Protease-resistant and Detergent-insoluble Prion Protein Is Not Necessarily Associated with Prion Infectivity*

PrPSc, an abnormal isoform of PrPC, is the only known component of the prion, an agent causing fatal neurodegenerative disorders such as bovine spongiform encephalopathy (BSE) and Creutzfeldt-Jakob disease (CJD). It has been postulated that prion diseases propagate by the conversion of detergent-soluble and protease-sensitive PrPC molecules into protease-resistant and insoluble PrPSc molecules by a mechanism in which PrPSc serves as a template. We show here that the chemical chaperone dimethyl sulfoxide (Me2SO) can partially inhibit the aggregation of either PrPSc or that of its protease-resistant core PrP27–30. Following Me2SO removal by methanol precipitation, solubilized PrP27–30 molecules aggregated into small and amorphous structures that did not resemble the rod configuration observed when scrapie brain membranes were extracted with Sarkosyl and digested with proteinase K. Interestingly, aggregates derived from Me2SO-solubilized PrP27–30 presented less than 1% of the prion infectivity obtained when the same amount of PrP27–30 in rods was inoculated into hamsters. These results suggest that the conversion of PrPC into protease-resistant and detergent-insoluble PrP molecules is not the only crucial step in prion replication. Whether an additional requirement is the aggregation of newly formed proteinase K-resistant PrP molecules into uniquely structured aggregates remains to be established.

PrP Sc , an abnormal isoform of PrP C (1), is the only known component of the prion, an agent causing fatal neurodegenerative disorders such as BSE 1 and CJD (2). It has been postulated that prion diseases propagate by the conversion of PrP C molecules into protease-resistant and -insoluble PrP Sc molecules by a mechanism in which PrP Sc serves as a template (3). The pathway for PrP Sc synthesis may feature the formation of PrP C -PrP Sc heterodimers (4). Alternatively, the nucleation-dependent protein polymerization model argues that the formation of new PrP Sc molecules depends on the presence of a seed composed of aggregated PrP Sc molecules and that new PrP Sc molecules join previously assembled prion polymers (5). Although many lines of evidence suggest that PrP Sc is the crucial and even the only prion component, until today infectivity could not be associated with PrP Sc like PrP molecules produced by an array of in vitro conversion protocols (6 -8).
The organic solvent dimethyl sulfoxide (Me 2 SO) was shown to block the formation of amyloid fibrils by A␤ peptide in vitro (9). After a single dose of Me 2 SO, the urine of human amyloidotic patients contained fibrils with the tinctorial properties of amyloids, suggesting that Me 2 SO can either break large amyloid fibrils or inhibit their formation, resulting in smaller structures that can be mobilized from the connective tissue and eliminated by the kidneys (10). Me 2 SO was also shown to inhibit the accumulation of PrP Sc in scrapie-infected neuroblastoma cells (11), suggesting that Me 2 SO, in its function as a "chemical chaperone," stabilized the conformation of PrP C molecules, thereby preventing them from undergoing the conformational changes required for the conversion of PrP C to PrP Sc . In this work, we investigated whether the hallmark properties of PrP Sc , i.e. resistance to proteases and insolubility in detergents, are affected by in vitro treatment with Me 2 SO. These biochemical properties of PrP have been traditionally linked to the presence of prion infectivity (12), although in some experimental setups, protease-resistant PrP could not be found in samples that contain prion infectivity (13)(14)(15). Interestingly, it was shown lately that in prion strains with long incubation times, PrP Sc is considerably less resistant to proteases than in short incubation time strains (16).
Our results show that when membranes prepared from brains of hamsters terminally ill with scrapie were incubated in the presence of Me 2 SO and detergents, as opposed to detergent only, part of the PK-resistant PrP molecules could neither be precipitated by high speed centrifugation nor did they aggregate into very large structures. Me 2 SO, although it can inhibit the aggregation of protease-resistant PrP molecules, could not solubilize previously aggregated PrP Sc . These soluble PrP27-30 molecules will aggregate upon the removal of Me 2 SO, albeit not to the characteristic rod structure obtained when scrapie brain membranes are extracted with Sarkosyl and digested with proteinase K (PK) (17). When inoculated into hamster brains, only traces of scrapie infectivity were associated with this prion-specific isoform.

EXPERIMENTAL PROCEDURES
Sucrose Gradients-Three hundred microliters of 10, 15, 20, 25, 30, and 60% sucrose in phosphate-buffered saline were loaded into TLS-55 ultracentrifuge tubes (Beckman Instruments) to form a zonal gradient. Microsomes from brains of scrapie-infected hamsters (60 -80 l containing about 15 g/ml protein) were diluted with STE buffer (100 mM NaCl, 10 mM Tris, pH 7.4, 1 mM EDTA) to 240 l. Sarkosyl (2%) and when appropriate Me 2 SO (10%) were added to the mixture and incubated for 16 h at 4°C before loading on top of the gradient and centrifuged at 100,000 ϫ g for 1 h at 20°C. After the centrifugation, gradient fractions of equal volume were collected and immunoblotted with anti-PrP mAb 3F4.
In Vivo Infectivity Experiments-Top and bottom fractions from sucrose gradients (after PK digestion) with and without Me 2 SO were precipitated by 4 volumes of methanol (to discard the Me 2 SO) and resuspended in 100 l of saline, 10% bovine serum albumin. 10 l of each sample was serially diluted in STE (3 ϫ 10), and all samples were immunoblotted with mAb 3F4. After comparing the protease-resistant PrP signal in control microsomes, bottom fractions from sucrose gradients with and without Me 2 SO, as well as top fractions of the Me 2 SO gradient, the top samples were diluted 10 times and the rest 100 times to produce solutions with similar concentrations of PrP27-30. Fourweek-old male Syrian hamsters were inoculated intracerebrally with samples to be tested for prion infectivity (50 l). Each sample was inoculated into five hamsters, and all experiments were performed in duplicates. The hamsters were tested daily. Prion titers were measured by monitoring the incubation period until the appearance of symptoms (18).
Cross-linking by DSS-Disuccinimidyl suberate, an N-hydroxy succinimide ester homobifunctional cross-linker (reacting with NH 2 groups), was dissolved in Me 2 SO (to 1 M) and subsequently diluted into double-distilled water to 500 M. Samples of sucrose gradient fractions were incubated with DSS at a final concentration of 125 M DSS (30 min, room temperature). Following the incubation, the reaction was terminated by the addition of 1 M Tris. The samples were precipitated by methanol and immunoblotted with mAb 3F4.
Histoblots-Histoblots were carried out as described by Taraboulos et al. (19). Shortly, glass slides carrying 8-m thick cryostat sections were quickly thawed and immediately pressed onto nitrocellulose membrane saturated with lysis buffer. The membranes were thoroughly air-dried, rehydrated for 1 h in TBST (10 mM Tris, pH 8, 100 mM NaCl, 1.5% Tween 20), and then subjected to limited proteolysis in digestion buffer containing 40 g/ml PK for 1 h at 37°C, followed by incubation of the blots in 3 M guanidine thiocyanate/10 mM Tris-HCl, pH 7.8. Subsequently, the blots were processed as for immunoblotting.

RESULTS
10-l microsomes (20) (15 g/ml protein) prepared from the brains of Syrian hamsters infected with experimental scrapie 263K (21) were diluted to 100 l in STE buffer (100 mM NaCl, 10 mM Tris, pH 7.4, 1 mM EDTA) and incubated with 2% sodium sarcosinate (Sarkosyl) in the presence or absence of 10% Me 2 SO at 4°C for 16 h before centrifugation for 1 h at 100,000 ϫ g. Me 2 SO was removed from the supernatants by methanol precipitation, and pellets were rinsed once with 70% methanol. Ethanol or methanol precipitation are established methods to precipitate PrP Sc and prions for infectivity assays as well as for biochemical manipulations (22)(23)(24)(25). All samples were resuspended in STE with 2% Sarkosyl, incubated with proteinase K (40 g/ml, 1 h, 37°C), and then analyzed by SDS-PAGE followed by immunoblotting with ␣PrP mAb 3F4 (26). Proteinase K completely digests PrP C and concomitantly converts PrP Sc into PrP27-30(1). In the presence of Me 2 SO (Fig. 1, lane 1), a considerable part of the protease-resistant PrP remained in the supernatant; otherwise almost all of it pelleted in the high speed spin (lane 3). Me 2 SO was almost ineffective in solubilizing PrP Sc if added 2 h after the detergent, when most PrP Sc molecules were already aggregated (compare lanes 2 and 4). This suggests that although Me 2 SO can inhibit PrP Sc aggregation, it does not solubilize previously aggregated PrP Sc . Me 2 SO was also unable to solubilize purified PrP27-30 or PrP Sc 33-35, which are already aggregated (not shown).
To test whether the results obtained in Fig. 1 resulted from the in vitro conversion of PrP C to a protease-resistant species by the Me 2 SO treatment or whether traces of Me 2 SO inhibit the activity of PK, microsomes from scrapie-infected hamster brains were pretreated with PK before the incubation in the presence or absence of Me 2 SO. The effect of Me 2 SO on PrP Sc aggregation was identical regardless of whether Me 2 SO was applied before or after digestion with PK ( Fig. 2). We conclude therefore that Me 2 SO does not confer protease resistance to otherwise PK-sensitive PrP molecules but rather inhibits the aggregation of PrP Sc molecules. Whether Me 2 SO just reveals and amplifies a pre-existing difference between two populations of PrP Sc molecules (such as preaggregation) or actively partitions PrP Sc molecules into two distinct populations remains to be established.
To estimate the degree of aggregation of Me 2 SO-solubilized PrP Sc , normal and scrapie-infected hamster brain microsomes were incubated for 16 h in Sarkosyl in the presence or absence of Me 2 SO, and the resulting lysates were sedimented through 10 -60% sucrose gradients containing 2% Sarkosyl. Molecules migrating to the bottom of such gradients are either in very large aggregates or of very large molecular weight, and detergent-soluble proteins of small size are expected to remain in the upper gradient fractions. As can be seen in Fig. 3, PrP C from normal microsomes, which solubilizes readily in the presence of detergents (20), remained in the upper fractions of the gradient in the presence or absence of Me 2 SO. In addition, the incubation of PrP C with Me 2 SO treatment did not convert the normal prion isoform into protease-resistant PrP (Fig. 3, a-d).
Microsomes from scrapie-infected brains, which are believed to contain both PrP C and PrP Sc (20), yielded a bimodal PrP distribution in the absence of Me 2 SO; about half of the PrP (presumably PrP C ) was found in the top of the gradient, whereas the rest migrated to the bottom (Fig. 3e). However, only the PrP at the bottom of the gradient resisted the action of proteinase K, as expected from PrP Sc (Fig. 3f).
When Me 2 SO was present during the lysis of scrapie microsomes, a profound change in the distribution of protease-resistant PrP was observed (Fig. 3h). A considerable portion of the PrP molecules at the top of the gradient resisted proteinase K-catalyzed proteolysis, which instead produced PrP27-30, the protease-resistant core of PrP Sc . Thus, the action of Me 2 SO on scrapie microsomes yielded protease-resistant, prion-specific Membranes from scrapie-infected brains were incubated with 2% Sarkosyl in the presence or absence of 10% Me 2 SO at 4°C. Following the incubation, all samples were digested with 40 g/ml PK for 1 h at 37°C and centrifuged at 100,000 x g. Pellets and supernatants (sup) were tested for the presence of PK-resistant PrP by immunoblotting with ␣PrP mAb 3F4. Lane 1, incubation with Sarkosyl and Me 2 SO for 16 h. Lane 2, Me 2 SO was added 2 h after Sarkosyl, and the sample was incubated for 14 additional hours. Lane 3, incubation with Sarkosyl alone for 16 h. Lane 4, Sarkosyl was added to this sample 2 h before the end of the incubation.

FIG. 2. Me 2 SO (DMSO) inhibits the aggregation of PrP27-30 molecules.
Membranes from scrapie-infected hamsters were incubated with Sarkosyl and in the presence (lanes b and d) and absence (lanes a and c) of Me 2 SO (as in Fig. 1). The samples presented in lanes a and b were digested for 1 h with 40 g/ml PK before the addition of Me 2 SO and Sarkosyl. In lanes c and d, samples were digested with PK after Me 2 SO treatment (as in Fig. 1). After Me 2 SO incubation and PK treatment (or vice versa), all samples were centrifuged and immunoblotted as in Fig. 1. Lanes a-d represent the supernatants of all samples after the centrifugation.
PrP structures that are contained in low degree oligomers.
To reinforce the conclusion that the Me 2 SO-soluble PrP27-30 is the protease-resistant core of PrP Sc and not a partially resistant PrP C promoted either by the Me 2 SO incubation or by methanol precipitation, the three lightest fractions from sucrose gradients c and g, respectively, in Fig. 3 were combined, precipitated with methanol, resuspended in Sarkosyl, and subsequently digested with low PK concentrations for short periods. It has been shown lately that under such mild conditions of PK digestion, PrP C reveals a PK-resistant core of a lower M r than PrP27-30, since the 3F4 epitope (residues 108 -111) is absent from this peptide (27). These PK-digested samples were immunoblotted either with mAb 3F4 or with mAb 13A5, which reacts with residue 138 of hamster PrP. As can be seen in Fig. 4, while in the protease-resistant core of Me 2 SO-treated PrP in the scrapie fractions, both epitopes were present after 1 h PK digestion, and this was not the case for PrP from normal brain. These results show that Me 2 SO-solubilized PrP from scrapie brains was indeed PrP Sc which, unlike PrP C , produced upon mild or harsh PK digestion a proteaseresistant core of the same molecular weight as aggregated PrP Sc (see also Figs. 3 and 8). In addition, these results also show that methanol precipitation does not confer or reduce protease resistance.
To investigate the degree of oligomerization of soluble PrP27-30 in Fig. 3, PK-digested Me 2 SO-treated scrapie microsomes were resolved on a sucrose gradient as described above, and the diverse gradient fractions were subsequently crosslinked by DSS. The rationale of this experiment is that although monomeric PrP Sc will not cross-link, aggregated PrP Sc will cross-link heavily and thereby not enter the polyacrylamide gel. Indeed, as can be seen in Fig. 5, while treatment of the heavy fractions from the Me 2 SO gradient with DSS resulted in a reduction in the PrP27-30 band seen by 3F4 immu-noblotting, PrP27-30 in the light fractions was resistant to cross-linking.
To test whether Me 2 SO-solubilized PrP Sc molecules will either aggregate or remain soluble following the removal of Me 2 SO, fractions from the top and from the bottom of sucrose gradients were incubated with PK and then precipitated with methanol. Me 2 SO is very soluble in this alcohol. Methanol pellets of top and of bottom fractions were resuspended in 2% Sarkosyl, and suspensions were either incubated overnight at 4°C or mixed prior to incubation. Suspensions were then resolved again in the same kind of sucrose gradient. Most of the soluble protease-resistant PrP27-30 (from the top fractions) sedimented to the bottom of the gradient under these conditions (Fig. 6). These results show that the Me 2 SO-solubilized PrP retained its propensity to aggregate upon Me 2 SO removal. When, instead of removal by methanol precipitation, Me 2 SO was dialyzed from the light fractions at 4°C against a Sarkosyl containing buffer, only part of the PrP27-30 could be recovered from the dialysis tube. Aggregated PrP Sc or PrP27-30 molecules are known for their nonspecific adhesion properties (28). Even under these conditions, only a fraction of the recoverable protease-resistant PrP was present in the light fractions of the new sucrose gradient (Fig. 6d), suggesting the reaggregation of PrP27-30 upon Me 2 SO removal precipitation is an intrinsic property of Me 2 SO-solubilized PrP Sc .
When the aggregates produced by Me 2 SO-solubilized PrP27-30 were looked upon by electron microscopy and compared with those in the original heavy fractions, a profound difference in structure was observed. Although PrP27-30 from the heavy fractions aggregated into the familiar rod-like structure (29), the new aggregates were amorphous (Fig. 7).
PK-digested heavy and light fractions of sucrose/Sarkosyl

FIG. 3. Me 2 SO-solubilized PrP Sc presents as a low density oligomer.
Membranes from normal or from scrapie-infected brains were incubated with 2% Sarkosyl in the presence and absence of 10% Me 2 SO and were subjected to a size-separating sucrose gradient as described under "Experimental Procedures." Fractions collected from the gradients were digested in the presence and absence of PK and immunoblotted with mAb 3F4. a, normal hamster brain membranes. b, normal hamster brain membranes digested for 1 h with 40 g/ml PK. c, Me 2 SOtreated normal hamster brain. d, Me 2 SO-treated normal hamster brain membranes digested for 1 h with 40 g/ml PK. e, scrapie hamster brain membranes. f, scrapie hamster brain membranes digested for 1 h with 40 g/ml PK. g, Me 2 SO-treated scrapie hamster brain membranes. h, Me 2 SO-treated scrapie hamster brain membranes digested for 1 h with 40 g/ml PK.
FIG. 4. Me 2 SO-solubilized PK-resistant PrP is not derived from PrP C . Membranes from normal and scrapie-infected hamsters were incubated in the presence of Me 2 SO and subjected to sucrose gradients as described in Fig. 3 and under "Experimental Procedures." The three lightest fractions from the normal and the scrapie gradient were combined, precipitated by methanol (4 volumes), and subsequently resuspended in STE ϩ 2% Sarkosyl before PK digestion at 20 g/ml for 0, 5, 10, or 60 min and immunoblotting with either mAb 3F4 (a) or mAb 13A5 (b).

FIG. 5. Me 2 SO-solubilized PrP Sc does not cross-link.
Me 2 SOtreated scrapie-infected membranes were subjected to a sucrose gradient as in Fig. 3. Fractions from this gradient were incubated in the presence or absence of 125 M DSS for 30 min at room temperature. The reaction was stopped by the addition of 1 M glycine. Samples were subjected to immunoblotting with mAb 3F4. gradients with and without Me 2 SO were tested for prion infectivity. The three lightest and heaviest fractions (out of 12) of each gradient were combined, concentrated by methanol, and then resuspended in 10% bovine serum albumin in saline. The same protocol was applied on a control sample consisting of scrapie brain microsomes. Before inoculation, the control, both heavy fractions, as well as the light fraction of the Me 2 SO gradient were adjusted by dilution to contain comparable concentrations of PrP27-30 (as judged by immunoblotting with the mAb 3F4). The light fraction prepared in the absence of Me 2 SO, which did not contain visible PrP Sc , was diluted for inoculation identically to the Me 2 SO-treated light fraction. Parallel samples were re-digested with PK and were shown to be identically resistant to digestion as the ones inoculated into hamsters. All samples were inoculated intracerebrally, and the hamsters were monitored for clinical signs of scrapie. The results are summarized in Fig. 8. Regardless of the Me 2 SO treatment, most of the infectivity (Ͻ99%) was found in the heavy fractions (that contained rods), whereas only about 1% of the infectivity was associated with light fractions. Thus, Me 2 SO-soluble PrP27-30 was associated with very low, if any, prion infectivity. Interestingly, although the amount of protease-resistant PrP present in the light fractions prepared in the presence of Me 2 SO was at least 100 times higher than in the light fractions without Me 2 SO (as judged by immunoblotting), no significant difference in the infectivity associated with these fractions could be observed.
A possible explanation for the infectivity results presented above is that Me 2 SO produced a profound change in the conformation of PrP Sc , therefore generating a new prion strain (30,31). To test this possibility, we looked for evidence of strainspecific parameters in the hamsters that were inoculated with samples from the top and the bottom fractions of Me 2 SOtreated and untreated microsomes. Clinical signs observed in all groups were similar and were characteristic of the parent Sc263K strain (32). PrP27-30 banding (Fig. 9a) and PrP Sc histoblot patterns (Fig. 9b) (33)(34)(35) were also similar in all the four groups. However, since there is a residual infectivity in the light fractions of untreated microsomes, these results do not rule out the presence, in the Me 2 SO-treated samples, of a new prion strain with a much longer incubation time than that of the parent Sc263 strain. DISCUSSION Although there is little doubt that PrP Sc plays a crucial role in prion diseases (2), the mechanism by which PrP C converts FIG. 6. Me 2 SO-solubilized PrP becomes insoluble after Me 2 SO removal. Fractions from a sucrose gradient of Me 2 SO-treated scrapie brain hamsters (a) were either precipitated with methanol (4 volumes) and resuspended in 2% Sarkosyl for 2 h or dialyzed against STE ϩ 2% Sarkosyl overnight. Light and heavy fractions by themselves or combined with each other were subjected to an additional sucrose gradient. b, heavy fraction after methanol precipitation. c, light fraction after Me 2 SO removal by precipitation. d, light fractions after Me 2 SO removal by dialysis. e, combined heavy and light fractions after Me 2 SO removal by precipitation.

FIG. 7. Me 2 SO-solubilized PrP Sc aggregates into non-rod structures.
Fractions from a sucrose gradient loaded with Me 2 SO-treated scrapie brain membranes were immunostained with mAb 3F4 and protein A gold (10 nM) and negatively stained with uranyl acetate and examined under an electron microscope. All samples were digested with 40 g/ml PK. a, gradient heavy (bottom) fractions (bar, 100 nm). b, light fractions after Me 2 SO removal (bar, 100 nm). c, sample b at a higher magnification (bar, 10 nm). into the scrapie isoform of the prion protein, as well as the specific role of the PrP isoforms before and during the disease, is still unknown. A puzzling point to the prion hypothesis is the lack of correlation between the number of PrP Sc molecules in a sample and its infectivity as revealed by animal bioassays. In most cases, the infectious unit is associated with as many as 10 5 PrP Sc molecules (36). However, in some cases, prion infectivity was transmitted even in the apparent absence of protease-resistant PrP Sc molecules (13,37). One explanation for such an effect depends on the different biochemical characteristics of distinct prion strains. It has been shown that each prion strain can be associated with a characteristic proteaseresistant core which probably results from a defined PrP Sc conformation for each strain (33, 38 -40). It was also shown that the prion incubation time depends on the degree of protease resistance of PrP Sc in each such strain (16). Indeed, PrP and infectivity in some CJD stains shows very little protease resistance (15), whereas the hamster 263K produces the most PK-resistant PrP identified among prion strains (16).
It is more difficult to speculate on how samples that contain the same prion strain can present different incubation times for the same concentration of PrP Sc . One explanation for this effect is that disaggregated PrP Sc could display a higher specific infectivity, perhaps because more infectious centers would exist in such a sample or perhaps due to the increased surface area of such a disaggregated agent. For instance, dispersion of prion rods into DLPC resulted in increased prion titer, suggesting that smaller and more open aggregates may have increased access to the molecules in the brain which participate in the transmission of infectivity (41). As opposed to dispersion into DLPC, treatment of PrP Sc with the amyloid-binding dye Congo Red reduces prion titer by stabilizing the aggregated state of PrP Sc (42,43). Based on these facts, we assumed that samples containing Me 2 SO-solubilized PrP Sc would display very high specific prion titers. Surprisingly, this was not the case. The light samples in the gradients shown in Fig. 3, with and without Me 2 SO, presented the same low prion titer, even though the control sample had no apparent PrP Sc in contrast to the Me 2 SO-treated sample, which contains large quantities of the protease-resistant PrP isoform. It should be noted that the DLPC dispersion experiments were performed mostly on preaggregated rods, which as described above contain a priori all the infectivity. In addition, it is possible that the inoculation of PrP Sc in DLPC, although dispersed into monomers or small oligomers, represents a favorable pharmacological pathway to infect cells with prions. Me 2 SO-solubilized PrP Sc may not present such a biological advantage.
Our results show that prion-specific PrP molecules can be differentiated into two distinct species of disparate physicochemical properties: "classical" PrP Sc and Me 2 SO-soluble PrP Sc . This demonstrates that prion-specific PrP molecules can exist as a soluble species and yet possess the protease-resistant core PrP27-30. Apart from being resistant to proteases, these soluble molecules differ from PrP C in that they have not lost their propensity to aggregate (albeit to amorphous, non-rod structures) when Me 2 SO is removed. However, this species is not associated with prion infectivity.
We have also shown here that soluble PrP Sc can be dissociated from prion infectivity only when aggregation is inhibited during membrane extraction and not by dissociation from previously aggregated PrP Sc . This is the fundamental difference between the experiments presented here and other approaches which failed to show non-infectious protease-resistant PrP Sc (44). As shown above, non-infectious and infectious PrP Sc have similar aggregation properties, and therefore, once aggregated in the presence of detergent they are biochemically indistinguishable.
Me 2 SO-solubilized PrP Sc may be a metabolic intermediate in the formation of infectivity-associated PrP Sc . If so, then the process of prion replication may be composed of more than one irreversible step, the conversion of PrP C to a protease-resistant species being only one of them. In vitro converted PrP Sc may also be such an intermediate, since although presenting the biochemical properties of PrP Sc , it has not been shown to be infectious (7,8). Prion infectivity may only be associated with the final step, which would involve the specific aggregation of PrP Sc into a structure with the pharmacological properties required for the biology of the infectious process.
Although it is not impossible that Me 2 SO caused a profound change in the structure of part of the PrP Sc molecules, the fact that Me 2 SO-soluble PrP Sc could only be generated from microsomes and not from preformed rods is more consistent with the possibility of more than one prion-specific PrP Sc existing before the addition of Me 2 SO and the extraction of scrapie brain membranes with detergents. Regardless of the mechanism, our results show that not all PK-resistant PrP molecules are associated with prion infectivity.
Me 2 SO-soluble PrP Sc species, although not infectious, could still play an important role in the neuropathology of prion diseases. At the last stages of the disease, when the load of total PrP Sc molecules in the brain is large, non-infectious PrP Sc molecules may replace most of the PrP C molecules or otherwise inhibit their normal function. A large load of non-infectious PrP Sc molecules may also be neurotoxic or contribute to brain degeneration in as yet unknown mechanisms. FIG. 9. a, PrP Sc banding pattern in brains of hamsters inoculated with Me 2 SO-treated or untreated samples. Ten percent brain homogenate of hamsters dying of scrapie due to infection with the samples described in Fig. 7 were digested with 40 g/ml PK and immunoblotted with mAb 3F4. Lane C, brain inoculated with control scrapie microsomes. Lane 1, brain inoculated with light fraction of gradient without Me 2 SO. Lane 2, brain inoculated with light fraction of gradient with Me 2 SO. Lane 3, brain inoculated with heavy fraction of lane 1. Lane 4, brain inoculated with heavy fraction of lane 2. b, brain histoblots of hamsters inoculated with Me 2 SO-treated or untreated light and heavy gradient fractions. Hippocampal histoblots of Syrian hamster brains inoculated with light fraction of gradient without Me 2 SO (blot 1). Light fraction of gradient with Me 2 SO (blot 2). Heavy fraction of blot 1 (blot 3). Heavy fraction of blot 2 (blot 4).