For determination of PrPSc, cell homogenates were digested with 5 g mL-1, COCS and C57BL/6 whole brain homogenates were digested with 25 g mL-1 PK (Roche) at a final volume of 20 L in PBS for 30 minutes at 37C. protein PrPC . The aggregated form, denoted PrPSc, is typically resistant to limited digestion with proteinase K (PK). The pathology brought on by prion infections, consisting of spongiosis, neuronal loss, astrogliosis, and microglial activation, is usually faithfully reproduced by administration of anti-prion antibodies targeting conformational epitopes around the globular domain name (GD) of PrPC [2, 3]. Toxicity requires the long flexible tail (FT) of PrPC, and antibodies against the octapeptide repeat (OR) domain name of the FT prevent the toxicity of anti-GD antibodies and antagonize neurodegeneration in prion infections . Therapeutic compounds conferring anti-prion protection are frequently effective also against toxic anti-prion antibodies, suggesting that GD antibodies and prions share common effector pathways . The striking similarities between the consequences of toxic anti-GD antibodies and of prion infections raised the question whether such antibodies might induce the generation of prions. By distorting the conformation of PrPC, antibodies may conceivably catalyze the formation of higher-order aggregates that would, in turn, act as nucleation sites for the growth of PrPSc fibers . This question is not only of academic importance, but it may also be of relevance to the biosafety classification of research with such antibodies. We therefore undertook to clarify T16Ainh-A01 whether POM1 induced infectious prions, and if so, whether this might explain its toxicity. We treated COCS homogenates, which have comparable prion propagation efficacies as whole brain homogenates , with the toxic anti-prion antibody POM1 and analyzed them for the presence of prions after passaging into prion-susceptible cells and PrPC-overexpressing mice . Results In order to minimize any possible effector functions and off-target effects of the antibodies, such as complement and Fc-binding, we generated single-chain variable fragments (scFv) of the neurotoxic anti-PrP antibody POM1 (scFvPOM1). PrPC-overexpressing COCS were then treated with either scFvPOM1 (400 nM) or with scFvPOM1 (400 nM) preincubated with a molar excess of recPrP23-230 (600 nM) for control. This treatment was maintained over 10 days with three medium changes per week; scFvPOM1 was replaced with each medium change. NeuN immunofluorescent stainings, which identify neurons, showed widespread neuronal degeneration in COCS treated with scFvPOM1, but not in COCS treated with antibody that had been preemptively blocked with recPrP23-230 (Fig 1A). To clarify whether this effect was induced by the aggregation of PrP, we analyzed pooled COCS homogenates treated with either scFvPOM1 (n = 8) or scFvPOM1 + recPrP23-230 (n = 5) for PrPSc, an isoform of PrP that is partially resistant to proteinase K (PK) and is universally employed as a surrogate marker for prion infectivity (Figs ?(Figs1B1B and S1) . Pooled slice homogenates from scFvPOM1-treated (n = 8) and (scFvPOM1+recPrP23-230))-treated (n = 5) COCS were analyzed by Western blotting and were found to be devoid of PrPSc. In contrast, PK digestion of prion-containing brain homogenate (RML6 = passage 6 of the Rocky Mountain Laboratory strain mouse-adapted scrapie prions), which served as positive control, showed the typical diagnostic shift towards a smaller PK-resistant core with un-, mono- and diglycosylated PrPSc. Open in a separate windows Fig 1 (A) Chronic treatment of COCS with scFvPOM1 induced profound neurodegeneration. Instead, no neurodegeneration was observed in COCS exposed to scFvPOM1 pre-incubated with recPrP23-230. *** p T16Ainh-A01 0.001. Scale bar = 500 m. (B) Pooled slice homogenates of scFvPOM1-treated (n = 8 slices) or (scFvPOM1+ recPrP23-230)-treated COCS (n = 5 slices) did not show PK-resistant species after digestion as is typically observed in RML6 brain homogenate (n = 1). (C) No PrPSc was observed after inoculation of the highly PrPSc-susceptible cell line CAD5 with pooled scFvPOM1-treated COCS homogenates. CAD5 cells were successfully infected with RML6 as shown by the typical diagnostic shift T16Ainh-A01 of T16Ainh-A01 TSPAN10 PK-digested RML6 with un-, mono- and diglycosylated PrPSc bands. RML6 brain homogenate served as a positive control (left band pair). The murine neuroblastoma cell line CAD5 is highly susceptible to prion contamination and serves as a sensitive bioassay for prion propagation . We hence spiked cell culture media of exponentially growing CAD5.