Introduction
Mammalian prions are proteinaceous pathogens responsible for a broad range of fatal neurodegenerative diseases in humans and animals [
20]. Prions are primarily formed of macromolecular assemblies of PrP
Sc, a misfolded, ß-sheet enriched form of the ubiquitously expressed, plasma membrane-anchored, variably N-glycosylated and α-helix rich, host-encoded prion protein PrP
C [
49]. This change is based on the self-sustained transfer of a structural information from the PrP
Sc conformer in the prion state to PrP
C, presumably through a seeding-polymerization process [
19]. Within defined host species, PrP
C can transconform in multiple prion variants or strains, differing in their PrP
Sc conformations at the level of the tertiary and/or quaternary structure, in their biological properties and in their relative capacity to replicate in cell lines or tissues [
5,
8,
14,
21,
37]. Prions can propagate within and between species, as exemplified by the emergence of human variant Creutzfeldt-Jakob disease (vCJD) through dietary exposure to prions responsible for the bovine spongiform encephalopathy (BSE) epidemics in cattle [
62]. Within defined species, prions can also form sporadically. In humans, the incidence of sporadic Creutzfeldt-Jakob disease (sCJD) ranges between 1 and 2 cases per million and per year, and affects mainly elderly people [
62]. Atypical BSEs (L-type and H-type) and atypical scrapie Nor98 are thought to develop spontaneously in aged ruminants [
8].
Characterizing PrP
Sc electrophoretic pattern following limited digestion with proteinase K (PK) can allow differentiation of prion strains. For sCJD, methionine/valine (M/V) polymorphism at codon 129 of the gene encoding PrP and the migration pattern and relative glycoform abundance of PK-resistant PrP
Sc (PrP
res) allow the definition of different molecular subtypes [
28,
45,
57]. These subtypes exhibit specific clinical and neuropathological features [
28,
45]. The most common form of sCJD is associated with the presence of type 1 (T1) PrP
res and homozygosity for methionine at codon 129. The unglycosylated band of T1 PrP
res migrates at 21 kDa in SDS-PAGE gels, and monoglycosylated forms predominate over diglycosylated ones. Rare forms of MM sporadic CJD with a Type 2 (T2) PrP
res type, - the unglycosylated form of which migrating at 19 kDa -, have been diagnosed. These forms are further subclassified as cortical and thalamic variants [
40,
45,
51]. The cortical variant is distinguished from all other sporadic forms by the absence of experimental transmission to knock-in mouse models expressing human PrP at physiological levels [
10,
30,
40].
Mice transgenic for PrP have been instrumental in deciphering prion strain diversity and in modeling experimentally the so-called species or transmission barrier that limits prions interspecies transmission (for review [
8]). In essence, such mice are generated to express specific sequences from mammalian PrP
C on a mouse PrP-ablated background, and are inoculated with prions. Analyzing the clinical outcome, attack rate and the presence of PrP
Sc in brain and peripheral tissues where prion replication can occur [
5,
27] allows the establishment of whether cross-interactions between host PrP
C and invading PrP
Sc structural landscapes are possible with regard to prion conversion. The conformational hypothesis [
21,
60] posits that PrP
C can adopt a limited portfolio of conformations in the PrP
Sc state, due to structural constraints in its amino acid backbone. If the infecting PrP
Sc conformation(s) is within the portfolio of possible conformations, cross-species transmission will occur. If not, the transmission barrier will be high, and can lead to an abrupt change in prion strain biological properties [
2,
4,
5,
9,
13,
29,
38,
46,
47,
54], a phenomenon referred to as a ‘mutation’. Whether the newly emerging strain is selected from an ensemble of pre-existing PrP
Sc conformations in the original inoculum (‘quasi-species’) or is generated
ex abrupto remains difficult to determine. A high transmission barrier does not lead systematically to prion strain ‘mutation’, as highlighted by the remarkable ability of classical BSE prions to retain their biological properties, despite intermediate passage to a range of different hosts [
12,
15,
35].
Here, we studied the strain biological properties of one rare subtype of sporadic CJD prions, the cortical MM2-form, upon transmission to either human or ovine PrP transgenic mice.
Discussion
We report here on the capacity of a rare subtype of human sCJD prions (MM2, cortical variant) to adapt onto the ovine PrP sequence, thereby modifying its apparent substrain composition and tissue tropism. We further show that the isolated variants can exhibit distinct replicative or biological behavior depending on the matrix or environment surrounding the prion conversion process.
Human PrP tg650 mice challenged with the MM2-cortical CJD subtype developed a clinical disease at full attack rate in less than 300 days. A T2 PrP
res profile was invariably observed in the inoculated mice. These characteristics were maintained over passaging, suggesting faithful propagation of this agent in this mouse line. Together, the incubation time in tg650 mice and the brain PrP
res profiles were unique among the panel of CJD cases transmitted so far to the tg650 line (Fig.
1 and unpublished data), thus fully supporting the contention that MM2-cortical CJD is a specific CJD subtype, as previously concluded from the absence of transmission to transgenic lines expressing physiological level of human PrP [
10,
30,
40]. Human PrP overexpression in tg650 mice may have been key in the success of transmission. Additionally, we used a cerebellum extract as inoculum, whereas previously, a cerebral cortex extract was used. Different brains and different brain regions harbor prion infectivity titers that can vary by 100-fold with regard to the number of infectious units per milligram of tissue [
1,
11]. While such a difference would barely affect the transmission rate of MM2-sCJD prions to tg650 mice (Table
2), this might be an issue in knock-in human PrP transgenic mice in which disease develops at slower pace [
10].
While MM2-sCJD prions propagated faithfully in human PrP mice, the outcome was different upon transmission to ovine PrP tg338 mice. Efficient transmission occurred on primary passage, at near full attack rate, albeit with a prolonged incubation period. A drastic reduction in the incubation time from 550 to 80 days occurred on secondary passage, consistent with the isolation of a variant, or a ‘mutant’ by analogy to mutational events observable with conventional microorganisms [
2,
4,
5,
8,
9,
13,
29,
46,
47,
54]. While a 19K-like PrP
res signature predominated in tg338 mouse brains as in human or human PrP mouse brain, a second signature, designated T1
Ov gradually emerged with subpassaging in tg338 mouse spleens. This distinct tissue tropism and the possibility to isolate, by conventional biological cloning or cell-free amplification, and further propagate separately and with fidelity the T2
Ov and T1
Ov types in tg338 mice or P2FJ6 cells indicate that a pair of
bona fide prion strains has been isolated on the ovine PrP sequence.
Once propagated alone, we failed to evidence any gradual shift of T2
OV towards T1
Ov prions in tg338 mice or in P2FJ6 cells, suggesting that the interspecies prion conversion events have been instrumental in the emergence of the T1
Ov substrain component. A recurrent and often perplexing question [
8,
21,
60,
61] is whether these agents have been generated
de novo on confrontation to the new PrP sequence or have been preferentially selected from pre-existing PrP
Sc conformations. Our data support the contention that T1
Ov prions have been generated
de novo: First, if the T1
Ov agent preexisted in the human MM2-sCJD source, it was not able to replicate at detectable levels in the spleens of tg650 mice, at variance with other human prion strain sources [
5,
6]. Second, we found no evidence of a T1 signature in the brain of tg650 mice inoculated with MM2-sCJD prions that may putatively be at the origin of T1
Ov prions onto the ovine PrP sequence (Fig.
1). Third, T1
Ov prions could be directly generated by PMCA from the human MM2-SCJD prions population passaged in tg650 mice. The lack of obvious difference in the PMCA efficacy between uncloned and cloned material (Fig.
4e), the latter being theoretically less populated in a potential T1
Ov subcomponent (to be compared with the dramatic reduction in the PMCA efficacy when tg338-passaged uncloned and cloned MM2-sCJD prions are amplified, Fig.
4d), suggests
de novo generation of T1
Ov on confrontation to the ovine PrP sequence. We acknowledge that definitive proof is lacking and it is possible that substrain heterogeneity may constantly arise even in cloned populations during propagation [
39].
Our observations are adding to the view that 2 distinct prion strains can propagate in neural and extraneural tissues of the same host individual [
5,
7]. We are currently comparing the retro-transmission properties of the T1
Ov and T2
Ov types to human PrP mice to examine whether the selection pressure imposed by the heterologous transmission was lower in spleen than brain tissues from tg338 mice, as previously observed [
5]. The situation may however be drastically different here, as the T1
Ov component emerged gradually in the spleen with subpassaging and adaptation, whereas previously, the spleen component emerged first, and at higher rate, than the brain component [
5].
The replicative properties of the two MM2-sCJD ovine prions markedly differed depending on the environment/matrix where ovine PrP conversion occurred: the two components fairly propagated in tg338 mice (yet at variable, tissue-dependent levels) and in P2FJ6 cells, whereas the T1
Ov component was selectively amplified by PMCA, despite the use of tg338 mouse brain as PrP
C substrate in the PMCA reaction. Comparative (ongoing) titration of isolated T1
Ov and T2
Ov prions suggest that both agents exhibit similar infectious titers (Table
4). Serial passaging in P2FJ6 cells also indicates that none of the component was outgrowing the other over passaging, suggesting roughly similar half-life and doubling time. Despite the point that our PMCA protocol is relatively promiscuous, allowing efficient amplification of several strains from different species [
41], it remains possible that T2
Ov amplification necessitates very specific PMCA conditions. Incidentally, PMCA appears as a convenient, alternative method to biological cloning, to isolate one prion substrain component from a mixture.
Table 4
Endpoint titration of isolated T1Ov and T2Ov MM2-sCJD prions in ovine PrP mice
| | T2Ov
b
| T1Ov
b
|
MM2-sCJD | 10−4
| 142 ± 8 (5/5) | nd |
| 10−5
| 146 (1/6) | 148 ± 4 (6/6) |
| 10−6
| 174; 192 (2/6)c
| 212 ± 6 (3/6)c
|
| 10−7
| > 450 (0/6)c
| > 450 (0/6)c
|
| 10−8
| >450 (0/6)c
| > 450 (0/6)c
|
Within the quasi-species concept applied to prions, -which proposes that prions are not constituting a single clone but are embedded with all PrP
Sc conformational variants [
21,
61]-, our observations would suggest that there has been a bottleneck event in tg338 mice (due to the heterologous PrP transmission) which has affected MM2-sCJD prions fitness, at least on primary passage, and has led to the emergence of two strain components. Key to the quasi-species definition is the existence of intra-population interactions, either complementation or interference [
44]. Complementation is difficult to accommodate with the observation that there is so far no apparent regeneration of the two T1
Ov and T2
Ov subcomponents after intermediate separation of one of the component, either by PMCA or biological cloning. As individuals, both components do not harbor different fitness with regard to incubation time, cell replicability and infectious titer; however, jointly, T2
Ov markedly outcompetes T1
Ov in either P2FJ6 cells or in tg338 brain. Neuroanatomically, the absence of replication of T1
Ov prions in certain brain target areas, that are free of T2
Ov prions (such as the lateral hypothalamic area (Fig.
3)) suggest local interfering mechanisms, independent of competition for the same PrP substrate [
55]. We thus propose that the MM2-sCJD ‘mutant’ isolated in tg338 mice is not a mere agglomeration of independently acting T1
Ov and T2
Ov conformations.
Since the discovery that RK13 cells expressing the ovine PrP VRQ allele were permissive to certain scrapie prions [
23,
58], we and others have made numerous attempts to infect these cells or more permissive clones such as the P2FJ6 one with ovine prions sources, passaged or not onto tg338 mice. Most attempts have failed, except for prions classified as ‘fast’, as based on their short incubation time in reporter tg338 mice (e.g. 127S and LA21K
fast, [
42,
56]). Here we show that two other relatively ‘fast’ agents originating from MM2-sCJD can fairly propagate in Rov cells at their maximum infectivity levels, as assessed by the tg338 bioassay. Many factors can account for the difficulties to replicate prions in immortalized cell lines. Obviously, prion-doubling time must surpass the cell division rate. For ‘fast’ strains such as 127S and LA21K
fast, it appears that a subset of small oligomers are by far the most active with respect to prion replication in vivo and to seeding activity by PMCA [
32]. This might be key to their sustained ability to replicate in cells. We are currently investigating the oligomeric state of the most active sCJD aggregates. In any case, this extended panel of strains replicating in cell culture opens the possibility to compare their biology in cells expressing wild type or mutated forms of PrP (Munoz-Montesino et al., submitted for publication).
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
JC, CC and MD carried out the cell studies, MM carried out the PMCA studies, FR, LH, EJ, HR, HL and VB carried out the mouse, biochemical and histopathological studies, IQ, APL and JB carried out the work related to the CJD patient, HR participated in the design of the study. VB designed the study and drafted the manuscript. All authors read and approved the final manuscript.