Prions efficiently cross the intestinal barrier after oral administration: Study of the bioavailability, and cellular and tissue distribution in vivo

Natural forms of prion diseases frequently originate by oral (p.o.) infection. However, quantitative information on the gastro-intestinal (GI) absorption of prions (i.e. the bioavailability and subsequent biodistribution) is mostly unknown. The main goal of this study was to evaluate the fate of prions after oral administration, using highly purified radiolabeled PrPSc. The results showed a bi-phasic reduction of PrPSc with time in the GI, except for the ileum and colon which showed sustained increases peaking at 3–6 hr, respectively. Plasma and whole blood 125I-PrPSc reached maximal levels by 30 min and 3 hr, respectively, and blood levels were constantly higher than plasma. Upon crossing the GI-tract 125I-PrPSc became associated to blood cells, suggesting that binding to cells decreased the biological clearance of the agent. Size-exclusion chromatography revealed that oligomeric 125I-PrPSc were transported from the intestinal tract, and protein misfolding cyclic amplification showed that PrPSc in organs and blood retained the typical prion self-replicating ability. Pharmacokinetic analysis found the oral bioavailability of 125I-PrPSc to be 33.6%. Interestingly, 125I-PrPSc reached the brain in a quantity equivalent to the minimum amount needed to initiate prion disease. Our findings provide a comprehensive and quantitative study of the fate of prions upon oral infection.

Most of the cases of scrapie, CWD, BSE and vCJD arise from oral exposure through contaminated food and environmental materials. The oral route of transmission was demonstrated by infectivity studies with experimental animals deliberately exposed to prions by mouth 5 . After oral inoculation, the agent accumulates in the gut-associated lymphoid tissues such as the Peyer's patches and mesenteric lymph nodes before neuroinvasion [5][6][7][8][9] . PrP Sc is absorbed across the intestinal epithelium mediated through several mechanisms. Studies in Caco-2 cellular models of the intestinal barrier have shown that PrP Sc can cross the barrier by endocytosis upon interaction with the laminin receptor 10 . On the other hand, Heppner and colleagues found that M cells mediate transcytosis of prions across the epithelial cell monolayers 11 . Mishra et al. showed that prion-ferritin complex is transcytosed in Caco-2 cell monolayers 12 . Kujala et al. found that the absorption of PrP Sc through enteric lymphatic tissues was independent from constitutive expression of PrP C 13 .
Regardless of the specific transcellular transport routes, once PrP Sc crosses the intestinal barrier from the apical (mucosal) surface to the basolateral (serosal) side, PrP Sc could be immediately absorbed by both dendritic cells and serousal fluid associated with mesenteric blood flow. Based on studies of infectivity and histological staining, it is generally assumed that the vasculature does not play a major role on the transport of PrP Sc to the target organs, but instead that transmission involves direct uptake by peripheral nerves located in the oral or gastrointestinal mucosa 5,7,14 . However, previous experiments in which isolated gut loop were inoculated with prions, showed appearance of PrP Sc rapidly in sub-mucosal lymphatics, much before detection in peyer's patches of the intestine 15 . Since no evidence was obtained for the inoculum being transported across the dome to the peyer's patches, the authors conclude that the most likely source of scrapie infection of lymphoid cells was blood. Moreover, since the pattern of PrP Sc deposition does not change when the agent is administered by different routes and because early PrP Sc accumulation was observed in circumventricular organs of sheep, it was proposed that the infectious agent may reach directly the brain from the blood through these structures where blood-brain barrier (BBB) is less tight 16 . Furthermore, a recent report from Denkers et al. 17 , showed that minor lesions in the oral cavity significantly facilitate CWD infection, suggesting that direct contact with blood circulation may play a role in prion spreading. In addition, the fact that prions acquired by oral exposure can be detected in various tissues, (including skeletal muscle) even before the onset of the disease 18 , is better explained by a hematogenous distribution of the infectious agent 19 . It is important to note that studies based on infectivity or conventional immuno-histochemical approaches do not have the sensitivity sufficient to elucidate the initial fate of PrP Sc within minutes or hours after exposure; indeed, these methodologies usually rely on detection of prions after peripheral replication. Thus, no direct study has been done to assess how much of PrP Sc given orally gets into the blood and what the fate is of ingested prions. In addition, it is unknown what percentage of PrP Sc survives gut metabolism and is taken up from the intestinal lumen (i.e. the bioavailability of PrP Sc ) and how the protein is then distributed in the body and reaches the brain.
Given the enormous social concerns, economic impact and public health consequences of prion transmission by infection and the unprecedented features of this protein-only infectious agent, it is highly surprising that this very important aspect of the disease pathogenesis has been almost completely neglected. Until now, the field has no quantitative information regarding the oral bioavailability, metabolism, and whole-body distribution profile of PrP Sc . The main goal of the present study was to analyze in a detailed and quantitative manner, the initial fate of prions upon oral exposure, including the estimation of oral bioavailability, spatio-temporal distributions throughout the GI tract, interaction with blood cells, uptake by peripheral organs and brain, thereby providing invaluable information to understand the pathogenesis of prion diseases which is important both for basic science and for risk assessment.

Results
Purity and size distribution of 125 I-PrP Sc . To study the in vivo fate of PrP Sc after oral administration, the protein was highly purified from the brain of mice infected with the RML prion strain, using a previously described procedure 20 . PrP Sc purity was estimated to be > 95% by silver staining and western blot (Fig. S1A). Thereafter, purified PrP Sc was labeled with 125 I, as previously described 20 . After radiolabeling, the molecular weight of 125 I-PrP Sc was analyzed by size-exclusion radio-HPLC chromatography (Fig. S1B). All the radioactivity measures after fractionation were normalized by TCA precipitation assay to separate free iodine and confirm that the peaks seen in the radio-chromatogram are protein associated. 125 I-PrP Sc appeared as one main speak with an estimated molecular weight of ~310 KDa and a minor peak of ~450 KDa (Fig. S1B), consisting with the expected size of the most infectious PrP Sc species 21 . It is likely that larger fibrillar aggregates were removed during purification or labeling procedures. As previously reported, radiolabeling by iodination does not change PrP Sc properties, including its resistance to proteolysis, biophysical stability against chemical denaturants and ability to self-propagate into infectious PrP Sc material 20 . In addition, pharmacokinetic parameters of radiolabeled PrP Sc were similar as those of full-length, unlabeled PrP Sc 20 . As such, we could take advantage of the sensitivity of radiolabeled PrP Sc to study the oral bioavailability and biodistribution in the present study.
Gastro-intestinal distribution of PrP Sc . Experimentally naïve mice fasted for 16 to 20 hr received a p.o. administration of purified radiolabeled 125 I-PrP Sc using oral gavages. Intact levels of 125 I-PrP Sc were measured by normalizing radioactivity recovered from various tissues by TCA precipitation to negate overestimation from free radioactive iodine detached from 125 I-PrP Sc . The results show a bi-phasic reduction in the levels of 125 I-PrP Sc in the stomach in fasted conditions (Fig. 1A). An initial phase of rapid reduction estimates the half-time of 1.29 hr followed by slower half-time of 25.9 hr. These findings suggest that, after relatively fast reduction of the majority of 125 I-PrP Sc , remaining portion of PrP Sc (around 3-4% of injected dose (ID)) was sequestered in the stomach and remained there even 24 hr after p.o. administration. The duodenum and jejunum showed similar patterns of 125 I-PrP Sc levels, which closely correlated with stomach over the study period (Fig. 1A). 125 I-PrP Sc in the duodenum showed a peak by 10 min after p.o. administration, and the levels decreased thereafter. The amount Scientific RepoRts | 6:32338 | DOI: 10.1038/srep32338 of 125 I-PrP Sc in this organ ranged from 18.7 to 0.3%ID/g. The jejunum showed similar retention of 125 I-PrP Sc to the duodenum, the maximal amount (43.4%ID/g) of 125 I-PrP Sc was found 30 min after the administration, thereafter the levels decreased with time (Fig. 1A). Interestingly, in the ileum where uptake of PrP Sc (mediated through Payer's patches and M cells) was reported to occur 11 , 125 I-PrP Sc was detected as fast as 10 min after p.o. administration, and the maximal levels (~18.4%ID/g) were reached by 1-3 hr followed by gradual decrease with time (Fig. 1B). The colon also showed small amount of 125 I-PrP Sc 10 min after administration, the levels peaked (~10%ID/g) at 6 hr, thereafter decreased to 2-3% ID/g by 24 hr. These findings suggest that fast distribution of 125 I-PrP Sc occurs in the GI-tract after oral ingestion, and that small amount of material are retained throughout the GI-tract over 24 hr.
Whole blood retained higher levels of PrP Sc mediated by blood cell association. We quantified intact 125 I-PrP Sc in the whole blood and plasma after p.o. or i.v. administrations, and blood cell associations of 125 I-PrP Sc observed by both routes were compared (Fig. 2). After i.v. injection, the levels of 125 I-PrP Sc in the whole blood were very similar as those seen in plasma and indeed the curves were not significantly different ( Fig. 2A, inset). This result indicates that when PrP Sc is directly deposited in blood, a high proportion of the material remains in the plasma. In contrast, the levels of 125 I-PrP Sc in the whole blood were constantly higher than that in plasma after p.o. administration ( Fig. 2A), reaching statistical differences from 12 to 24 hr. Intact 125 I-PrP Sc in plasma peaked at 30 min (0.69%ID/ml) and the levels sustained by 6 hr, whereas in whole blood, intact 125 I-PrP Sc reached the maximal amount at 3 hr after p.o. administration.
The finding that the proportion of 125 I-PrP Sc in plasma and blood cells was different depending on the route of administration suggest that there is a large difference in blood cell association depending on the mode by which PrP Sc get access into the blood. This is most clearly observed when blood/plasma ratios for intact 125 I-PrP Sc after i.v. and p.o. administrations were compared (Fig. 2B). While the ratios remained less than 1 for all time points in i.v. injected animals, indicating that only ~14-48% of 125 I-PrP Sc was cell associated, the ratios after p.o. administration increased with time, leading to an estimation that about 87% of 125 I-PrP Sc in blood got cell-associated at 24 hr (Fig. 2B). Two-way ANOVA revealed that these two administration routes produced statistically different time profiles of PrP Sc in relation to cell association. At this point we cannot know whether 125 I-PrP Sc was taken up by certain type of blood cells or remained attached to the cell surface. Next, we further investigated the localization of 125 I-PrP Sc in whole blood by fractionating plasma, peripheral blood mononuclear cells (PBMC), and red blood cells (RBC) at the time (3 hr) when whole blood levels peaked after p.o. administration, and fractional distribution of 125 I-PrP Sc compared with i.v. injection (Fig. 3). In p.o. administration (Fig. 3A), about 44% of 125 I-PrP Sc existing in whole blood was distributed in the plasma fraction, 17% found in the PBMC, and 39% associated to RBC fraction, indicating that more than half (56%) of 125 I-PrP Sc in whole blood was eventually cell associated. In contrast, localization of 125 I-PrP Sc in whole blood after i.v. injection was markedly different; majority (80%) of 125 I-PrP Sc was found in plasma, and 7% and 13% was present in PBMC and RBC fractions, respectively (Fig. 3B). The explanation for the striking finding that the degree and manner of PrP Sc cell association depends on the route of exposure, likely comes from differences in spatio-temporal absorption of 125 I-PrP Sc to blood. Figure 3C shows that the levels of 125 I-PrP Sc per respective cell counts in whole blood indicating about 1000-fold higher amount of the agent was accumulated in PBMCs compared to that of RBCs, regardless of the route of administration.
Tissue distribution and oral bioavailability of PrP Sc . Pharmacokinetic parameters were calculated using a statistical moment analysis, which is pharmacokinetic model-independent approach, to estimate disposition parameters for 125 I-PrP Sc , such as terminal half-life (t 1/2 , time in which 50% of 125 I-PrP Sc is eliminated from systemic circulation), steady state volume of distribution (Vdss, estimates the volume in which PrP Sc distribute o. administration were 14-48%, and 62-87%, respectively. Dashed line represents the ratio of 0.5, which corresponds to the ratio when all material remains in plasma. Blood collected at various time points during the same experiment described in Fig. 1 were separated in plasma and cellular package and radioactivity was quantified and corrected for TCA precipitation. Asterisks indicate statistical difference in the levels between the whole blood and plasma at various time points. Statistical difference was evaluated by two-tailed student T-test with Welch's correction, and two-way ANOVA followed by the Bonferroni's test. *P < 0.05, and ***P < 0.001. Mean values with standard error terms are presented. (n = 4-7). ID: Injected dose. Pharmacokinetic parameters for 125 I-PrP Sc after p.o. and i.v. administrations were calculated by these time-course profiles and summarized in Table 1. (C) The levels of 125 I-PrP Sc expressed by unit cell number in whole blood. 125 I-PrP Sc was administered at 1 × 10 6 cpm in each route. Plasma, buffy coat (PBMC), and RBC fractions were freshly separated by Ficoll and TCA precipitable radioactivity was measured. Total radioactivity of intact 125 I-PrP Sc in whole blood was defined as 100%, and fractional percentage was calculated for each fraction. Statistical evaluation was performed with One-way ANOVA followed by Neuman-Keuls post-hoc test. **P < 0.01, and ***P < 0.001. Mean values with standard error terms are presented. (n = 3-4). ID: Injected dose.
when the system has reached the equilibrium between tissue distribution and elimination), total body clearance rate (CLtot, estimates the rate in which PrP Sc is removed from the body) and mean residual time (MRT represents the time required for clearing distributed 125 I-PrP Sc in the body) and oral bioavailability ( Table 1). Considering that the levels of PrP Sc in plasma and blood cells were different depending on the route of administration, we determined the pharmacokinetic parameters for both plasma and whole blood. While terminal half-lives of plasma 125 I-PrP Sc ranged 5.5 to 6.5 hr regardless of administration routes, prolonged terminal half-life (15.7 hr) was observed in whole blood after p.o. administration, indicating that cell association increased the terminal half-life of the agent. Mean residual time was estimated 20.7 hr based on whole blood profile in p.o. administration. This was mostly because of larger volumes of distribution at pharmacokinetic steady-state. The area under the curves for 125 I-PrP Sc by p.o. and i.v. routes were compared to estimate the percentage of 125 I-PrP Sc absorbed through the GI tract into the body, which is defined as absolute bioavailability of 125 I-PrP Sc . Estimated bioavailability of 125 I-PrP Sc based on whole blood was 33.6% which was 2.4-fold higher than plasma based value of 14.3% ( Table 1). The rate of the bioavailability was also shown as Tmax and Cmax values which refer to the time to reach maximal amount and the maximal concentration at Tmax, respectively.
Once 125 I-PrP Sc became blood-borne after p.o. administration, systemic circulation allows the delivery of prions throughout the body. We investigated the distributions of 125 I-PrP Sc in brain and peripheral organs (Fig. 4). The levels in the liver, spleen, and kidney showed a similar time profile as in systemic circulation ( Fig. 2A). The volumes of 125 I-PrP Sc distribution were 9-10 times higher than the vascular space in each tissue measured separately by 125 I-albumin (Fig. 4B), suggesting that the majority of PrP Sc found in these tissues was indeed in the  tissue and not in the blood associated to the tissue. The quantity of material reaching the brain was very low, but measurable at 3 hr after p.o. administration which is in agreement with prior studies showing that PrP Sc can cross full-width of the blood-brain barrier 20,22,23 . Brain distributed 125 I-PrP Sc was ~1.5 fold higher than vascular space in the brain.  (Fig. 5). This result indicate that oligomeric forms of 125 I-PrP Sc were effectively transported across the gut lumen and a fractional portion (0.1%ID/ml, Fig. 2A) of the agent was circulating in plasma even at long times after the administration. and distributed in peripheral organs through the blood stream retained the ability for self-replication, we subjected samples to Protein Misfolding Cyclic amplification (PMCA), a technique that mimics prion replication in vitro. For these studies we employed the non-radioactive isotope iodine-127 which was incorporated into PrP Sc as a control for the iodination procedure. After 3 hr following p.o. administration of 127 I-PrP Sc , the agent was detected in all GI-tract tissues including stomach, duodenum, jejunum, ileum, and colon by the 2 nd round of PMCA (Table 2 and Fig. S2). Among non-GI peripheral organs, 127 I-PrP Sc was detected in the liver as early as by the 2 nd round of PMCA, suggesting a first pass effect for 127 I-PrP Sc after crossing the GI tract. Lower levels of PrP Sc were detected in the spleen, kidney, whole blood, and plasma 3 hr after p.o. administration. In the GI tissues and peripheral organs obtained at 24 hr after p.o. administration, PMCA amplifiable PrP Sc was constantly found in these samples even though there was no detectable PrP Sc in the blood and plasma 24 hr after p.o. administration, suggesting minute amounts of PrP Sc deposition in these organs. PMCA with brain tissues obtained at 3 and 24 hr after p.o. administration found that brain retained self-replicating PrP Sc (Fig. 6), suggesting hematogenous route of brain distribution upon oral ingestion. There was no PrP signal in tissue samples from control mice receiving a mock injection. These data suggest that PrP Sc distributed in the brain and peripheral organs was not only intact but retained its capacity to sustain prion replication which is essential for the infectious property of PrP Sc .

Discussion
General tissue disposition of an exogenous substance is governed by a homeostatic balance between tissue uptake and clearance. Therefore, investigating systemic pharmacokinetics and biodistribution as a dynamic system integrating the CNS and the periphery is critical to understand the consequence of body exposure to infectious prions. In the present study, we found that the absorption, distribution and accumulation of PrP Sc in various organs and tissues are highly dynamic and change substantially with time after infection in relation to the routes of prion exposure. Our experimental approach does not permit to study protease-sensitive PrP Sc , because for radiolabeling PrP Sc needs to be highly purified which necessitates strong PK treatment to remove the bulk of protein contaminants. Thus, in our study we focus exclusively on the bioavailability of the protease-resistant fraction of PrP Sc . While it has been reported that a large proportion of RML infectivity is associated to protease-sensitive PrP Sc 24 , it remains to be study whether these species behave similarly to the protease-resistant forms in terms of bioavailability and pharmacokinetic properties. In this sense, it is important to highlight that in our previous study we compared the pharmacokinetic profiles of non-PK digested PrP Sc (studied by western blots) and protease-resistant PrP Sc (studied by radiolabeling) and both preparations showed an indistinguishable initial-phase distribution in blood 20 . In the present study, the amount of 125 I-PrP Sc orally administered was about 15.4 ng per mouse. To put this number in perspective, we compared to the quantity used for infectivity bioassays. The PrP Sc concentration in a clinically sick animal has been estimated to be 0.6-2 × 10 −5 g per g of brain 25,26 . For simplicity we will consider 10 μ g/g of brain, in other words, 1 μ g of PrP Sc per ml of a 10% brain homogenate. For infection by intracerebral injection, researchers traditionally use 1-10 μ l of this solution, i.e. 1-10 ng of PrP Sc . For oral administration, 10-100 μ l of this material is commonly used, i.e 10-100 ng of PrP Sc . Therefore, the quantity of PrP Sc used in our experiments is very similar to that used for infectivity bioassay.
In our previous studies 20,22,23 , we reported that purified PrP Sc injected i.v. can reach the brain exclusively by crossing the blood-brain barrier, without the involvement of peripheral nerves, and PrP Sc appeared in both the brain parenchyma and the cerebrospinal fluid after i.v. administration 22 . In the current study, we found that the brain received ~0.03%ID of PrP Sc per g of brain after p.o. administration (Fig. 4A) which represents ~4.5 pg of PrP Sc per g of brain weight. Since blood contains 1.1%ID/ml of PrP Sc , it is possible to estimate that ~3.3 pg of PrP Sc are present in the cerebral blood vessel per g of brain, based on the brain vascular space 20 . Thus, brain parenchymal distribution could be up to ~1.2 pg per g of brain. As stated before, we have estimated that the PrP Sc Figure 6. Replication-competent PrP Sc in brain after p.o. administration. Whole brains from mice that received a p.o. administration of 127 I-PrP Sc were collected 3 and 24 hr post administration. Control tissue samples were from mice which received a mock injection. After precipitation with sarkosyl to remove tissue components that interfere with PMCA, samples were subjected to 4 serial rounds of PMCA, as described in Methods. Western blotting was performed with 6D11 antibody at a dilution of 1:10,000 after treatment of the samples with proteinase K. We employed non-radioactive iodine-127 to label PrP Sc for PMCA analysis. Four brain samples obtained at a designated time are indicated 1 through 4, lanes + or − represent positive or negative control for PrP Sc (RML), respectively. Lane C indicates non-PK digested 10% normal brain homogenate. The results indicate that 50% of brain samples have replication-competent PrP Sc 3 hr and 24 hr after p.o. administration.
Scientific RepoRts | 6:32338 | DOI: 10.1038/srep32338 concentration in the brain of a terminally sick animal is around 0.6-2 × 10 −5 g per g of brain 25,26 . Since in these models the last infectious dilution of brain homogenate injected by intra-cerebral inoculation is ~10 −8 32 , we estimate that the minimum quantity of PrP Sc that can initiate disease in the brain is ~0.2 pg per g of brain. Thus, numerically, the amount of PrP Sc that reached the brain from the hematogenous route after oral administration appears to be comparable to the minimum amount needed to initiate the disease. We are aware that this view may seem contradictory to previous studies [33][34][35][36] which have solidly established that neuroinvasion is dependent of peripheral replication of PrP Sc in lymphatic tissues including secondary lymphatic organs and transport thru the peripheral nerves. Our present findings do not intend to contradict these widely accepted pathways for neuroinvasion, but merely make the point that the amount of prions reaching the brain directly by the hematogenous route is not negligible and the exact contribution of this route should be further investigated.
Also, our findings do not diminish the importance of the lymphatic system and retrograde nerve transport of prions into the CNS. On the contrary, our data showing high affinity and rapid association of orally administrated PrP Sc to PBMC further confirms the involvement of the lymph-reticular system in prion diseases. Indeed, after oral administration up to 87% of PrP Sc in blood was cell-associated whereas only ~20% was present in blood cells when the protein was directly loaded into the blood. Considering the volume of buffy coat fraction in whole blood is < 1%, the fact that PBMCs contain 17% of 125 I-PrP Sc present in blood (Fig. 3A), suggests a higher affinity of PBMCs for PrP Sc compared to RBCs. We estimate that PBMCs contain about 1000 times higher concentrations of PrP Sc than RBCs when the number of cells in whole blood is taken into account (Fig. 3C). Nevertheless, the large amount of PrP Sc in the RBCs fraction (48%) suggests these cells may serve as a reservoir of PrP Sc after p.o. administration. The blood/plasma ratio indicates that significant amount of PrP Sc present in blood at long times after p.o. administration (> 12 hr) is cell-associated, suggesting a delay in biological clearance of this portion of PrP Sc . Surprisingly, PrP Sc directly injected into blood stream largely remained in plasma fraction, and only about 20% of PrP Sc got cell associated (Fig. 3B). The differences in blood cell association of PrP Sc can be explained by the fact that PrP Sc crossing the GI tract is immediately exposed to lymphatic tissues and follicle associated white blood cells 13 . On the contrary, PrP Sc directly infused into vascular system remains in plasma fraction after saturating bindings to circulating blood cells. Considering that different routes of exposure to prions lead to substantial differences in PrP Sc distribution in distinct blood fractions, it is likely that whole body distribution of prions may also differ at different stages of the disease when prions could be shed into blood from either peripheral replication or leakage from the brain 25 .
By comparing pharmacokinetic parameters obtained through p.o. and i.v. administration (Table 1), we estimated the percentage of PrP Sc that was absorbed into the body. The estimated bioavailability of prions was 33.6%, which is very high considering the protein nature of the infectious agent. PrP Sc crosses the intestinal barrier as an oligomeric form that retains the capacity to self-replicate. Once PrP Sc became blood-borne, its distribution half-life was about 5 min, tissue sequestration occurred in the brain and majority of peripheral organs, without much degradation of PrP Sc in blood 20 . A caveat from our study is that the pharmacokinetic parameters obtained in mice cannot directly translate to humans. In addition, the use of stomach gavages instead of putting the material into the animal food may not replicate exactly the natural situation. The reason we used stomach gavage administration because this procedure gives the best control over the amount that is really getting in the gastro-intestinal system. The exposure route and amount of infectious prions upon initial contact in natural case may be less robust compared to present experimental setting. Nevertheless, our data show that intact, replication-competent PrP Sc can effectively cross the intestinal barrier, distribute in blood flow, and temporally appeared in the brain. Purification of PrP Sc from infected mouse brain. PrP Sc was purified from mice infected with Rocky Mountain Laboratory (RML) scrapie strain as previously described 20,37,38 . Briefly, brain tissue was homogenized at 10% w/v in phosphate buffered saline containing 10% sarkosyl and subjected to a series of differential centrifugations employing a Beckman TL-100 ultracentrifuge (OptimaMAX Ultracentrifuge, Beckman-Coulter) with the final step consisting of a sucrose gradient. The material was then treated with proteinase K (PK) (100 μ g/ml) at 37 °C for 2 hr followed by ultracentrifugation to precipitate PrP Sc . The purity of PrP Sc was confirmed by silver staining and estimated to be > 95%. PrP Sc concentration was measured by micro BCA protein assay reagent (Pierce).

Methods
Radiolabeling. Purified PrP Sc and albumin (Fraction G, bovine serum albumin, Sigma) were radioactively labeled by the iodobead method with [ 125 I]Na (Perkin-Elmer) or [ 127 I]Na (Sigma), as previously described 20,39 . Briefly, ultra-purified PrP Sc (10 μ g) or albumin (5 μ g) was mixed with [ 125 I]Na (2 mCi) or equivalent amount of 127 INa in 250 mM chloride-free sodium phosphate buffer (pH 7.4), and the protein labeling was initiated by adding one iodobead into the mixture. After 15 min, the reaction was terminated by removing the bead from the mixture. Each labeled agent was purified by Sephadex G-10 chromatography to remove free iodine. The labeled PrP Sc or albumin was diluted with phosphate buffered saline and further centrifuged in albumin pre-coated Microcon filtration tube (Mw cutoff: 10 kDa) at 12,000 rpm for 30 min to further remove free iodine from the G-10 eluate. Labeled PrP Sc and albumin was extensively washed by this procedure. 125 I-PrP Sc had a specific radioactivity of 1.3 × 10 5 cpm/ng. The radioactively labeled PrP Sc preparations we used had over 95% precipitation with trichloroacetic acid (TCA). Materials were freshly prepared on the day of the experiment.
Separation of blood fractions by Ficoll-Paque density gradient. Whole blood was diluted with 2-fold volume of phosphate buffered saline, and gently floated onto equal volume of Ficoll-Paque Premium (GE Healthcare, London). The sample was centrifuged at 400 × g for 30 minutes at room temperature. Fractions were separated into plasma, PBMC, and RBC fractions. The radioactivity from 125 I-PrP Sc was measured after TCA precipitation of each fraction by a gamma counter (Packard), and results were presented as %ID/ml of original whole blood volume. Representative numbers of cells in PBMC and RBC fractions were counted, and the values were 1 × 10 6 cells/ml and 4 × 10 8 cells/ml of whole blood, respectively. Pharmacokinetic analysis. The arithmetic mean values of concentration-time profiles of intact 125 I-PrP Sc in whole blood and plasma were used for pharmacokinetic analysis based on the statistical moment theory according to Yamaoka and colleagues 40 , using the following equations: where AUC and ID accompanied with a suffix represent the value obtained by each injection route.
To estimate gastric emptying half-time, the concentration-time profile for 125 I-PrP Sc in the stomach was fitted with following bi-exponential equation using the nonlinear least squares method, where A and B represent y-intercepts at the initial and late phases after p.o. administration, respectively. The rate constants in initial (α) and late (β ) phases were employed to estimate gastric emptying half-times as ln2/α and ln2/β , respectively. HPLC analysis. The stability of 125 I-PrP Sc in the plasma, blood, brain, liver, spleen, kidney, stomach, duodenum, jejunum, and colon was examined using HPLC. Mice received a p.o. administration of 125 I-PrP Sc of 1 × 10 6 or 2 × 10 7 cpm/mouse. The samples from animals sacrificed at 3, or 24 hr were used for the HPLC analysis. Plasma (300 μ l) was directly loaded onto the HPLC column. The tissues were mechanically homogenized in a 9-fold volume of mobile phase, centrifuged at 20,000 × g for 20 minutes, and the supernatant was collected. Each tissue extract was injected onto HPLC (Shimadzu UPLC system LC-20AB). Radio-HPLC chromatography was conducted with the size exclusion column BioSep-SEC-S4000 (7.8 mm × 300 mm, Phenomenex, CA) and a guard cartridge. The mobile phase consisted of 25 mM sodium phosphate buffer (pH 7.4). Fractions were collected at 1-min interval for first 5 min, thereafter at 10-sec per fraction at the flow rate of 1.0 ml/min. Column recovery of loaded radioactivity was > 80%. Radioactivity was measured by gamma counter. TCA precipitation of fractions was also performed. The retention time for molecular size markers was separately measured, and the markers included thyroglobulin (669 kDa), aldolase (158 kDa), conalbumin (75 kDa), carbonic anhydrase (29 kDa), aprotinin (6.5 kDa), and iodine-125 (125 Da). UV absorbance at 280 nm was used to characterize the retention time of the markers, except for iodine-125 which was characterized by radioactivity. All markers were obtained through commercial sources.
PMCA procedure. Aliquots of purified PrP Sc either iodinated or not were used to seed conversion of mouse PrP C in 10% normal mouse brain homogenate. The detailed protocol for PMCA has been described elsewhere 41,42 .
Scientific RepoRts | 6:32338 | DOI: 10.1038/srep32338 Briefly, tissues from infected animals were homogenized (10% w/v) in phosphate buffered saline containing complete protease inhibitors (Roche) by a mechanic homogenizer and hard tissue tubes. Tissue homogenates were incubated with equal volume of 20% sarkosyl solution for 10 min in agitation at room temperature. Thereafter, the mixture was centrifuged at high speed (100,000 × g) for 60 min at 4 °C. Pellet was washed with one volume of PBS and centrifuged for 30 mins at the same speed. The final pellet was resuspended in 100 μ L of 10% normal brain homogenate for PMCA reaction. Four rounds of PMCA were conducted, and durations of PMCA reaction per round were 3 days for 1 st round and 2 days for 2 nd through 4 th round. Reaction tubes were positioned on an adaptor placed on the plate holder of a microsonicator (Misonix Model Q700) and programmed to perform sonication cycles of 20 sec every 30 min at an amplitude of 20, in an incubator set at 34 °C.