A single protective polymorphism in the prion protein blocks cross-species prion replication in cultured cells

Abstract

The bank vole (BV) prion protein (PrP) can function as a universal acceptor of prions. However, the molecular details of BVPrP’s promiscuity for replicating a diverse range of prion strains remain obscure. To develop a cultured cell paradigm capable of interrogating the unique properties of BVPrP, we generated monoclonal lines of CAD5 cells lacking endogenous PrP but stably expressing either hamster (Ha), mouse (Mo), or BVPrP (M109 or I109 polymorphic variants) and then challenged them with various strains of mouse or hamster prions. Cells expressing BVPrP were susceptible to both mouse and hamster prions, whereas cells expressing MoPrP or HaPrP could only be infected with species-matched prions. Propagation of mouse and hamster prions in cells expressing BVPrP resulted in strain adaptation in several instances, as evidenced by alterations in conformational stability, glycosylation, susceptibility to anti-prion small molecules, and the inability of BVPrP-adapted mouse prion strains to infect cells expressing MoPrP. Interestingly, cells expressing BVPrP containing the G127V prion gene variant, identified in individuals resistant to kuru, were unable to become infected with prions. Moreover, the G127V polymorphic variant impeded the spontaneous aggregation of recombinant BVPrP. These results demonstrate that BVPrP can facilitate cross-species prion replication in cultured cells and that a single amino acid change can override the prion-permissive nature of BVPrP. This cellular paradigm will be useful for dissecting the molecular features of BVPrP that allow it to function as a universal prion acceptor.

1 |. INTRODUCTION

Prions are infectious protein aggregates that cause a variety of fatal neurodegenerative conditions in humans and animals, collectively referred to as prion diseases (Colby & Prusiner, 2011; Scheckel & Aguzzi, 2018; Watts et al., 2006). The naturally occurring prion diseases include scrapie in sheep, chronic wasting disease in cervids, as well as Creutzfeldt–Jakob disease, Gerstmann–Sträussler–Scheinker disease, fatal familial insomnia, and kuru in humans. The infectious proteins underlying these diseases are composed of an endogenously expressed protein called the cellular prion protein (PrP^C) (Oesch et al., 1985; Prusiner, 1982; Stahl et al., 1987). During disease, PrP^C undergoes a conformational rearrangement from a predominantly α-helical structure to an aggregation-prone conformer that is enriched in β-sheet content termed PrP^Sc (Hallinan et al., 2022; Kraus et al., 2021; Manka et al., 2022; Pan et al., 1993; Riek et al., 1997). PrP^C is post-translationally modified by the addition of up to two N-linked glycans as well as a glycosylphosphatidylinositol (GPI) anchor that tethers the protein to the extracellular face of the plasma membrane. Compared to PrP^C, PrP^Sc exhibits distinct biochemical features, including insolubility in non-ionic detergents and partial resistance to proteolytic digestion (Prusiner, 1982; Prusiner et al., 1977). PrP^Sc can template the conversion of PrP^C into additional copies of PrP^Sc, enabling prion replication and cell-to-cell spreading both within and between tissues. Templated misfolding is a central feature of prion diseases, and can be recapitulated through animal bioassays (Watts & Prusiner, 2014), prion infections in cultured cells (Krance et al., 2020), or in cell-free conversion assays like Protein Misfolding Cyclic Amplification or Real-Time Quaking-Induced Conversion (RT-QuIC) (Saborio et al., 2001; Wilham et al., 2010). PrP^Sc can also exist as distinct strains, which are believed to be encoded by structural variation within PrP^Sc aggregates (Block & Bartz, 2022).

Animal bioassays are costly, both in terms of time and resources. Therefore, cultured immortalized cell models of prion replication have served as an alternative paradigm for characterizing the biology associated with prion infection (Heumuller et al., 2022; Krance et al., 2020). Murine N2a neuroblastoma cells were the first cells shown to be capable of becoming persistently infected with the RML strain of mouse-adapted scrapie prions (Butler et al., 1988; Race et al., 1987). These cells have been extensively used to assess the consequences of prion infection in a cellular context and have also served as a platform for identifying and characterizing small molecule anti-prion compounds (Krance et al., 2020). Several small molecules, including quinacrine, Anle138b, and 2-aminothiazoles such as IND24 reduce levels of mouse PrP^Sc when applied to prion-infected cells or added to cell-free prion conversion assays (Ghaemmaghami et al., 2010; Korth et al., 2001; Wagner et al., 2013). Treatment with either IND24 or Anle138b significantly extends lifespan in mice following infection with mouse prion strains (Berry et al., 2013; Wagner et al., 2013). Unfortunately, none of these molecules has shown any efficacy against human prion strains, limiting their therapeutic potential (Berry et al., 2013; Giles et al., 2015). The inability for these anti-prion compounds to translate across prion strains from different species highlights the need for cellular paradigms capable of replicating a diverse range of prion strains, including those derived from species other than mice.

The species barrier, which is thought to arise from structural incompatibilities between PrP^C and PrP^Sc during templated misfolding, limits cross-species prion infection when there is an amino acid mismatch between two ortholog prion proteins (Collinge & Clarke, 2007; Prusiner et al., 1990). Even single amino acid changes can have profound effects on the efficiency of prion transmission (Barron et al., 2001). For instance, the G127V substitution in human PrP, which was originally found in individuals seemingly resistant to kuru infection (Mead et al., 2009), provides complete resistance to human prions (Asante et al., 2015). At the other end of the prion susceptibility spectrum, PrP from bank voles (BV) has been shown to function as a near-universal acceptor of prions, enabling efficient replication of a wide variety of prion strains stemming from several different animal species (Agrimi et al., 2008; Arshad et al., 2020; Burke, Mark, Walsh, et al., 2020; Burke, Walsh, Mark, et al., 2020; Cosseddu et al., 2011; Di Bari et al., 2013; Espinosa et al., 2016; Nonno et al., 2006, 2019; Orru et al., 2015; Piening et al., 2006; Pirisinu et al., 2016; Watts et al., 2014). BVPrP is polymorphic at codon 109, where either a methionine (M) or isoleucine (I) residue can be present (Cartoni et al., 2005). While both the M109 and I109 variants can act as efficient prion replication substrates, for certain prion strains the I109 polymorphism enables more rapid replication kinetics (Di Bari et al., 2013; Pirisinu et al., 2016, 2022). In transgenic mice, over-expression of BVPrP containing the I109 polymorphism causes a fatal neurodegenerative illness characterized by spontaneous generation of prions in the brain (Watts et al., 2012).

While primary glial cultures from bank voles enable cross-species prion replication (Schwenke et al., 2022), no immortalized bank vole cell lines currently exist, hindering our ability to understand and exploit the unique prion transmission properties exhibited by BVPrP. Expression of non-mouse PrP alleles in cell lines lacking endogenous PrP expression has expanded the range of prion strains that can be propagated in cultured cells (Bian et al., 2010; Courageot et al., 2008; Vilette et al., 2001). More recently, advances in gene editing technology have been used to generate PrP-null (PrP^−/−) versions of prion-permissive cell lines that can act as a blank canvas for expression of a desired PrP sequence (Arshad & Watts, 2022). For instance, murine CAD5 cells, which are susceptible to many different mouse prion strains (Mahal et al., 2007), can be rendered susceptible to hamster prions by ablation of mouse PrP and transfection with a hamster PrP-encoding vector (Bourkas et al., 2019). Furthermore, challenge of BVPrP-expressing CAD5-PrP^−/− with chronic wasting disease prions resulted in positive prion seeding activity in the RT-QuIC assay (Walia et al., 2019).

Given that CAD5 cells can become infected with a wide range of prion strains (Bourkas et al., 2019; Mahal et al., 2007), we hypothesized that expression of BVPrP in this line may facilitate cross-species prion infection. In this paper, we show that expression of either the M109 or I109 variants of BVPrP in CAD5-PrP^−/− cells enables cross-species prion infection with several different strains of mouse and hamster prions. Propagation of prions in BVPrP-expressing cells results in varying degrees of prion strain adaptation, leading to altered susceptibility to anti-prion small molecules for certain strains. Finally, the G127V substitution completely blocks the ability of cultured cells expressing BVPrP to become infected with prions, likely because it hinders the ability of BVPrP to polymerize into aggregates.

2 |. METHODS

2.1 |. General information

Institutional ethical approval was not required for this study. The study was not pre-registered. No blinding was performed, and there were no pre-determined data exclusion criteria. To observe a 50% difference in sample means with a power of 0.80 and an α-value of 0.05, a power calculation reveals that a minimum sample size of 3 is required, assuming an average standard deviation of 20%. Thus, a minimum of 3–4 independent replicates per condition were utilized, which is consistent with our previous studies performed using similar experimental paradigms (Arshad et al., 2021; Bourkas et al., 2019).

2.2 |. Cell culture

Murine CAD5 cells, which were obtained from Charles Weissmann, are a subclone of the catecholaminergic Cath.a-differentiated (CAD) line (Mahal et al., 2007; Qi et al., 1997). Although CAD5 cells are not a commonly misidentified cell line, their authenticity was not validated. Wild-type CAD5 cells, CAD5-PrP^−/− cells (clone D6) (Bourkas et al., 2019), and monoclonal lines of stably transfected CAD5-PrP^−/− cells were cultured in Opti-MEM media (Thermo Fisher #31985088) containing 5–10% (v/v) fetal bovine serum (Thermo Fisher #12483020), 1× Glutamax (Thermo Fisher #35050061), and 0.2× penicillin/streptomycin (Thermo Fisher #15140122). Cells were maintained at 37°C in a humidified incubator containing 5% CO₂ and were passaged at dilutions of 1:3–1:5 every 2–4 days, depending on the confluency. For passaging, cells were washed with PBS and then dissociated using an enzyme-free dissociation reagent (Millipore #S-014-B). Polyclonal lines of stably transfected CAD5-PrP^−/− cells were maintained in media containing 0.2 mg/ml G418. Cells were lysed following washing with PBS using lysis buffer (50 mM Tris–HCl pH 8.0, 150 mM NaCl, 0.5% (w/v) sodium deoxycholate, and 0.5% (v/v) NP-40). The lysates were kept on ice for 15 min, with vortexing every 5 min for 30 s, and then centrifuged at 5000 g for 5 min to pellet cellular debris. Protein concentrations in the supernatant were quantified using the bicinchoninic acid assay (Thermo Fisher #23227).

2.3 |. Generation of monoclonal and polyclonal stable cell lines

Expression vectors for mouse, hamster, and bank vole PrP (containing either the M109 or I109 polymorphism) were generated by inserting the coding sequences between the BamHI and XbaI sites of the plasmid vector pcDNA3, which also encodes resistance to the antibiotic G418 (Thermo Fisher). The G127V substitution was introduced into the bank vole PrP plasmids by performing site-directed mutagenesis using the following primers for mutagenic PCR: 5’-GTGGGGGGCCTGGGTGTCTACATGCTGGGGAGCG-3′ (forward) and 5’-CGCTCCCCAGCATGTAGACACCCAGGCCCCCCAC-3′ (reverse). CAD5-PrP^−/− cells were transfected using Lipofectamine 2000 (Thermo Fisher #11668019) according to the manufacturer’s protocol. Briefly, 2 μg of plasmid DNA was mixed with 4 μl of Lipofectamine 2000 in serum-free Opti-MEM media. Transfection mixtures were added to cells plated in 6-well dishes at a density of 5–7 × 10⁵ cells per well. After 24 h, the cells were passaged into 10 cm dishes containing 1 mg/ml G418. Stably transfected clones were selected over 10–14 days in antibiotic, and then divided into multiple 10 cm dishes. For polyclonal lines, stably transfected cells were pooled without further subcloning. For generation of monoclonal cell lines, stably transfected cells were diluted to 1.5 cells per 100 μl medium and then plated in 96-well plates. The plates were monitored over 14 days for the presence of single-cell-derived clones. Individual colonies were amplified sequentially in 24-well, 12-well, and then 6-well plates. Cells were lysed at the 24-well stage and checked for PrP^C expression by immunoblotting. Clones with high expression of PrP^C were picked and expanded further. Low passage number cells were used for all experiments. All cell lines will be shared upon reasonable request.

2.4 |. Immunoblotting

Samples were prepared in 1x Bolt LDS sample buffer (Thermo Fisher #B0007), with or without the addition of 2.5% (v/v) β-mercaptoethanol, boiled, and then loaded onto 10% Bolt Bis-Tris gels and run for 35 min at 165 V. Following SDS-PAGE, gels were transferred onto Immobilon-P PVDF membranes for 1 h at 25 V, using a Tris-glycine transfer solution (100 mM Tris–HCl pH 8, 137 mM Glycine). Following transfer, membranes were blocked for 1 h at 22°C in blocking buffer (5% (w/v) skim milk prepared in TBS containing 0.05% (v/v) Tween-20) and then probed with antibodies diluted in blocking buffer overnight at 4°C. The following day, blots were washed three times for 10 min each in TBS containing 0.05% (v/v) Tween-20 (TBST). Blots were then incubated with secondary antibody diluted in blocking buffer for 1 h at 22°C. Blots were again washed three times for 10 min each with TBST followed by incubation with horseradish peroxidase-conjugated secondary antibodies for 1 h at 22°C. At this point, the blots were developed using Western Lightning ECL Pro (PerkinElmer), followed by exposure to x-ray film or imaging using the LI-COR Odyssey Fc. For quantification of band intensity, densitometry was performed using ImageJ. The following anti-PrP antibodies were used in this study: recombinant humanized Fabs HuM-D18 (1:5000 dilution), HuM-D13 (1:5000 dilution), and HuM-P (1:10000 dilution) (Safar et al., 2002; Williamson et al., 1998); or the mouse monoclonal antibody POM1 (1:2000 dilution; Millipore Sigma #MABN2285) (Polymenidou et al., 2008). HuM-D13 was a generous gift from Stanley Prusiner (University of California San Francisco). Actin reprobes were performed using the 20–33 antibody (1:10000 dilution; Sigma Aldrich #A5060).

2.5 |. Immunofluorescence microscopy and cell surface PrP quantification

CAD5 cells or their derivatives were plated in 24-well dishes with a glass-like polymer coverslip bottom (Cellvis #P24-1.5P) coated with Poly-D-Lysine (Thermo Fisher #A3890401). Prior to the addition of cells, the coated wells were thoroughly washed and then air dried. Cells were plated at a density of 2.5 × 10⁵ cells/well and then cultured for 24–48 h until they reached ~90% confluency. Cells were then fixed with 4% (v/v) paraformaldehyde for 15 min and then washed twice with PBS. Cells were not permeabilized. Following the washes, the cells were blocked in 2% goat serum (diluted in PBS) for 1 h at 22°C. Cells were then incubated with the anti-PrP antibody POM1 (1:500 dilution) overnight at 4°C in PBS containing 2% goat serum. The following day, the cells were washed twice with PBS, and then incubated with an AlexaFluor 488-conjugated secondary antibody (Thermo Fisher; 1:500 dilution in PBS containing 2% goat serum) for 2 h at 22°C. Subsequently, cells were washed two additional times with PBS. After the final wash, the cells were incubated with DAPI (1:1000 dilution in PBS) for 10 min, and then washed with PBS. The cells were then stored in PBS until quantitative analysis of cell surface PrP levels using the BMG CLARIOstar plate reader or qualitative immunofluorescence imaging of PrP subcellular localization using a Zeiss LSM880 confocal microscope. Using the CLARIOstar, plates were analyzed for PrP expression (excitation: 488 ± 7 nm; emission: 535 ± 15 nm) and DAPI staining (excitation: 405 ± 10 nm; emission: 460 ± 15 nm) using a gain setting of 1500. Relative cell surface PrP expression levels were calculated using the PrP/DAPI signal ratio.

2.6 |. Prion strains

The hamster prion strains 263 K, 139H, HY, and DY were obtained as 10% (w/v) brain homogenates in PBS from terminally ill infected hamsters and were provided by Jason Bartz (Creighton University) or Valerie Sim (University of Alberta). To generate a cellular source of 263 K prions, CAD5-PrP^−/− cells stably expressing hamster PrP were infected with 263 K prions, passaged, and then expanded. The mouse prion strains RML, 22 L and ME7 were derived from the brains of terminally ill prion-infected non-transgenic mice. The brains of these mice were homogenized in 9 volumes of PBS to generate 10% (w/v) homogenates using a Minilys personal homogenizer and CK14 soft tissue homogenization tubes (Bertin). Homogenization was performed using 3 cycles of 60 s each at maximum speed, with 10 min of incubation on ice between cycles. To generate a cellular source of the mouse strains, wild-type CAD5 cells were infected with prion-containing brain homogenate, passaged, and then expanded. Cellular homogenates used as prion inocula were generated by washing prion-infected cells in 10 cm dishes with ice-cold PBS and then scraping the cells in PBS into CK14 soft tissue homogenization tubes. The tubes, which contain 1.4 mm zirconium oxide beads, were supplemented with additional 0.5 mm zirconia beads as well as 50 units/ml of Benzonase (EMD Millipore #70746-4). Cells were homogenized using the Minilys homogenizer as described above. Following homogenization, the homogenate was aliquoted and stored at −80°C. Brain- and cellular-derived prion inocula were used interchangeably in this study.

2.7 |. Cellular prion infections

Cells to be infected were plated in 12-well dishes at a density of 1.5 × 10⁵ cells/well. The next day, cells were infected with 100 μg of cellular homogenate or 1% (w/v) brain homogenate in 300 μl of growth media. For titration experiments, cells were infected with different amounts of cellular homogenate. After 72 h, cells were washed with PBS and then passaged at 1:3 dilution. Cells were continuously passaged in 12-well plates for at least 3–6 passages, and then scaled up to 6 cm dishes prior to lysis for analysis of infection status. As negative controls, cells were exposed to non-infected brain homogenate or cellular homogenate from non-infected cells.

2.8 |. Enzymatic digestions

For the detection of the PrP C1 cleavage fragment, 100 μg of cell lysate was denatured and then deglycosylated using 0.5 μl of PNGase F (New England BioLabs #P0704S) overnight at 37°C according to the manufacturer’s instructions. For the analysis of PrP^Sc, 0.5–1 mg of cell lysate was digested with proteinase K (PK) (Thermo Fisher #EO0491) at a concentration of 50 μg/ml (enzyme to protein ratio of 1:50) to allow for an aggressive digestion of naïve PrP^C. Digestions were conducted at 37°C with shaking at 600 rpm for 1 h. The digestions were stopped with 2 mM PMSF, and then sarkosyl (Sigma Aldrich #61747) was added to a final concentration of 2% (v/v), followed by ultracentrifugation at 100000 g for 1 h at 4°C to isolate the PK-resistant detergent insoluble fraction. Following ultracentrifugation, the supernatant was discarded, and the pellet was resuspended in 1x LDS sample buffer, boiled, and analyzed by immunoblotting as described above.

2.9 |. Conformational stability assays

Cellular homogenates from prion-infected cells were detergent extracted by adding 0.1 volumes of 10x detergent buffer (5% NP-40, 5% sodium deoxycholate in PBS). Following centrifugation (5000 g for 5 min) to clear cellular debris, 20 μl of detergent extracted cellular homogenate was mixed with 20 μl of 2x stocks of guanidine hydrochloride (GdnHCl) to generate final GdnHCl concentrations of 1, 1.5, 2, 2.5, 3, 3.5, and 4 M. Following the addition of GdnHCl, samples were incubated at 22°C with shaking at 800 rpm for 2 h. Next, all samples were normalized to 0.4 M GdnHCl and then digested with proteinase K at a concentration of 20 μg/ml. This digestion was conducted at 37°C with shaking at 600 rpm for 1 h. The digestion was stopped with the addition of PMSF (2 mM) followed by the addition of sarkosyl to a final concentration of 2%. A white precipitate often forms during the digestion; to address this, the samples were kept on ice with vortexing every 10 min until the precipitate disappeared. The PK-digested samples were then ultracentrifuged at 100000 g for 1 h at 4°C. Following ultracentrifugation, the supernatant was discarded, and the pellet resuspended in 1x LDS sample buffer. The resuspended pellet was boiled and then analyzed by immunoblotting.

2.10 |. Drug treatment of prion-infected cells

Quinacrine dihydrochloride was purchased from Sigma Aldrich (#Q3251) and was dissolved in dH₂O. IND24 and Anle138b were custom synthesized by Sundia (Sacramento, CA) and were dissolved in DMSO. Prion-infected cell lines were plated at roughly 7.5 × 10⁵ cells/well in 6-well dishes. The following day, the cells were either left untreated, treated with DMSO, or were treated with quinacrine, IND24, or Anle138b at various concentrations. The plates were incubated for 72 h, followed by lysis and analysis of proteinase K-resistant PrP by immunoblotting. For longer treatments, the cells were passaged three times in the continuous presence of 2 μM quinacrine, IND24, or Anle138b, and then lysed.

2.11 |. Thioflavin T aggregation assays

Untagged recombinant BVPrP (residues 23–231) containing either the M109 or I109 polymorphism, as well as their counterparts containing the G127V substitution, were expressed and purified in E. coli Rosetta2(DE3) cells as previously described for recombinant hamster PrP(23–231) (Bourkas et al., 2019). Recombinant mouse PrP(23–230) was produced similarly (Arshad et al., 2021). Purified recombinant PrP was stored at −80°C in 10 mM sodium phosphate buffer, pH 5.8. For Thioflavin T (ThT) assays, recombinant PrP was dialyzed into 10 mM sodium phosphate buffer, pH 7.4. Following dialysis, proteins were ultracentrifuged at 100000 g for 1 h at 4°C to remove any pre-existing aggregates. The reaction mixture was as follows: 0.1 mg/ml recombinant PrP, 10 μM ThT, 135 mM NaCl in a 100 μl reaction volume. The reaction mixtures were plated in a black, clear bottom 96-well Nunc plate (Thermo Fisher #265301), and then the plate was sealed with an optically clear film (VWR #89134-428). No beads were added to the plates. The plate was then placed in a BMG CLARIOstar plate reader and was incubated at 37°C for 83.25 h with cycles of shaking (1 min) and rest (4 min). ThT readings (excitation 444 ± 5 nm; emission: 485 ± 5 nm) were taken every 5 minutes, and the lag phases were quantified as previously described (Arshad et al., 2021). Samples that failed to aggregate were assigned a lag phase of 83.25 h.

2.12 |. Statistical analysis

The distribution of the data, except for the lag phases from the ThT assays, was assumed to be normal, but this was not formally tested. Standard deviations were not assumed to be equal between experimental groups. When comparing the means of two samples, two-tailed unpaired t tests with Welch’s correction were used. When comparing the means of more than two samples, a Welch’s ANOVA with Dunnett’s T3 multiple comparisons test was used. For comparing lag phases for recombinant PrPs in the ThT aggregation assay, a non-parametric Kruskal–Wallis test followed by Dunn’s multiple comparisons test was utilized. No data points were excluded from analysis and no formal test for outliers was performed. All statistical analysis was performed using GraphPad Prism software (version 9.4.1) with a significance threshold of p < 0.05.

3 |. RESULTS

3.1 |. Susceptibility of monoclonal lines of PrP-transfected CAD5-PrP^−/− cells to prions

We previously reported that polyclonal lines of CAD5-PrP^−/− cells that had been stably transfected with vectors encoding expression of either hamster (Ha) or mouse (Mo) PrP rendered the cells susceptible to hamster and mouse prions, respectively (Bourkas et al., 2019). However, the presence of the selective agent G418 in the culture medium, which was required for maintenance of PrP^C expression in the polyclonal lines, significantly hindered the ability of the cells to become infected with prions (Arshad et al., 2021). To circumvent the inhibitory effect of G418 and the drift that occurs in a heterogeneous population of cells, we isolated monoclonal cell lines from pools of stably transfected CAD5-PrP^−/− cells expressing either MoPrP, HaPrP, or BVPrP. Monoclonal cell lines were generated for both the M109 and I109 BVPrP polymorphic variants, referred to hereafter as BV^M and BV^I, respectively. Each clone was selected based on its relative amount of PrP^C expression, with all clones expressing PrP^C at similar or higher levels than the level of MoPrP present in wild-type CAD5 cells. (Figure 1a). In all the monoclonal lines, PrP^C expression at the cell surface was detected by immunofluorescence in the absence of permeabilization (Figure 1b). The relative levels of cell surface PrP^C were quantified and were in general agreement with the levels of total PrP^C observed by immunoblotting (Figure 1c). The transfected PrPs were processed similarly to MoPrP in wild-type CAD5 cells, as revealed by the presence of the C1 endoproteolytic fragment (Altmeppen et al., 2012) following removal of N-linked glycans (Figure 1a). The lower signals observed for HaPrP-expressing cells when using the POM1 antibody likely reflect a reduced ability of POM1 to recognize HaPrP rather than lower levels of protein expression (Figure S1).

FIGURE 1.

Expression of bank vole PrP permits cross-species prion replication in monoclonal lines derived from CAD5-PrP^−/− cells. (a) Immunoblots of PrP^C levels in cell lysates from monoclonal lines of CAD5-PrP^−/− cells stably expressing either mouse PrP (Mo), hamster PrP (Ha), bank vole PrP-M109 (BV^M), or bank vole PrP-I109 (BV^I). Lysates from wild-type (wt) CAD5 cells and CAD5-PrP^−/− cells are included as controls. In the bottom blot, lysates were treated with PNGaseF to detect de-glycosylated full-length PrP and the C1 cleavage fragment. Blots were probed with the anti-PrP antibodies HuM-D13 or POM1. The HuM-D13 blot was also probed with an antibody against actin. (b) Immunofluorescence images of non-permeabilized wt CAD5 cells, CAD5-PrP^−/− cells, and transfected monoclonal lines expressing the indicated PrP molecules. Cell surface PrP expression (green) was detected using the antibody POM1. Scale bar = 20 μm. (c) Quantification of cell surface PrP expression relative to DAPI levels in the various cell lines, normalized to levels in wt CAD5 cells (n = 2 independent replicates). (d) Immunoblots of proteinase K (PK)-resistant PrP^Sc (PrP^res) levels in lysates from cell lines expressing mouse PrP, hamster PrP, or bank vole PrP after six passages following challenge with normal brain or cellular homogenate (Ctrl), RML prions, or 263 K prions. PrP was detected using the antibody HuM-P. (e) Immunoblots of PrP^res levels in cells expressing mouse PrP or bank vole PrP after six passages following challenge with the indicated amounts of cell homogenate containing RML prions. PrP was detected using the antibody HuM-P. (f) Immunoblots of PrP^res levels in cells expressing hamster PrP or bank vole PrP after six passages following challenge with the indicated amounts of cell homogenate containing 263 K prions. PrP was detected using the antibody HuM-D13.

The monoclonal cell lines were assessed for their susceptibility to infection with either the RML strain of mouse prions or the 263 K strain of hamster prions. Cells were exposed to prions for 3 days and then passaged 3–6 times to remove all traces of the initial inoculum and to allow for sufficient de novo accumulation of prions. Prion infection was gauged by the presence of detergent insoluble, proteinase K-resistant PrP (PrP^res) in cell lysates. As expected, MoPrP- and HaPrP-expressing lines were susceptible to their cognate prion strains, but resistant to cross-species prion conversion (Figure 1d). On the other hand, cells expressing BV^M or BV^I were susceptible to both mouse RML and hamster 263 K prions. No PrP^res was observed in any of the cell lines exposed to samples that did not contain prions. The susceptibility of BVPrP-expressing cells to RML and 263 K prions relative to cells expressing MoPrP or HaPrP, respectively, was examined by performing titration experiments in which different amounts of prion inocula were applied to the cells at the time of infection. The BV^M and BV^I-expressing cells exhibited comparable susceptibility to RML prions, which was only marginally reduced compared to the MoPrP-expressing cells (Figure 1e). Similarly, the cells expressing HaPrP were only slightly more susceptible to 263 K prions than cells expressing BV^M or BV^I (Figure 1f). Following infection with hamster prions, levels of PrP^res were more pronounced in cells expressing HaPrP compared to cells expressing BV^M or BV^I (Figure S2), which is consistent with results from BVPrP-expressing transgenic mice (Watts et al., 2014).

3.2 |. Infection of BVPrP-expressing cells with additional mouse and hamster prion strains

To further assess the prion susceptibility of BVPrP-expressing CAD5-PrP^−/− cells, the monoclonal BV^M and BV^I lines were challenged with a variety of mouse and hamster prion strains. The RML, 22 L, and ME7 mouse strains were examined, which are all strains of mouse-adapted scrapie prions (Chandler, 1961; Kimberlin et al., 1989). For hamster prions, the 263 K and 139H strains of hamster-adapted scrapie prions were examined (Kimberlin et al., 1989; Kimberlin & Walker, 1977), as well as the Hyper (HY) and Drowsy (DY) strains (Bessen & Marsh, 1992b), which are derived from transmissible mink encephalopathy. After six passages following prion exposure, both the BV^M and BV^I cell lines were readily infected with the RML and 22 L strains of mouse prions (Figure 2a). Similarly, abundant PrP^res was observed in BV^M and BV^I cells challenged with the 263 K and HY strains of hamster prions (Figure 2b). Mouse ME7 and hamster 139H strains accumulated to much lower levels in the BV^M and BV^I cells relative to the other strains, but following additional passages, PrP^res could be readily detected for both strains in the two lines (Figure 2c,d). This is consistent with the low conversion efficiency of BVPrP^C when templated with mouse ME7 prions in an in vitro conversion assay (Piening et al., 2006).

FIGURE 2.

Susceptibility of monoclonal bank vole PrP-expressing CAD5-PrP^−/− cells to a panel of mouse and hamster prion strains. (a, b) Immunoblots of proteinase K (PK)-resistant PrP (PrP^res) levels in lysates from monoclonal cell lines expressing either bank vole PrP-M109 (BV^M) or bank vole PrP-I109 (BV^I) after 6 passages following challenge with the 22 L, RML, and ME7 strains of mouse prions (a) or the 263 K, 139H, Hyper (HY), and Drowsy (DY) strains of hamster prions (b). As a negative control, cells were exposed to normal brain homogenate or PBS (ctrl). (c, d) Immunoblots of PrP^res levels in lysates from monoclonal cell lines expressing either BV^M or BV^I after additional passages to allow accumulation of difficult-to-replicate strains such as hamster 139H prions and mouse ME7 prions. (e) Immunoblot of PrP^res levels in lysates from monoclonal CAD5-PrP^−/− cells stably expressing BV^M at passage 6 following challenge with BV^I-adapted HY (BV^I.HY) and DY (BV^I.DY) prions. In all panels, PrP was detected using the antibodies HuM-P or HuM-D13, as indicated.

While the BV^I-expressing cells were susceptible to infection with DY prions (Figure 2b), all attempts to infect the BV^M-expressing cells with DY prions failed. DY infection in the BV^I cells was characterized by faster electrophoretic mobility of the unglycosylated PrP^res band compared to the corresponding band in HY-infected BV^I cells, which is consistent with the differences observed in the original hamster strains (Bessen & Marsh, 1992a). Interestingly, BV^M cells could be infected with the DY strain following adaptation to the BV^I sequence (Figure 2e). As in the cells expressing BV^I, PrP^res from the BV^M cells infected with BV^I-adapted DY prions migrated more rapidly than the PrP^res from cells infected with BV^I-adapted HY prions. Thus, expression of BVPrP in CAD5-PrP^−/− cells permits cross-species prion infection with several distinct strains of mouse and hamster prions.

3.3 |. Biochemical characterization of BVPrP-adapted prion strains

Cross-species prion replication can often lead to PrP^Sc conformational adaptation and the emergence of strains with distinct biochemical and pathological properties (Baskakov, 2014). While BVPrP can act as a promiscuous substrate for prion strains, the consequences of prion strain adaptation to the BVPrP sequence remain unclear. Given that the BV^M-expressing cells could be efficiently infected with the RML and 263 K strains, we examined the biochemical properties of these BVPrP-adapted strains, which will be referred to as BV^M.RML and BV^M.263 K, respectively. To do this, we employed the conformational stability assay, which assesses the ability of PrP^Sc aggregates to resist solubilization in the presence of increasing concentrations of guanidine hydrochloride (GdnHCl). The GdnHCl concentration at which half the PrP^Sc aggregates are solubilized is referred to as the [GdnHCl]₅₀ value, which can be used as a prion strain-specific measure of PrP^Sc stability (Peretz et al., 2001). When comparing the stabilities of RML and BV^M.RML prions, we found that adaptation of mouse RML prions to the BVPrP sequence resulted in a significant increase in [GdnHCl]₅₀, from ~1.0 M to ~1.7 M (Figure 3a–c). A similar result was obtained with hamster 263 K prions, with a [GdnHCl]₅₀ of ~1.4 M for the original 263 K prions compared to ~1.9 M for the BV^M.263 K prions (Figure 3d–f). While the change in [GdnHCl]₅₀ did not reach statistical significance, the increased stability of BV^M.263 K prions was consistently observed. Thus, adaptation to the BVPrP sequence increases the conformational stability of the RML and 263 K strains.

FIGURE 3.

Conformational stability and glycoform analysis of bank vole PrP-adapted prion strains. (a) Representative immunoblots for proteinase K (PK)-resistant PrP (PrP^res) in lysates from RML-infected wild-type CAD5 cells or monoclonal CAD5-PrP^−/− cells stably expressing bank vole PrP-M109 (BV^M) subjected to the conformational stability assay, which involves treatment with increasing concentrations of guanidine hydrochloride (GdnHCl). (b) GdnHCl denaturation curves for mouse RML prions (black) and bank vole PrP-adapted RML (BV^M.RML) prions (red). Data are mean ± SD from 3 to 4 independent replicates. (c) Quantification of [GdnHCl]₅₀ values for RML and BV^M.RML prions. The BV^M.RML prions were significantly more stable than RML prions, as determined by a two-tailed, unpaired Welch’s t test (df = 3.685, t = 3.881, p = 0.0208). Data are mean ± SD from 3 to 4 independent replicates. (d) Representative immunoblots for PrP^res in lysates from 263 K-infected CAD5-PrP^−/− cells stably expressing either hamster PrP or BV^M subjected to the conformational stability assay. (e) GdnHCl denaturation curves for 263 K prions (blue) and BV^M.263 K prions (red). Data are mean ± SD from 3 to 4 independent replicates. (f) Quantification of [GdnHCl]₅₀ values for 263 K and BV^M.263 K prions. While the BV^M.263 K prions were consistently more stable than 263 K prions, this result did not reach statistical significance, as determined by a two-tailed, unpaired Welch’s t test (df = 3.053, t = 2.335, p = 0.0984). Data are mean ± SD from 3 to 4 independent replicates. (g) Percentage of diglycosylated (left plot), monoglycosylated (middle plot), and unglycoslyated (right plot) PrP^res glycoforms for cell-derived RML and 263 K prions as well as their bank vole PrP-adapted counterparts. Data are mean ± SD from five independent samples per strain. Significant differences were assessed using Welch’s ANOVA followed by Dunnett’s T3 multiple comparisons test. Statistical reports are shown in Table S1. To outline the three PrP^res glycoforms, an immunoblot of PrP^res from 263 K prion-infected cells expressing hamster PrP is also shown. In panels a and d, PrP was detected with the antibody HuM-P.

Since the relative proportions of the three PrP^res N-linked glycoforms (di-, mono-, and unglycosylated) can also be used to distinguish prion strains (Collinge et al., 1996), we quantified the glycosylation patterns of RML and 263 K prions and compared them to their BVPrP-adapted equivalents (Figure 3g). For both the BV^M.RML and BV^M.263 K strains, PrP^res glycoform distributions differed significantly from the parental strains. The changes were most notable for the diglycosylated and unglycosylated PrP^res glycoforms. Compared to RML prions, an increased proportion of the diglycosylated species at the expense of the unglycosylated species was observed for BV^M.RML prions. The opposite was true for 263 K prions, whereby a decrease in the proportion of diglycoslated species as well as a compensatory increase in the unglycosylated glycoform was found for BV^M.263 K prions. Therefore, adaptation to BVPrP results in modest alterations to the PrP^res glycoform profile, which could potentially indicate conformational divergence from the parental strains. An important caveat to the conformational stability and glycoform analyses is that the parental and BVPrP-adapted prion strains have different amino acid sequences, which can complicate strain comparison (Otero et al., 2022). Nonetheless, there are examples of prion strains that maintain their properties when propagated using very different PrP sequences (Vidal et al., 2013).

3.4 |. Comparative susceptibility of BVPrP-adapted prion strains to anti-prion compounds

We next compared the susceptibility of native and BVPrP-adapted prion strains to the anti-prion compounds quinacrine, IND24, and Anle138b. Although the mechanisms of action of these drugs remain unknown and may overlap, they were selected based on their ability to delay the onset of prion disease in prion-inoculated mice (IND24 and Anle138b), modulate PrP^res levels in prion-infected cells (quinacrine, IND24, and Anle138b), or to exhibit strain-specific efficacy (Berry et al., 2013; Bian et al., 2014; Burke, Mark, Kun, et al., 2020; Doh-ura et al., 2000; Korth et al., 2001; Wagner et al., 2013). Initially, we treated BV^M.RML- or BV^M.263 K-infected cells with various concentrations of the drugs ranging from 0 to 5 μM for 72 h. However, we found that at this timepoint, drug-treated cells infected with BVPrP-adapted prion strains showed almost no change in PrP^res levels (Figure S3). Therefore, we treated RML and BV^M.RML-infected cells with 2 μM quinacrine, IND24, or Anle138b over three passages (13 days) and then analyzed the reduction in PrP^res levels relative to DMSO-treated or untreated cells (Figure 4a). For RML prions, a significant (>90%) reduction in PrP^res levels was observed for all three anti-prion compounds (Figure 4b,c). In cells infected with BV^M.RML prions, levels of PrP^res were reduced by ~80–85% by treatment with the three anti-prion drugs (Figure 4d,e). Thus, although RML prions undergo strain adaptation when templated to BVPrP, they remain largely susceptible to the drugs that are efficacious against the original strain. Drug treatment experiments were also conducted using 263 K and BV^M.263 K prions. Neither IND24 nor Anle138b was effective at reducing PrP^res levels for either strain (Figure 4f–i). However, when treated with 2 μM quinacrine over three passages, cells infected with 263 K prions, but not cells infected with BV^M.263 K prions, showed a drastic reduction in PrP^res.

FIGURE 4.

Susceptibility of bank vole PrP-adapted prion strains to anti-prion small molecules. (a) Schematic of the drug treatment paradigm. (b) Representative immunoblot of proteinase K (PK)-resistant PrP (PrP^res) in cell lysates from RML-infected monoclonal CAD5-PrP^−/− cells expressing mouse PrP (Mo) left untreated or treated with DMSO, 2 μM quinacrine, 2 μM IND24, or 2 μM Anle138b for three passages. (c) Quantification of PrP^res levels in drug-treated, RML-infected CAD5-PrP^−/− cells expressing mouse PrP. (d) Representative immunoblot of PrP^res in cell lysates from BV^M.RML-infected CAD5-PrP^−/− cells expressing bank vole PrP-M109 (BV^M) left untreated or treated with the indicated compounds for three passages. (e) Quantification of PrP^res levels in drug-treated, BV^M.RML-infected CAD5-PrP^−/− cells expressing BV^M. (f) Representative immunoblot of PrP^res in cell lysates from 263 K-infected monoclonal CAD5-PrP^−/− cells expressing hamster PrP (Ha) left untreated or treated with the indicated compounds for three passages. (g) Quantification of PrP^res levels in drug-treated, 263 K-infected CAD5-PrP^−/− cells expressing hamster PrP. (h) Representative immunoblot of PrP^res in cell lysates from BV^M.263 K-infected monoclonal CAD5-PrP^−/− cells expressing BV^M left untreated or treated with the indicated compounds for 3 passages. (i) Quantification of PrP^res levels in drug-treated, BV^M.263 K-infected CAD5-PrP^−/− cells expressing BV^M. In panels b, d, f, and h, data are mean ± SD from 3 to 4 independent replicates. Statistical significance was determined by Welch’s ANOVA followed by Dunnett’s T3 multiple comparisons test. Statistical reports are shown in Table S1. In panels b, d, f, and h, PrP was detected with the antibody HuM-P.

3.5 |. Transmission properties of BVPrP-adapted prion strains

Having demonstrated that the properties of BVPrP-adapted prion strains are not identical to the original strains from which they are derived, we next assessed whether the BVPrP-passaged strains retained memory of the original strains and could be “retro-transmitted” to their native host. To test this, we exposed monoclonal CAD5-PrP^−/−(MoPrP) and CAD5-PrP^−/−(HaPrP) cells to BVPrP-adapted mouse and hamster prion strains. The original mouse and hamster strains were used as positive controls. Cells were passaged six times and then analyzed for the presence of PrP^res. We found that CAD5-PrP^−/−(MoPrP) cells could be infected with RML, 22 L, and ME7 prions, but were resistant to their BVPrP-adapted counterparts (BV^M. RML, BV^M.22 L, and BV^M.ME7) (Figure 5a). Additionally, we found that CAD5-PrP^−/−(MoPrP) monoclonal cells could not be infected with BV^M.263 K prions. In contrast, CAD5-PrP^−/−(HaPrP) cells were susceptible to three hamster prion strains (263 K, HY, and 139H) as well as their BVPrP-passaged counterparts, yet were resistant to RML and BV^M.RML prions (Figure 5b). To confirm that the BVPrP-adapted mouse prions were actually infectious, we performed a “second passage” experiment in which monoclonal CAD5-PrP^−/− cells expressing BV^M were treated with several different BVPrP-adapted prion strains (Figure 5c). Our results indicate that BVPrP-adapted prion strains can be robustly propagated upon second passage.

FIGURE 5.

Retro-transmission of bank vole PrP-adapted prion strains to CAD5-PrP^−/− cells expressing mouse or hamster PrP. (a) Immunoblots of proteinase K (PK)-resistant PrP (PrP^res) levels in lysates from monoclonal CAD5-PrP^−/− cells stably expressing mouse PrP (Mo) at passage 6 following challenge with the indicated mouse (RML, 22 L, ME7), hamster (263 K), or bank vole PrP-adapted (BV^M.RML, BV^M.22 L, BV^M.ME7, BV^M.263 K) prion strains. (b) Immunoblot of PrP^res levels in lysates from monoclonal CAD5-PrP^−/− cells stably expressing hamster PrP (Ha) at passage 6 following challenge with the indicated mouse (RML), hamster (263 K), or bank vole PrP-adapted (BV^M.RML, BV^M.263 K, BV^M.139H, BV^M.HY) prion strains. (c) Immunoblots of PrP^res levels in lysates from monoclonal CAD5-PrP^−/− cells stably expressing BV^M at passage 6 following challenge with the indicated bank vole PrP-adapted prion strains. In all panels, PrP was detected using the antibodies HuM-P or HuM-D13, as indicated.

3.6 |. The G127V substitution blocks cross-species and species-matched prion transmission in cells expressing BVPrP

Given the ability of BVPrP to facilitate cross-species prion infection in cultured cells, we were curious whether the protective G127V substitution would influence the prion permissiveness of BVPrP. To do this, we generated polyclonal lines of CAD5-PrP^−/− cells that were stably transfected with either wild-type BV^M or BV^I or with G127V-mutant versions of both PrP alleles. Levels of PrP expression were comparable across all four lines, and PrP expression was observed at the cell surface (Figure 6a,b). The cell lines were then challenged with two strains of hamster prions (263 K and Hyper) and two strains of mouse prions (RML and 22 L). Following six passages, cells were lysed and then analyzed for the presence of PrP^res. These experiments were performed in the presence of a low concentration of G418 (0.2 mg/ml) to promote maintenance of PrP expression while minimizing the inhibitory effects of G418 on de novo prion infection (Arshad et al., 2021). Cells expressing wild-type BV^M were readily infected with all prion strains, whereas cells harboring BV^M with the G127V substitution could not be infected with any of the four strains (Figure 6c). Cells expressing the BV^I allotype behaved similarly, as cells expressing wild-type BV^I could readily become infected with hamster HY prions, but their counterparts with the G127V substitution were resistant (Figure 6d). To check if the G127V substitution could also prevent prion infection when exposed to BVPrP-adapted prions, cells expressing either wild-type BV^M or BV^M(G127V) were challenged with BV^M.RML and BV^M.263 K prions. Abundant accumulation of PrP^res was observed in the wild-type BV^M cells, whereas none was detected in cells expressing the G127V variant (Figure 6e). These results demonstrate that the protective properties associated with the G127V substitution can even counteract a highly prion-permissive substrate such as BVPrP.

FIGURE 6.

The G127V polymorphism blocks cross-species prion infection in polyclonal lines of bank vole PrP-expressing CAD5-PrP^−/− cells. (a) Immunoblot of PrP^C levels in cell lysates from polyclonal lines of CAD5-PrP^−/− cells stably expressing wild-type (wt) bank vole PrP-M109 (BV^M) or bank vole PrP-I109 (BV^I), or stably expressing BV^M or BV^I with the G127V substitution (V127). PrP was detected using the antibody POM1, and the blot was also probed with an antibody against actin. (b) Immunofluorescence images of non-permeabilized CAD5-PrP^−/− cells as well as stably transfected polyclonal lines expressing the indicated PrP molecules. Cell surface PrP expression (green) was revealed using the antibody POM1. Scale bar = 20 μm. (c) Immunoblots of proteinase K (PK)-resistant PrP (PrP^res) levels in lysates from polyclonal cells expressing either wt or G127V-mutant BV^M after six passages following challenge with mouse prions (RML or 22 L strains) or hamster prions (263 K or HY strains). PrP was detected using the antibody HuM-P. (d) Immunoblot of PrP^res levels in lysates from polyclonal cells expressing either wt or G127V-mutant BV^I after six passages following challenge with hamster HY prions. PrP was detected using the antibody HuM-P. (e) Immunoblots of PrP^res levels in lysates from polyclonal cells expressing either wt or G127V-mutant BV^M after six passages following challenge with BV^M.RML prions or BV^M.263 K prions. PrP was detected using the antibody HuM-P.

3.7 |. The G127V substitution delays the aggregation kinetics of BVPrP

To understand how the G127V substitution might inhibit the ability of BVPrP to facilitate cross-species prion infection, recombinant untagged full-length (residues 23–231) wild-type BVPrP and BVPrP(G127V) were generated by expression in E. coli. Both the M109 and I109 allotypes of BVPrP were analyzed (Figure 7a). The relative abilities of the recombinant (rec) BVPrPs to spontaneously aggregate were assessed using a Thioflavin T (ThT) kinetic assay. Following incubation at 37°C with periodic shaking in a physiological buffer containing ThT, wild-type recBV^M rapidly formed ThT-positive aggregates with a mean lag phase of ~10 h (Figure 7b,d). The same was true of wild-type recBV^I, which exhibited a mean lab phase of ~7 h (Figure 7c,d). All the wild-type recBV^M (36/36) and recBV^I (18/18) replicates formed aggregates within the timeframe of the experiment (Figure 7d). In contrast, recBV^M(G127V) rarely aggregated, with only 33% of replicates (12/36) becoming positive, and in the instances in which aggregation occurred, it was significantly delayed (Figure 7b,d). While a higher proportion of recBV^I(G127V) replicates formed aggregates (72%; 13/18) compared to recBV^M(G127V), the aggregation lag phases were still greatly extended compared to wild-type recBV^I (Figure 7c,d). Overall, we found that wild-type recBVPrP forms aggregates very rapidly, whereas the presence of the protective G127V substitution significantly delays aggregation or inhibits it entirely.

FIGURE 7.

The G127V polymorphism delays spontaneous aggregation of recombinant bank vole PrP. (a) SDS-PAGE followed by staining with Coomassie blue of recombinant (rec) bank vole PrP-M109 (BV^M) and bank vole PrP-I109 (BV^I) (residues 23–231) along with their G127V-containing counterparts. (b) Representative aggregation curves for recBV^M (black) and recBV^M(G127V) (red), as determined using a Thioflavin T (ThT) assay. For each protein, data are mean ± SD from six independent replicates. (c) Representative aggregation curves for recBV^I (black) and recBV^I(G127V) (red), as determined using a ThT assay. For each protein, data are mean ± SD from six independent replicates. (d) Quantification of the aggregation lag phases for the ThT assays. For both recBV^M and recBV^I (black) the G127V substitution (red) significantly lengthened the lag phase, as determined by a Kruskal–Wallis test followed by Dunn’s multiple comparisons test. Blue data points indicate replicates that failed to form aggregates within the experimental timeframe. Data are mean ± SD from 36 independent replicates for the BV^M proteins and 18 independent replicates for the BV^I proteins.

4 |. DISCUSSION

This study showcases the utility of PrP-transfected CAD5-PrP^−/− cells as a flexible paradigm for investigating key aspects of prion biology, such as the species barrier, prion strain adaptation, and the susceptibility of various prion strains to anti-prion drugs, within an intact cellular environment that contains glycosylated, GPI-anchored PrP. Both the M109 and I109 polymorphic variants of BVPrP functioned as permissive substrates, enabling cross-species prion replication when exposed to various strains of hamster and mouse prions. These results confirm and extend recent findings that BVPrP can permit cross-species prion infection in cultured cells (Schwenke et al., 2022; Walia et al., 2019). However, the monoclonal BVPrP-expressing CAD5-PrP^−/− lines described herein offer several important advantages. First, PrP^res was readily observed following infection with either hamster or mouse prions, whereas in a distinct line of CAD5-PrP^−/− cells expressing BVPrP, prion infection could often only be detected using the ultrasensitive RT-QuIC technique (Walia et al., 2019). This suggests that prions accumulate to higher levels in our cellular models, perhaps because of their monoclonal nature. Second, unlike primary glia derived from bank voles (Schwenke et al., 2022), which do not divide in culture, it is easy to distinguish between de novo PrP^res produced by conversion of PrP^C in the cultured cells and residual levels of PrP^res because of persistence of the prion inoculum. Finally, since PrP^−/− cells are used a starting point in our system, it is straightforward to test the effects of amino acid changes in PrP, such as G127V, on prion susceptibility. This should help to clarify the molecular determinants that allow BVPrP to function as a universal prion acceptor (Agrimi et al., 2008; Burke, Mark, Walsh, et al., 2020; Kurt et al., 2017; Piening et al., 2006).

All tested mouse and hamster strains were able to be propagated in CAD5-PrP^−/− cells expressing BVPrP, except for hamster DY prions, which could only be replicated in cells expressing BV^I. This is consistent with the observation that I109 bank voles are better at replicating certain atypical prion strains than M109 bank voles (Nonno et al., 2019; Pirisinu et al., 2016, 2022). Interestingly, once the DY strain was templated onto the sequence of BV^I, it could then be replicated in cells expressing BV^M, with maintenance of its unique migration pattern on an immunoblot. Strangely, despite multiple attempts, we were unable to observe any signs of PrP^res in monoclonal cells expressing HaPrP following infection with DY prions, despite the cells expressing PrP that is perfectly sequence-matched to DY PrP^Sc. However, it remains possible that lower levels of prions might be present in the DY-challenged HaPrP-expressing cells, which may be detectable using a more sensitive technique such as RT-QuIC. Indeed, another group has found that polyclonal CAD5-PrP^−/− cells expressing HaPrP can display very low levels of PrP^res upon infection with DY prions (Cortez et al., 2021). These experiments suggest that the DY strain may require specific co-factors that are absent in CAD5 cells to be efficiently propagated using HaPrP. Because of its permissive nature, BV^I may permit replication of DY prions in the absence of co-factors. Once templated onto BV^I, sequence similarity may enable transfer to the BV^M sequence in the absence of the hypothetical co-factor. Indeed, prion strain-specific co-factor preferences are predominantly determined by PrP^Sc conformation, not its primary sequence (Burke, Walsh, Mark, et al., 2020).

Mouse prions underwent substantial adaptation upon templating onto BVPrP, as evidenced by alterations to the PrP^res glycoform profile and an increase in conformational stability for the RML strain, as well as the inability of all BVPrP-passaged mouse strains to be retro-transmitted to cells expressing MoPrP. This is generally consistent with results from transgenic mice, where although BVPrP-adapted RML prions could be retro-transmitted to transgenic mice expressing MoPrP, the incubation periods were significantly longer than would be predicted had no strain adaptation occurred (Watts et al., 2014). On the other hand, hamster prions experienced only minor adaptation, as the 263 K, HY, and 139H strains could be easily retro-transmitted back to the original HaPrP host substrate. This lack of significant adaptation of hamster prions upon passage to BVPrP was also observed using transgenic mice (Watts et al., 2014). One possible explanation for the differential adaptation behavior of mouse and hamster prions relates to their N-glycosylation patterns. While diglycosylated PrP^res dominated in the hamster 263 K strain, monoglycosylated PrP^res was the most prevalent glycoform for the mouse RML strain. All the BVPrP-adapted strains featured a diglycosylation-dominant PrP^res banding pattern. Thus, strains that naturally exhibit diglycosylation-dominant PrP^res may replicate easier using BVPrP as a substrate, leading to fewer structural alterations, which is in general agreement with in vitro prion replication experiments using diglycosylated and unglycosylated PrP substrates (Burke, Walsh, Mark, et al., 2020). The differences in the retro-transmission properties of the BV^M.RML and BV^M.263 K strains also highlight the fact that they remain distinct strains, despite sharing certain biochemical features. Therefore, prions do not converge into a single strain upon adaptation to BVPrP. Furthermore, since BV^M.263 K prions failed to transmit to cells expressing MoPrP and BV^M.RML prions failed to transmit to cells expressing HaPrP, our data reaffirm that although bank vole PrP^C can function as a near-universal acceptor of prions, it does not behave as a “universal donor” of prions. It is important to note that we only characterized mouse and hamster prion strains upon adaptation to the BV^M sequence. It remains to be determined whether any of the observed strain alterations also occur when prions are propagated using BV^I.

Despite the adaptation of the RML strain, we found that anti-prion compounds active against mouse RML prions in cultured cells were also efficacious against the BV^M.RML strain. This finding suggests that although certain biochemical properties of RML are significantly altered upon adaptation to BVPrP, drug susceptibility remains consistent. However, this was not always the case, as we observed that quinacrine drastically reduced PrP^res levels in cells infected with 263 K prions but failed to show any efficacy against BV^M.263 K prions. Therefore, despite the relatively limited adaptation of hamster prions when propagated using BVPrP, subtle alterations to PrP^Sc can still alter susceptibility to anti-prion compounds. None of the three anti-prion drugs tested were effective against all four types of prions, and none of the drugs were able to reduce PrP^res levels in cells infected with BV^M.263 K prions. This highlights the importance of testing anti-prion drugs using a panel of prion strains in cultured cells to identify those that are active against a wide range of strains. Such compounds may be more likely to be efficacious against human prions. Given that BVPrP-adapted strains appear to be more resilient to treatment with anti-prion drugs, perhaps because of their increased conformational stability, screening putative drugs against BVPrP-adapted strains may represent a more rigorous test of anti-prion activity.

As BVPrP can serve as a permissive prion replication substrate in a variety of experimental paradigms, it was surprising that the G127V substitution completely blocked both cross-species and species-matched prion replication in cultured cells. It remains conceivable that additional passaging or more sensitive detection techniques (i.e., RT-QuIC) could reveal low levels of prion infection in cells expressing BVPrP with the G127V substitution. However, previous studies in transgenic mice revealed that G127V is as effective as PrP ablation at preventing the replication of several human prion strains (Asante et al., 2015). Our findings suggest that the protective effect of G127V can be generalized to other prion strains, such as those from mice and hamsters. The G127V substitution has been shown to alter the structure of PrP^C, including in a region known to modulate cross-species prion replication (Hosszu et al., 2020), and there are conflicting reports regarding the effect of this substitution on PrP dimerization (Hosszu et al., 2020; Sangeetham et al., 2021; Zheng et al., 2018; Zhou et al., 2016). We found that the addition of G127V to recombinant BVPrP significantly interfered with its ability to polymerize into aggregates in vitro, which is consistent with previous studies utilizing human or mouse PrP (Hosszu et al., 2020; Huang et al., 2020; Sabareesan & Udgaonkar, 2017). We speculate that the G127V substitution may interfere with the conformational flexibility of BVPrP^C, preventing it from adopting structures or forming assemblies that enable cross-species prion replication. Alternatively, G127V may impede the formation of a stable PrP^Sc structure. While exploiting the protective nature of the G127V polymorphism for therapeutic purposes by performing gene editing in human brains is likely to be challenging, modifying PrP so that it is unable to be converted into PrP^Sc is nonetheless an attractive strategy for preventing or delaying prion replication. Since G127V blocks the replication of many different types of prions, even when in the context of a highly prion-permissive substrate such as BVPrP, it represents an ideal theoretical approach for treating a wide variety of prion diseases that could be used in conjunction with or as an alternative to strategies that lower PrP^C levels (Minikel et al., 2020).

Abbreviations:

BV

bank vole

DY

drowsy prion strain

GdnHCl

guanidine hydrochloride

GPI

glycosylphosphatidylinositol

Ha

hamster

HY

Hyper prion strain

Mo

mouse

PrP

prion protein

PrP^res

proteinase K-resistant prion protein

rec

recombinant

RT-QuIC

real-time quaking induced conversion

ThT

Thioflavin T

DATA AVAILABILITY STATEMENT

All data generated or analyzed during this study are included in the article.