mice suggest that T1 21 and T1 20 are distinct prion strains even though they are associated with similar histotypes in sCJDVV1.

Features of resPrP D of the inocula
The inocula containing resPrP D T1 20 were obtained from three cases of sCJDVV1 while inocula harboring T1 21 and T2 were each isolated from one case of sCJDVV1 and sCJDVV2, respectively; T2 was used as control. T1 21−20 * corresponds to T1 variant with a ~ 21-20 kDa doublet co-existing with T2 (the latter accounting for ~ 5% of the total resPrP D ) harvested from a case of sCJDVV1-2 that had histotype mostly consistent with sCJDVV1 (Additional file 5: Table S1). Immunoblotting characterization of the inocula confirmed the previously established electrophoretic profiles of T1 20 and T1 21 resPrP D variants and excluded the presence of T2 (Additional file 1: Figure S1). The consistent predominance of the ~ 21 kDa component in the T1 21−20 * variant was also confirmed. Of note, the small T2 component of T1 21−20 * was detected with the T2-specific Ab Tohoku-2 (data not shown), but not with the type generic 3F4 (Additional file 1: Figure S1). T2 in sCJDVV2 was harvested from the frontal cortex and putamen, which were used as separate inocula.

Transgenic mice
Two Tg mouse lines, the Tg362 and Tg340, were used [34,35]. They express the human PrP C -129V (Tg362) and PrP C -129M (Tg340) at ~ eightfold and ~ fourfold normal human brain levels, respectively, and are hereafter identified as Tg129V and Tg129M.

Intracerebral inoculations
Twenty microliters of ten percent (wt/vol) brain homogenates (BH) in 5% glucose generated from the sCJDVV cases were inoculated intracerebrally according to previously described procedures [35]. A total of 99 Tg mice were inoculated in this study. Brain homogenates from Tg129M mice challenged with each of the three T1 variants were used for a second passage in the same mouse line.

Histology, immunohistochemistry, lesion profiles, and morphometric analysis
Histological and immunohistochemical examinations were carried out on four brain levels at approximately bregma 0.5 mm, − 1.7 mm, − 3.8 mm and − 6.0 mm, as previously described [10]. Paraffin sections were stained with hematoxylin and eosin (H.E.) or probed with the Ab 3F4 [22,47] to human PrP (residues 106-110) at 1:1,000 and 1:400 dilutions as previously described [10]. Lesion profiles were performed using semi-quantitative evaluation for severity of SD, which was rated on a 0-3 scale on H.E.-stained sections (0 = not detectable; 1 = mild, 2 = moderate, and 3 = severe) [14]. Each point of the lesion profiles and bar graphs in Figs. 2 and Additional file 3: S3 were expressed as mean ± standard error of the mean (SEM). The eight brain regions examined included the cerebral cortex, hippocampus, basal ganglia, thalamus, hypothalamus, superior and inferior brainstem, and cerebellum. The semi-quantitative assessment of gliosis severity and neuronal loss in the cerebellum was rated on a 0-3 scale as noted above. Morphometric analysis to assess vacuole-size was carried out on the cerebral cortex at the level of bregma − 1.7 mm, and measured by the software Image-Pro Plus (Media Cybernetics, Inc.) [23].

Preparation of brain homogenates, PK digestion and Western blot analysis
Ten percent (wt/vol) BH of human cases were prepared using 1X LB100 (100 mM NaCl, 0.5% Nonidet P-40, 0.5% sodium deoxycholate, 10 mM EDTA, 100 mM Tris-HCl, pH 8.0), and supernatants (S1) were collected following centrifugation at 1000 × g for 5 min (min) at 4 °C. For the mouse brains, 10% BH prepared in 5% glucose was mixed with an equal volume of 2X LB100 (pH 8.0) and centrifuged at 1000 x g for 5 min. Human and mouse S1 were subjected to enzymatic digestion with 10 Units/ml (U/ ml) PK (Sigma Aldrich), which was used at 48 U/mg PK specific activity (1 U/ml is equal to 20.8 μg/ml PK) at 37 °C for 1 h (h). Enzymatic reaction was stopped by the addition of 2 mM phenylmethylsulfonyl fluoride (PMSF) prior to the dilution of each sample with an equal volume of 2 × Laemmli buffer (6% SDS, 20% glycerol, 4 mM EDTA, 5% β -mercaptoethanol, 125 mM Tris-HCl, pH 6.8) and then denaturation at 100 °C for 10 min.

Transmission features and characterization of resPrP D variants in Tg mice
In Tg129V mice, all three T1 variants and T2 transmitted with 100% attack rate. Incubation periods or days post inoculation (dpi) varied; it was the longest for T1 21 (425 ± 66 dpi) and the shortest for T1 21−20* (315 ± 66 dpi) even though the difference did not rich statistical significance (Tables 1 and Additional file 5: S2). T2 transmitted in 215 ± 18 dpi, which was significantly different (P < 0.0001) from the incubations of all T1 variants combined. T1 20 and T1 21 electrophoretic mobility and Ab immunoreactivity were indistinguishable from those of the respective inocula (Fig. 1a, c, d). By contrast, the T1 21−20 * inoculum (with predominance of the T1 21 component and presence of ~ 5% T2) was reproduced only as T1 21 , with the addition of a weak band of ~ 19 kDa that was detectable with the T2-specific Ab Tohoku-2 but not with the type generic 3F4, mirroring the T2 of the inoculum (Table 1 and Fig. 1a). T2 was faithfully replicated as the typical ~ 19 kDa resPrP D T2 (Fig. 1e). Ancillary transmission studies with hemizygous Tg129V mice challenged with T1 20 and T2 isolates led to results similar to those obtained with homozygous mice with the exception of longer incubation periods (data not shown).  Tg129M mice showed a 100% attack rate following challenge with T1 20 , T1 21−20 * and T2, whereas T1 21 failed to replicate. T1 20 and T1 21−20 * essentially shared the incubation periods (554 ± 53 dpi and 570 ± 60 dpi, respectively), which, on average, were 1.7 times longer than those of Tg129V (Tables 1 and Additional file 5: S2). The incubation period following inoculation with T2 was nearly three times longer than Tg129V (Tables 1 and Additional file 5: S2). Overall, resPrP D replication was much less accurate: T1 20 replicated as T1 21−20 * with ~ 20 kDa preponderance and no detectable T2, whereas T1 21−20 * inoculation engendered both ~ 21 and ~ 20 kDa fragments but with an inverted ratio as compared to that of the inoculum, and with traces of T2 (Fig. 1b, d). Furthermore, T2 was replicated as T1 20 , which immunoreacted with the resPrP D type non-specific 3F4 Ab but not with the resPrP D T2-specific Tohoku-2 Ab, clearly indicating that the resPrP D T2 of the inoculum was not replicated (Tables 1 and Additional file 5: S2, Fig. 1e). This finding contrasts with the apparently faithful replication of the ~ 19 kDa T2 by the Tg129V mice inoculated with sCJDVV2, and it is puzzling considering that bona fide T2 ~ 19 kDa fragment is reproduced by Tg129M mice after inoculation with T1 21−20 * (Tables 1 and Additional file 5: S2, Fig. 1e). The second passage in Tg129M mice as the first, resulted in the replication of T1 20 and T1 21−20 * only with indistinguishable electrophoretic profiles (Table 1 and Additional file 2: Figure S2). However, the incubation periods were respectively reduced ~ 1.6and ~ 1.9-fold due to the strain adaptation (P < 0.0001). Furthermore, the ~ 50 days longer incubation period of mice inoculated with T1 20 also was statistical significant (T1 20 : 338 ± 30 dpi; T1 21−20 *: 292 ± 16 dpi; P < 0.009) (Tables 1 and Additional file 5: S2).

Histopathological and immunohistochemical features of inoculated Tg mice
Inoculations of resPrP D T1 20 , T1 21 and T1 21−20 * variants to Tg129V mice generated similar histopathological features ( Table 2, Figs. 2, 3) consisting of prominent spongiform degeneration (SD) and astrogliosis of neocortex, hippocampus and basal ganglia, which progressively subsided caudally (except for a small peak in the brain stem) reaching the lowest level in the cerebellum. Vacuoles commonly were of medium or intermediate size, and plaques were not detected (Fig. 2a, b).
Matching PrP immunostaining (IHC) showed some topographic variation. In the cerebral cortex the pattern was similar in the three T1 variants and consisted of individual granules or clusters of variable sizes that codistributed with SD (Table 2 and Fig. 3a). However, while in T1 21 and T1 21−20 *, the granular deposits were limited to the cerebral neocortex and hippocampus, T1 20 inoculated Tg mice displayed PrP granules also in subcortical regions. Furthermore, the cerebellum showed PrP deposition in the granule cell layer in T1 20 and T1 21−20 * but it was entirely negative in T1 21 Tg mice ( Table 2, Fig. 3b).
In Tg129M mice inoculated with the T1 20 or T1 21−20 * variants, SD severity and brain regional distribution or lesion profile, did not significantly differ from those of Tg129V mice (Fig. 2a, d, and Additional file 3: Figure  S3 A and B) although vacuoles were significantly larger (P < 0.0001) (Fig. 2a, e). Mice challenged with T1 21 were free of lesions up to ~ 700 dpi (Table 1 and Fig. 2). Lesion profiles and vacuole size in second passage Tg129M mice challenged with T1 20 and T1 21−20 * overlapped with those of the first passage (data not shown).
In T1 20 Tg129M, PrP IHC pattern with granular aggregates in the cerebral and cerebellar cortices mirrored that of matching Tg129V, while T1 21−20 * Tg mice showed rare granular aggregates in subcortical regions but not in the cerebral cortex and cerebellum (Fig. 3a, b). No plaques were detected (Table 2 and Fig. 3b).
Following T2 inoculation, both Tg129V and Tg129M mice showed scant SD that, contrary to T1 variants, displayed an inversed severity gradient that increased progressively from the cerebral cortex, where it was virtually iii-iv, VV1 21 : Mouse brain free of lesions. iii-iv, VV2: Cortical plaques; inset, iii: high magnification of a plaque. b and c: Profiles of brain distribution and severity of SD were similar in Tg129V mice challenged with VV1 20 , VV1 21 , and VV1 21−20 * (b), and in Tg mice challenged with VV2 (c). d: Profiles in Tg129M mice inoculated with VV1 20 and VV1 21−20 * were similar; VV1 21 -inoculated mice were free of lesions. e: Vacuole size averaged from nine VV1 20 and three VV1 21−20 * challenged mice was ~ 8 µm 2 greater in Tg129M than Tg129V. *P < 0.05, **P < 0.003, ***P < 0.0001. CC: Cerebral cortex, HI: hippocampus, BG: basal ganglia, TH: thalamus, Hth: hypothalamus, BS.s: brainstem, superior, BS.i: brainstem, inferior, CE: cerebellum absent, to the thalamus ( Table 2, Fig. 2a, c). Furthermore, SD was made of small vacuoles. In the cerebellum, astrogliosis also was significantly more severe than that observed in mice inoculated with T1 variants (Additional file 3: Figure S3 C and E) although granule cell depopulation did not reach statistical significance (Additional file 3: Figure S3D and E). PrP IHC showed plaques-like aggregates in the cerebral cortex of the Tg129V mice while real plaques were seen only in the brain stem and septal nuclei in one mouse (Figs. 3 and Additional file 4: S4). By contrast, plaques were widespread in Tg129M mice and populated the cerebral cortex, thalamus, the border between the hippocampal alveus and the corpus callosum, the brain stem and cerebellum in the majority (70%) of the inoculated mice (Figs. 3 and Additional file 4: S4).

Discussion
Previous transmission studies to Tg mice expressing human wild-type or mutated PrP did not examine the mouse replications of the sCJDVV1 T1 variants that we have recently described [4,12,13,16,21,24,26,34,45]. We now show that T1 20 and T1 21 are faithfully reproduced in Tg129V mice with no significantly different incubation periods and slightly different histotypes reminiscent of that associated with the -VV1 subtype (Tables 1, 2). Transmissibility characteristics clearly  distinguished T1 20 from T1 21 following inoculation to  the Tg129M mice where T1 20 accumulated as T1 21−20 *  whereas T1 21 was not detected. T1 20 and T1 21−20 * transmission to Tg129M required an incubation period nearly 60% longer than that of the Tg129V mice consistent with the effect of the 129 genotype barrier. This assumption is further supported by the significant reduction in the incubation period following second passage in Tg129M with T1 20 and T1 21−20 * (Tables 1 and Additional file 5: S2). A similar phenomenon has been observed following second passage of sCJDVV1 prions to Tg129M mice [12].
In contrast to the accurate reproduction of the T1 20 and T1 21 variants, T1 21−20 * inoculated to Tg129V mice accumulated as T1 21 . Conversely, Tg129M mice faithfully accumulated T1 21−20 *; both mouse lines accumulated T1 21−20 * with the additional presence of T2 traces (also present in the inoculum) which may have impacted the replication. The two T1 21−20 * variants generated in Tg129M following inoculation of T1 20 and T1 21−20 *, respectively, had significantly different incubation period on second passage and differed in the histotype based on the lack of cerebral cortical and cerebellar pathology in the latter. Furthermore, second passage in Tg129M mice confirmed the lack of transmission of T1 21 . An unexpected phenotypical distinction between T1 inoculated Tg129V and− 129M mice was the size of the vacuoles, which was significantly larger in the Tg129M mice consistent with an effect of the PrP 129MV polymorphism on this distinctive histopathological feature. Vacuole size and lesion profiles were virtually identically in Tg129M mice of the 1st and 2nd passage.
Transmission of sCJDVV2 T2 used as control revealed expected results. In contrast to the faithful replication of -VV2 T2 by the Tg129V mice, a T1 20 variant was reproduced in the Tg129M after an incubation period that was three times longer than that in Tg129V. Our data resemble those recently described in a transmission study employing the same Tg129M mouse line as in our study [12]. These findings confirm the incompetence of human PrP C -129M to reproduce -VV2 T2 [12,24] as opposed to the faithful transmission of -MM2 to Tg129M mice [30,34].
The original classification of major sCJD subtypes based on histotype and PrP D characteristics has undergone recent revisions [2,11,12,27,32,43]. Sporadic CJDMM(MV)1 (a combination of -MM1 and -MV1, which share histotype and PrP D characteristics) as well as -MM2 (also referred to as MM2C) and -VV2, are seen as definitely distinct subtypes [5,18,19,32]. They are associated with PrP D variants that show distinct conformational and transmissible characteristics but have straightforward electrophoretic profiles of either PrP D type 1 or 2. By contrast, the -MV2 and -VV1 subtypes have shown considerable electrophoretic heterogeneity [11,32,33]. The subtype -MV2 is now subdivided into two variants; the first, -MV2C, is currently viewed as a phenocopy of -MM2 in terms of histotype and PrP D characteristics; the second, -MV2K, is characterized by the presence of kuru (K) plaques and heterogeneous PrP D inclusive of at least two components: (i) a ~ 19 kDa PrP D variant with gel mobility and conformational features similar to the -VV2 ~ 19 kDa, and (ii) a ~ 20 kDa PrP D (also termed "intermediate" type or "type i") of uncertain origin. Recently, however, the convergence of transmission and mass spectrometry data basically indicates that (i) the -MV2C and -MV2K phenotypes and respective PrP D characteristics are directly related to the representation of the resPrP D -129M and -129V components, respectively [32], (ii) the -MV2K ~ 20 kDa variant is made exclusively of the minority resPrP D -129 M component, and (iii) the -MV2K ~ 20 kDa appears to be an adaptation of the VV2 PrP D type 2 to the 129MM or 129MV background ( [24,25,32] and this study).
Our previous study showed that in sCJDVV1 resPrP D presents an even higher level of complexity given that it features three combinations of resPrP D kDa: T1 20 , T1 21 and T1 21−20 [11]. The T1 20 and T1 21 difference of ~ 1 kDa in electrophoretic mobility of the two resPrP D variants, although minor, is not negligible since it implies that the span of the PK-resistant region (i.e., the abnormal secondary structure generated during the PrP C to PrP D conversion) is different in the T1 20 and T1 21 variants. Indeed, T1 20 and T1 21 isolated from sCJDVV1 brains show features (e.g., resistance to enzymatic degradation by PK and propensity to unfold following exposure to the denaturing agent guanidine hydrochloride) that differ significantly, which further supports the conclusion that these two T1 variants have distinct conformational characteristics even though they are associated with similar histotypes [11]. It is noteworthy that the association of conformationally distinct prions strains with similar phenotypes has been previously reported [1,44]. Our present findings are consistent with this conclusion given that both T1 20 and T1 21 can be faithfully replicated in Tg129V mice but display opposite transmission characteristic in Tg129M mice; furthermore, mimicking sCJDVV1, T1 20 and T1 21 are associated with essentially similar histotypes in the Tg mice.
Our study also offers the opportunity to directly compare the histotype of the T1 20 variant associated with -VV1 with that of the T1 20 variant generated after inoculation of -MV2K and -VV2 PrP D to Tg129M mice [24,25]. Tg129M mice inoculated with T1 20 from -VV1 subjects are characterized by medium size vacuole SD, predominantly impacting the cerebral cortex, and lack of PrP plaques. By contrast, Tg129M inoculated with -MV2K and -VV2 isolates (also reported to harbor a T1 20 variant) [24] displayed ubiquitous plaques along with small vacuole SD occupying mostly subcortical regions. These two distinct histotypes are thus reminiscent of the -VV1 and -MV2K/-VV2 subtypes, respectively. The nature of the molecular features-besides the M and V incongruity at PrP residue 129-underpinning the complex and major impact on the histotype associated with the two T1 20 variants, remains to be resolved.

Conclusions
The present study further contributes to understand the molecular features of T1 variants in sCJDVV1 [11]. Our present data along with the previous conformational studies are consistent with the conclusion that T1 21 and T1 20 resPrP D are two distinct human prion strains that generate similar clinico-histopathological phenotypes. The lack of transmissibility of T1 21 VV1 to Tg129M mice suggests that subjects with the PrP-129MM genotype may not be at risk of acquiring prion disease from sCJDVV1 donor harboring the T1 21 variant. Understanding the molecular properties of PrP D T1 associated with sCJDVV1 may shed light into the common early presentation of this subtype and be essential for strain-sensitive therapeutic approaches [3,15,20,29].