Photodegradation illuminates the role of polyanions in prion infectivity

Abstract

Understanding the mechanism by which prion infectivity is encoded by the misfolded protein PrP^Sc remains a high priority within the prion field. Work from several groups has indicated cellular cofactors may be necessary to form infectious prions in vitro. The identity of endogenous prion conversion cofactors is currently unknown, but may include polyanions and/or lipid molecules. In a recent study, we manufactured infectious hamster prions containing purified PrP^Sc, co-purified lipid, and a synthetic photocleavable polyanion. The polyanion was incorporated into infectious PrP^Sc complexes, and then specifically degraded by exposure to ultraviolet light. Light-induced in situ degradation of the incorporated polyanion had no effect on the specific infectivity of the samples as determined by end-point dilution sPMCA and scrapie incubation time assays. Furthermore, prion strain properties were not changed by polyanion degradation, suggesting that intact polyanions are not required to maintain the infectious properties of hamster prions. Here, we review these results and discuss the potential roles cofactors might play in encoding prion infectivity and/or strain properties.

The prion diseases are infectious diseases that are believed to be caused by the conformational change of a host-encoded protein, PrP^C, to a pathogenic conformer PrP^Sc. The controversial “protein-only” hypothesis posits that the infectious agent is composed solely of the misfolded conformer PrP^Sc. There have been many attempts to create infectious prions from purified recombinant PrP protein. However, all of the samples generated in these experiments display relatively low levels of specific infectivity when inoculated intracerebrally into wild-type animals.1–4 Several lines of evidence suggest that cellular cofactors, such as polyanionic molecules, facilitate the formation of the infectious conformation.5–14

The first in vitro PrP conversion assay used radiolabeled PrP^C substrate purified from mammalian cells mixed with a stoichiometric excess of unlabeled PrP^Sc. This cell free assay produced a protease-resistant, radioactive product termed PrP-res.15 These pioneering studies showed for the first time that PrP could be specifically transformed in vitro, but the yield using purified substrates was low. Using a modification of the cell free assay in which crude brain homogenate replaced purified PrP^C as the substrate, our laboratory was able to amplify PrP^Sc 6-fold over input prion seed, suggesting that non-PrP constituents of crude brain homogenate might be required for efficient PrP^Sc formation in vitro.16 Using this system, we discovered that nuclease treatment of hamster brain homogenates abolished PrP^Sc amplification in vitro, and that reconstituting the nuclease-treated reactions with purified mammalian RNA rescued the amplification process.5 PrP^Sc amplification could also be obtained by adding certain synthetic homopolymeric nucleic acids to immunopurified PrP^C.6 Taken together, these surprising results argue that non-proteinaceous, host-encoded cofactors such as RNA molecules might facilitate prion conversion through a structural (as opposed to encoding) mechanism.8 The high efficiency of the serial protein misfolding amplification (sPMCA) technique developed by Soto and colleagues has allowed researchers to amplify prion infectivity as well as PrP^Sc molecules.17,18 Using sPMCA, we showed that infectious PrP^Sc molecules could be formed from immunopurified PrP^C, co-purified lipid and synthetic RNA molecules. Moreover, even unseeded reactions containing these defined components were capable of generating prions with high specific infectivity in a prion-free environment, showing for the first time that wild type infectious prions could be produced de novo.7

Additional studies in this purified system showed that PrP^C molecules undergo a time-dependent conformational change upon interaction with RNA. When this change occurs, PrP^C adopts an intermediate conformation that mimics some of the characteristics of PrP^Sc, such as detergent insolubility and reactivity to PrP^Sc-specific antibodies, but remains sensitive to proteinase K digestion.8 When incubated with a heterogeneous size mixture of homopolymeric [³²P] poly(A) molecules during PMCA, hamster PrP^C molecules incorporated a specific size subset (1–2.5 kb) of the RNA molecules into nuclease-resistant complexes. The physical interaction between RNA and PrP^Sc was confirmed by fluorescence microscopy experiments showing that fluorescein-labeled RNA molecules became integrated into nuclease-resistant complexes with PrP^Sc molecules. Interestingly, neuropathologic analysis of scrapie-infected hamsters revealed that endogenous RNA molecules stained with acridine orange co-localized with large extracellular PrP aggregates.8 Taken together, these studies suggest that PrP interacts specifically with polyanionic molecules in vitro and in situ, and raised the possibility that polyanions might be a necessary component of infectious prions.

Jeong et al. investigated whether endogenous RNA molecules might be required for prion infectivity by treating scrapie brain homogenates with LiAlH4 (lithium aluminum hydride), a strong reducing agent that can cleave the phosphodiester bond in RNA molecules.19 Interestingly, treatment of hamster scrapie brain homogenates with LiAlH₄ caused an ∼3-fold increase in scrapie incubation period measured by bioassay, suggesting that RNA may be an important component of infectious prions and therefore may play a role in stabilizing PrP^Sc structure. However, LiAlH₄ is not a specific reagent, and can damage a variety of other macromolecules, including proteins. Therefore, the decrease in infectivity measured in this study cannot be specifically ascribed to degradation of the polyanion.19

We recently reinvestigated the potential role of polyanion in maintaining prion infectivity by using a more targeted approach.20 Specifically, we utilized a synthetic oligonucleotide that could be selectively hydrolyzed by treatment with ultraviolet (UV) light. The photocleavable oligonucleotide was synthesized by inserting a photocleavable linker in between every fives bases of a poly(dT) 100-mer. Exposure to UV light quantitatively converted the oligonucleotide into five base fragments. During incubation with excess recombinant PrP, the photocleavable oligonucleotide became incorporated into a nuclease-resistant nucleoprotein complex, but remained sensitive to photocleavage. This novel system allowed us to study the role of a polyanion molecule incorporated into infectious prions in situ (Fig. 1).

We used PMCA to create PrP^Sc molecules that contained either the photocleavable oligonucleotide or a non-photocleavable control analog. After treatment with UV light, the infectivity of each sample was measured using a combination of end-point dilution sPMCA and animal bioassays. The end-point dilution PMCA assay showed a ∼1 log decrease in the seeding ability of PrP^Sc samples treated with UV light, but this effect was not specific since a similar decrease was measured in samples containing the control nucleic acid. In the bioassay, there was no change in the incubation periods of animals inoculated with PrP^Sc samples treated either in the presence or absence of UV light. Neuropathological analysis of inoculated animals also showed no differences in neurotropism between the two groups. Degradation of the nucleic acid had no effect on the molecular migration or structural stability of PrP^Sc samples as determined by SDS-PAGE and urea denaturation assays, respectively. There were also no differences in the molecular migration or glycosylation profile of the PrP^Sc molecules produced in the brains of animals inoculated with light- versus mock-treated inocula, and urea denaturation assays showed no differences in PrP^Sc stability. These results collectively demonstrate that the presence of intact polyanion molecules is not required to maintain the infectious, biochemical or strain properties prions generated in vitro.

These results are consistent with the stringent “protein-only” hypothesis, but do not yet provide definitive proof. The purified PrP^C molecules used as substrate in these experiments contain a stoichiometric amount of co-purified lipid7 that may play a role in the generation of prion infectivity.9 Also, although the efficacy of photocleavage conditions was carefully confirmed in control reactions, it is possible that some intact oligonucleotide survived UV treatment at a level below detection. Alternatively, the remnant five base nucleic acid fragments may remain incorporated within the PrP^Sc molecule and play a role in maintaining the infectious conformation. Even in this scenario, our results would place a significant geometric constraint on the role of incorporated polyanion. While polyanions ≥40 bases facilitate the formation infectious prions in vitro,8 our results suggest that polyanions >5 bases are not necessary to maintain the infectious properties the prion. The exact role polyanions play in prion formation is still unclear, but it is tempting to speculate that they may serve as scaffolds that facilitate prion conversion by (a) bringing PrP^C and PrP^Sc seed together for templating to occur or (b) acting as a catalyst which is necessary to reduce the activation energy of refolding to the PrP^Sc form. Future studies will need to be performed to differentiate between these two hypotheses. It is also possible that polyanions are completely dispensable for maintaining PrP^Sc structure, and it is the co-purified lipid molecules that serve this role instead. Consistent with this possibility, we recently discovered that mouse PrP^Sc can be serially propagated in vitro in the absence of nucleic acids.21 Finally, it is possible that either polyanions or lipids can function equally well as stabilizers of the infectious PrP^Sc conformation. More work is required to distinguish between these possibilities.

Generating high levels of specific infectivity solely using purified recombinant PrP remains the ultimate proof of the “protein-only” hypothesis. To date, evidence suggests that cellular cofactors are necessary to create infectious prions but may or may not be required to maintain infectivity once formed. Significantly, Wang et al. showed that bona fide prions could be formed from recombinant PrP, synthetic lipid and RNA molecules.9 Although no completely pure preparations of misfolded PrP possessing significant levels of specific infectivity have yet been produced, it should eventually be possible to produce such a preparation if the “protein-only” hypothesis is correct. On the other hand, a rigorous refutation of the hypothesis would require demonstrating that PrP^Sc and infectivity can be dissociated.

Figures and Tables

Figure 1 Selective photodegradation of an incorporated polyanion in situ.