Vaccination with Prion Peptide-Displaying Polyomavirus-Like Particles Prolongs Incubation Time in Scrapie-Infected Mice

Prion diseases like scrapie in sheep, bovine spongiform encephalopathy (BSE) in cattle or Creutzfeldt–Jakob disease (CJD) in humans are fatal neurodegenerative diseases characterized by the conformational conversion of the normal, mainly α-helical cellular prion protein (PrPC) into the abnormal β-sheet rich infectious isoform PrPSc. Various therapeutic or prophylactic approaches have been conducted, but no approved therapeutic treatment is available so far. Immunisation against prions is hampered by the self-tolerance to PrPC in mammalian species. One strategy to avoid this tolerance is presenting PrP variants in virus-like particles (VLPs). Therefore, we vaccinated C57/BL6 mice with nine prion peptide variants presented by hamster polyomavirus capsid protein VP1/VP2-derived VLPs. Mice were subsequently challenged intraperitoneally with the murine RML prion strain. Importantly, one group exhibited significantly increased mean survival time of 240 days post-inoculation compared with 202 days of the control group. These data show that immunisation with VLPs presenting PrP peptides may represent a promising strategy for an effective vaccination against transmissible spongiform encephalitis agents.


Introduction
Transmissible spongiform encephalopathies (TSE) are fatal neurodegenerative disorders that are caused by proteinaceous infectious particles (prions). So-called prion diseases include Creutzfeldt-Jakob Disease (CJD) and Kuru in humans, bovine spongiform encephalopathy (BSE) in cattle, scrapie in sheep and goats, and chronic wasting disease (CWD) in cervids [1].
Prion diseases are caused by the misfolding of the normal, host-encoded cellular prion protein (PrP C ) into a pathogenic isoform (PrP Sc ). Cellular PrP C is linked to the cell membrane by a glycophosphatidyl inositol (GPI) anchor and contains two N-linked carbohydrate groups, and a single disulfide bond [1,2]. PrP C is found in all mammals, and most highly expressed in the central nervous system, but also in a vast majority of immune cells like lymphocytes, natural killer cells and monocytes. Moreover follicular dendritic cells (FDC) are essential sites of prion replication in lymphoid tissues [3]. The exact biological function of PrP C is still unknown but there are indications for a role in signal transduction [4], immunoregulation [5], synaptic plasticity through promoting neurite outgrowth [6], Cu 2+ binding [7,8] and modulating N-methyl-D-aspartate receptor (NMDAR) activity [9]. The potential neuroprotective function of PrP C is consequently lost during its conversion to PrP Sc disturbing synaptic functionality and leading finally to synapse degeneration [10].

Generation, Purification and Electron Microscopy Analysis of Chimeric and Pseudotype Virus-Like Particles (VLPs)
The chimeric VP1 and VP2 proteins of HaPyV were produced in yeast S. cerevisiae strain AH22 -214 (a, leu2 his4). Purification of recombinant proteins was performed as described previously [35,40]. Briefly, yeast cells transformed with constructed plasmids (Table 1) were first cultured in glucose-, then galactose-containing induction media for 24 and 18 h, respectively and then collected by centrifugation, washed with distillate water, and stored at −20 • C until purification. For purification of chimeric proteins, yeast cells after suspension in DB450 buffer (450 mM NaCl, 1 mM CaCl 2 , 0.001% 0.25 M L-arginine, Triton X-100, 10 mM Tris/HCl, pH 7.2) with 2 mM phenylmethylsulfonylfluorid (PMSF) and EDTA-free Complete Protease Inhibitor Cocktail tablets (Roche Diagnostics, Mannheim, Germany) were disrupted with glass beads using Bead-Beater GB26 (BioSpec Products, Inc., Bartlesville, OK, USA). The chimeric proteins were purified from the supernatant of yeast cell lysate by ultracentrifugation overnight at 4 • C in 20-60% sucrose gradient (100,000× g) and then for 18 h in cesium chloride (CsCl) gradient (1.23-1.38 g/mL). After pooling and diluting, purified recombinant protein-containing fractions in DB150 buffer were precipitated by ultracentrifugation for 4 h at 100,000× g. Pellets containing recombinant chimeric proteins were dissolved in phosphate-buffered saline (PBS, pH 7.2), dialyzed against PBS, aliquoted (25 and 50 µg), lyophilised and stored until use.
To confirm the VLP assembly of purified recombinant proteins, samples of each construct were placed on 300-mesh carbon-coated palladium grid. Adhered to the grid, samples were negatively stained with 2% aqueous uranyl acetate solution and examined by JEM-100S electron microscope (JEOL, Tokyo, Japan).
Accumulated PrP Sc was examined by a western blotting protocol using phosphotungstic acid (PTA) precipitation [41,42]. In short, 10% (wt/vol) brain homogenates of all samples were homogenized in 0.42 mM sucrose solution containing detergents (0.5% deoxycholic acid sodium salt/0.5% Nonidet P-40) and incubated with proteinase K (PK, final concentration: 50 µg/mL) at 55 • C for 60 min. Subsequently the digestion was stopped by addition of Pefabloc (Roche, Mannheim, Germany) and heating at 95 • C for 5 min. Subsequently, PTA was added to the samples (final concentration, 0.3%), incubated at 37 • C for 60 min and centrifuged at 13,300 rpm for 30 min at RT. Samples were resuspended in loading buffer, heated for 5 min at 95 • C and loaded on 16% Tris-PAGE. Proteins were transferred onto PVDF membrane (Millipore) and the membranes were blocked for 1 h in PBST-milk. PrP Sc was detected using mAb SAF70 (Bertin Technologies, Montigny le Bretonneux, France) by incubating the membranes for 1 h at RT. After washing the membranes 3 times with PBST a secondary antibody-alkaline-phosphatase-conjugated anti-mouse immunoglobulin (Dianova, hamburg, Germany)-was added in PBST for 1 h at RT. The membranes were finally washed 3 times in PBST and bound antibodies were detected with the chemiluminescence substrate CDP Star (Thermo Fisher) and signals were visualized and calculated with Versa Doc system (Quantity One, Bio-Rad, Munich, Germany) and by the analysis software Quantity One (BioRad, Munich, Germany) using background substraction method.

Propagation of Scrapie Strain and Immunisation Scheme
Ten groups of C57/Bl6 mice (5 individuals each) were subcutaneously (s.c.) immunized in a prime-boost regimen three times at intervals of three weeks (priming, first and second boost) with 50 µg of the corresponding chimeric or pseudotype VLPs or VP1/VP2derived VLPs (control). Immunized mice were subsequently challenged with 50 µL of a 1% brain homogenate of mouse scrapie strain RML intraperitoneally (i.p.). The schematic course of application is depicted in Figure 1. The used mouse scrapie strain RML was originally designated the "Chandler" strain and was the result of a passage of natural sheep scrapie brain to mice [43]. The health status of the mice was analysed daily, and body weight was recorded weekly. After the onset of TSE-associated clinical symptoms including weight loss, abnormal tail tonus, hind limb paralysis, the animals were euthanized. The incubation times were calculated as the time between inoculation and necropsy. The brains were removed and stored at −20 • C. Mice that died from other causes than scrapie were excluded from the dataset. Animal experiments were approved by German law (reference number: LALLF M-V/TSD/7221.3-1.1-039/05).

Enzyme-Linked Immunosorbent Assay (ELISA)
All serum samples were tested in an in-house indirect IgG enzyme-linked immunosorbent assay (ELISA) based on full length recombinant murine PrP and HaPyV VP1 protein. Maxisorb immunoplates (Nunc, Roskilde, Denmark) were coated with 1 μg of the corresponding protein, diluted in 0.05 M carbonate-bicarbonate buffer (pH 9.6). After incubating the plates overnight at 4 °C, they were washed three times with 300 μL washing buffer PBST, followed by a blocking step with 200 μL/well 10% skim milk powder diluted in PBS for 1 h at 37 °C. Serum (5 μL diluted in 120 μL 2% skimmed milk) was added, plates were incubated at 37 °C for 1 h and subsequently washed. 100 μL of HRP-conjugated protein G (Calbiochem), diluted 1:5000 in dilution buffer, were then added to each well. After another 1 h incubation step, plates were washed again and 100 μL substrate (2′-Azinobis [3-ethylbenzothiazoline-6-sulfonic acid]-diammonium salt (ABTS); Roche, Mannheim, Germany) was added to each well. Reactions were run for 30 min at RT in the dark, stopped by addition of 1% SDS and the optical density (OD) value at 405 nm was determined. Relative OD was calculated as OD from samples at corresponding timepoints t0, t21 and t42 subtracted from t0 value and divided by t0 [(ODtx−OD t0)/ODt0.]

Statistical Analysis
Survival times were analyzed by Kaplan-Meier survival analysis using the log-rank test to compare the curves. The statistical analysis was done using SigmaBlot statistical software (Systat Software, San Jose, CA, USA). Survival times are expressed as means ±

Enzyme-Linked Immunosorbent Assay (ELISA)
All serum samples were tested in an in-house indirect IgG enzyme-linked immunosorbent assay (ELISA) based on full length recombinant murine PrP and HaPyV VP1 protein.
Maxisorb immunoplates (Nunc, Roskilde, Denmark) were coated with 1 µg of the corresponding protein, diluted in 0.05 M carbonate-bicarbonate buffer (pH 9.6). After incubating the plates overnight at 4 • C, they were washed three times with 300 µL washing buffer PBST, followed by a blocking step with 200 µL/well 10% skim milk powder diluted in PBS for 1 h at 37 • C. Serum (5 µL diluted in 120 µL 2% skimmed milk) was added, plates were incubated at 37 • C for 1 h and subsequently washed. 100 µL of HRP-conjugated protein G (Calbiochem), diluted 1:5000 in dilution buffer, were then added to each well. After another 1 h incubation step, plates were washed again and 100 µL substrate (2 -Azinobis [3-ethylbenzothiazoline-6-sulfonic acid]-diammonium salt (ABTS); Roche, Mannheim, Germany) was added to each well. Reactions were run for 30 min at RT in the dark, stopped by addition of 1% SDS and the optical density (OD) value at 405 nm was determined. Relative OD was calculated as OD from samples at corresponding timepoints t 0 , t 21

Statistical Analysis
Survival times were analyzed by Kaplan-Meier survival analysis using the log-rank test to compare the curves. The statistical analysis was done using SigmaBlot statistical software (Systat Software, San Jose, CA, USA). Survival times are expressed as means ± standard deviation.

Biochemical and Electron-Microscopical Characterization of VLPs
Partial sequences of MPrP were inserted into major capsid protein VP1 or fused at N-terminus of a N-terminally truncated capsid protein VP2 of HaPyV. This included MPrP amino acid residues 128-164 (P1), 174-219 (P2), 128-219 (P3), 51-128 (P4) as well as the full length mature MPrP sequence (residues 23-231, MPrP). The structure of the nine different VP1 or VP2 fusion proteins is schematically illustrated in Figure 2a. All constructs assembled into polyomavirus-derived VLPs are as shown by electron microscopy (Figure 2b). The constructs were detected by Coomassie blue staining (Figure 3a) as well as in Western blot assay using PrP specific as well as VP1/VP2 specific antibodies (Figure 3b Table 1). The constructs were detected by Coomassie blue staining (Figure 3a) as well as in Western blot assay using PrP specific as well as VP1/VP2 specific antibodies (Figure 3b-f). All VLPs displayed immunoreactivity against VP1-specific mAb 6D11 (Figure 3b) and all pseudotype VP1/VP2 VLPs (K3, K5, K8) in addition against polyclonal anti-VP2 antiserum (Figure 3c). Detection of MPrP peptides was undertaken with mAb SAF70 (Figure 3d 109-115, 145-180, respectively. The detected protein bands corresponded to the predicted molecular mass of constructed chimeric proteins (see Table 1).

Serological Analysis of VLP-Immunized Mice
To assess, whether antibodies were induced by the immunisation procedure, sera at time point 0 (priming), day 21 (first boost) and day 42 (second boost) were collected and analysed by an indirect ELISA using homologous constructs as antigens (Figure 4). The individual data are shown in Table S1. There was no correlation between individual ELISA values and incubation times of corresponding mice (Table S2).

Serological Analysis of VLP-Immunized Mice
To assess, whether antibodies were induced by the immunisation procedure, sera at time point 0 (priming), day 21 (first boost) and day 42 (second boost) were collected and analysed by an indirect ELISA using homologous constructs as antigens (Figure 4). The individual data are shown in Table S1. There was no correlation between individual ELISA values and incubation times of corresponding mice (Table S2).
Five constructs (K2, K4, K5, K8 and K9) including the VP1/2 control developed a significant increase in the anti-VP1 antibody response against the applied VLP construct. Four constructs (K1, K3, K6 and K7) did not provoke a significant induction of anti-VP1 antibody response. No PrP specific antibodies were detected in ELISAs on days 21 and 42, when full length recombinant MPrP was used as immunogen). To reveal low antibody levels mouse sera taken at necropsies were pooled (100 µL) (from each group of 5 animals) and used in Western blot tests (final dilution 1:20). A specific signal was obtained only for pooled sera from the K9 immunized animals. No signals were seen for any other immunogen and none for the VP1/2 control (Figure 5a). The semi-quantitative analysis is shown in Figure 5b. The corresponding Western blot test is depicted in Figure S1.  Five constructs (K2, K4, K5, K8 and K9) including the VP1/2 control developed a significant increase in the anti-VP1 antibody response against the applied VLP construct. Four constructs (K1, K3, K6 and K7) did not provoke a significant induction of anti-VP1 antibody response. No PrP specific antibodies were detected in ELISAs on days 21 and 42, when full length recombinant MPrP was used as immunogen). To reveal low antibody levels mouse sera taken at necropsies were pooled (100 μL) (from each group of 5 animals) and used in Western blot tests (final dilution 1:20). A specific signal was obtained only for pooled sera from the K9 immunized animals. No signals were seen for any other immunogen and none for the VP1/2 control (Figure 5a). The semi-quantitative analysis is shown in Figure 5b. The corresponding Western blot test is depicted in Figure S1.

In Vivo Efficacy of VLP Immunisation against Scrapie Infection
The protective potential of all VLP constructs was measured by determination of the mean survival time after challenge ( Figure 6). Untreated mice had a mean incubation time of 201.6 +/− 4.5 days. Mice treated with the VP1/2 control alone exhibited an incubation time of 213.8 +/− 8.7 days. A slight increase was observed in mice groups treated with

In Vivo Efficacy of VLP Immunisation against Scrapie Infection
The protective potential of all VLP constructs was measured by determination of the mean survival time after challenge ( Figure 6) Table S2. The PrP Sc accumulation in brains of mice immunized with construct K9 and untreated terminally sick mice was analyzed by Western blot test (Figure 7). The PrP Sc accumulation in brains of mice immunized with construct K9 and untreated terminally sick mice was analyzed by Western blot test (Figure 7). mean survival times ± the standard deviation for mice immunized with VLP constructs K1-K9, VP1/2 control and nonimmunized control (Co) mice (a). Columns are labeled with elongation time [days] compared to untreated control. Kaplan-Meier survival analysis of intraperitoneally (i.p.) infected mice after s.c. treatment with construct K9. Controls included untreated controls (Co) and VP1/2 controls (b).
The PrP Sc accumulation in brains of mice immunized with construct K9 and untreated terminally sick mice was analyzed by Western blot test (Figure 7).  Western blot analysis of PK resistant PrP Sc fragments demonstrated no differences in the glycosylation pattern nor molecular weight of PrP Sc between VLP (K9) immunized and control mice. The analysis demonstrates further that the amount of PrP Sc accumulation was similar in control compared to immunized mice. Similar findings were observed for mice immunized with the other constructs ( Figure S2).

Discussion
The results demonstrate that HaPyV capsid protein-derived VLPs containing PrP fragments represent a promising prophylactic strategy for direct immunisation. In total, nine VLP variants displaying different PrP fragments were tested and one construct (K9)-encompassing PrP sequence 128-164-significantly elongated the survival time. Mice immunized with this construct K9 exhibited a humoral immune response and generated auto-antibodies that detect PrP as determined by western blot assay.
Interestingly, construct K1 chimeric VLPs harboring the same peptide P1 (amino acids 128-164) also showed, although it was not significant, a prolongation of the incubation period (+34.6 versus +38.8 for K9). The different potential of K9 and K1 chimeric VLPs might be explained by the presence/absence of the GSSG linker or the different insertion sites (site #1 in K9 versus site #4 in K1). The lacking or much lower effect of pseudotypebased VLPs might be explained by the insertion of other peptides; the lower PrP epitope density on the VLP surface might be an additional reason to be tested in further experiments with P1 peptide fused to VP2 protein of HaPyV. Moreover, a slight, but non-significant increase in the incubation time (12 days) was generally observed only for pure VP1/VP2 derived VLPs. VLPs have been shown to trigger a cellular immune response including the release of pro-inflammatory cytokines and a cytotoxic T-lymphocyte (CTL) response [44]. This stimulation activates macrophages to phagocytosis leading to a partial degradation within the phagolysosomal compartments of the cells [45].
In a previous study, direct intracerebral inoculation of mouse scrapie strain RML lead to an incubation time of 164 days in C57BL/6 mice [46]. In contrast intraperitoneal inoculation of this strain in our study prolonged incubation time by 38 days due to initial dissemination of PrP Sc via peripheral nerves and lymphoreticular organs to the CNS followed by brain neuroinvasion [47]. Application of construct K9 exhibits a similar delay of 38 days (240 vs. 202 days) and underlines the significant immunisation effect of these chimeric VLPs.
The use of VLPs comprising PrP sequences to elicit an immune response against PrP have been evaluated in just a few studies: in one approach a peptide of nine amino acids of the murine PrP (DWEDRYYRE; amino acids 144-152) was inserted within the L1 major capsid protein of bovine papillomavirus type 1 (BPV-1). The assembled VLPs induced antibodies in rabbits, that inhibited the synthesis of PrP res in prion-infected ScN2a Viruses 2021, 13, 811 11 of 14 neuroblastoma cells [48]. In an alternative setting murine leukemia virus (MLV) was used to create VLPs displaying murine PrP that induced anti-PrP C specific antibodies even in wild-type mice [49]. Interestingly, this was only observed for constructs encoding the C-terminal prion fragment (amino acids 121 to 231). However, in both cases, no additional studies have been performed so far to further evaluate these findings in prophylactic and therapeutic settings.
The elongation of the incubation time of about 18% (240 vs 202 days) observed here is in a similar range compared to other immunisation strategies: immunisation with murine PrP linked to Keyhole Limpet Hemocyanin (KLH) prior to infection prolonged incubation time by 11.2% for scrapie in mice [50] and scrapie infected mice that were vaccinated with aggregated murine PrP (suspended in complete Freund's adjuvant) survived approximately 28 days longer than naive mice [51]. Immunisation with KLH-linked hamster PrP peptides resulted in a prolongation of about 13.0-18.7% in scrapie infected hamsters [52] and the immunisation of hamsters with dendritic cells loaded with PrP peptides (encompassing amino acids 98-127) delayed the onset of clinical scrapie by 18.7% [53].
Another promising approach to overcome PrP self-tolerance was the immunisation with recombinant dimeric murine PrP with different adjuvants which induced in vitro efficacious anti-PrP antibodies [54,55]. A similar approach with monomeric and dimeric PrP induced auto-antibodies against cervid PrP and significantly prolonged incubation times in CWD-infected transgenic elk PrP (TgElk) mice [56,57]. Interestingly, immunisation with heterologous monomeric mouse PrP caused the largest prolongation by 60% in TgElk mice [57].
The exact mechanism that causes the significant delay remains to be elucidated, but the generated antibodies could interfere with the binding and/or conversion into the infectious PrP Sc isoform, since the expressed PrP sequence in construct K9 encompasses a specific region of the globular domain (amino acids 128-164). This includes an α-helix located at amino acid positions 143 to 153 and two short β-sheets (residues 127-130 and 160-163). The first helix plays a significant role in binding cellular PrP C to infectious PrP Sc and the two β-sheets are initial point of the conformational transition into a β-solenoid architecture and subsequent PrP Sc amyloid fibril formation [58]. Antibodies targeting this region have been demonstrated to have a significant anti-PrP effect, e.g., mAb 6H4 (raised against amino acids 144-152) which clears PrP Sc from prion infected neuroblastoma cells [59]. Moreover, passive immunisation with mAb ICSM18 (targeting amino acid residues 146-159) lead to a prolongation of the incubation period by more than 500 days [23]. Therefore, the VLP-induced antibodies could act as inhibitors of PrP conversion. On a molecular level, antibodies could interfere directly with host encoded PrP C to avoid PrP Sc binding and to prevent subsequent conversion into pathogenic isoform due to lack of substrate. Alternatively antibodies could directly bind to cellular PrP C and trigger a decrease of PrP C levels in host cells which are then not available as a source for further conversion and subsequent PrP Sc replication. This mechanism has been proposed for mAb targeting PrP residues 144-152, which causes clearance of cellular PrP in prion-infected cell cultures [60]. However, the proposed inhibitory mechanisms are only transient, as terminal sick mice, either VLP treated or not, develop a similar PrP Sc aggregate pattern and do not effectively cure the disease.

Conclusions
Taken together, the obtained results demonstrate that a homologous chimeric VLPbased PrP vaccine can abrogate host encoded immune tolerance to cellular PrP C and induces an albeit weak antibody response that prolongates survival time in scrapie infected mice. Further studies are needed to compare the chimeric to the pseudotype PrP VLP vaccine and determine the potential of a PrP-VLP vaccine for prophylactic or even therapeutic interventions in currently fatal and incurable prion diseases.

Supplementary Materials:
The following are available online at https://www.mdpi.com/article/10 .3390/v13050811/s1, Table S1: Optical density (OD) of indirect VP1 IgG ELISA displaying raw data from mice sera after immunisation with corresponding VLP constructs at three different timepoints, Table S2: Survival time (individual values) of RML-infected C57/Bl6 mice immunized with virus-like particles, Figure S1: Western blot analysis Western blot analysis of pooled sera from immunized animals, Figure S2: PrP sc accumulation in the brains of selected mice challenged with RML.