Toxicogenomics: roadblocks and new directions.

The persistence of infectious biomolecules in soil constitutes a substantial challenge. This holds particularly true with respect to prions, the causative agents of transmissible spongiform encephalopathies (TSEs) such as scrapie, bovine spongiform encephalopathy (BSE), or chronic wasting disease (CWD). Various studies have indicated that prions are able to persist in soil for years without losing their pathogenic activity. Dissemination of prions into the environment can occur from several sources, e.g., infectious placenta or amniotic fluid of sheep. Furthermore, environmental contamination by saliva, excrements or non-sterilized agricultural organic fertilizer is conceivable. Natural transmission of scrapie in the field seems to occur via the alimentary tract in the majority of cases, and scrapie-free sheep flocks can become infected on pastures where outbreaks of scrapie had been observed before. These findings point to a sustained contagion in the environment, and notably the soil. By using outdoor lysimeters, we simulated a contamination of standard soil with hamster-adapted 263K scrapie prions, and analyzed the presence and biological activity of the soil-associated PrP Sc and infectivity by Western blotting and hamster bioassay, respectively. Our results showed that 263K scrapie agent can persist in soil at least over 29 months. Strikingly, not only the contaminated soil itself retained high levels of infectivity, as evidenced by oral administration to Syrian hamsters, but also feeding of aqueous soil extracts was able to induce disease in the reporter animals. We could also demonstrate that PrP Sc in soil, extracted after 21 months, provides a catalytically active seed in the protein misfolding cyclic amplification (PMCA) reaction. PMCA opens therefore a perspective for considerably improving the detectability of prions in soil samples from the field.


INTRODUCTION
Transmissible spongiform encephalopathies (TSEs) comprise a group of fatal neurodegenerative diseases such as bovine spongiform encephalopathy (BSE) in cows [1], chronic wasting disease (CWD) in deer (Odocoileus spp.) and elk (Cervus elaphus nelsoni) [2][3][4], scrapie in sheep and goats [5][6][7] and Creutzfeldt-Jakob disease (CJD) in humans [1]. The exact molecular pathomechanisms underlying TSEs have not yet been fully elucidated but it is generally accepted that a pathologically misfolded and/or aggregated isoform of the normal cellular prion protein (PrP), referred to as PrP Sc and PrP C , respectively, is the key pathogenic factor for this group of diseases [1].
Among the known TSEs, only scrapie and CWD are contagious diseases which show horizontal transmissibility under natural conditions [2,4,8]. CWD is the only TSE known to affect freeranging animals [3]. The regular occurrence of scrapie in affected areas [7] and the spread of CWD in North America and Korea [9,10] among mule deer, white-tailed deer and elk indicates that a contagion in the environment is responsible for the occurrence of these TSEs [4,[11][12][13][14], and even raises the possibility of a crossspecies transmission under natural conditions. Recent findings demonstrated that saliva from deer with CWD harbours infectivity and can transmit this TSE upon peroral uptake [15]. Other studies pointed to transmission of scrapie among sheep by vectors like mites, fly larvae or other ectoparasites [16][17][18][19]. It has also been hypothesized that sporadically occurring TSEs may be induced by insecticides or by a disproportion of manganese and copper in soil leading to an enrichment of manganese in animals [20,21]. Alternatively, an influence of the Fe/Mn ratio in forage has been discussed in connection with TSEs [22]. However, on the balance of all evidence available so far, contaminated soil appears as one of the most likely sources of infection in the natural transmission of scrapie and possibly also CWD. It has been known for decades that sheep can become infected with scrapie while grazing on pastures where infected sheep have been kept before, and scrapie occurs often in areas where it has already occurred previously [7,8,23]. Furthermore, Brown and Gajdusek found that scrapie agent remains infectious after persisting in soil for 3 years as evidenced by intracerebral bioassay in Syrian hamsters [24]. The putative transmission of scrapie and CWD via soil is also corroborated by recent studies showing that prion infectivity binds to soil components with high affinity [25,26], thereby retaining its pathogenic biological activity [25]. Thus, soilassociated TSE agents in the environment represent a potential hazard. This holds true not least since prion infectivity exhibits an unusually pronounced resistance against both physical and chemical methods of inactivation, as described in detail elsewhere [27][28][29][30][31][32][33][34][35][36][37].
The contamination of soil with TSE infectivity can occur from several sources. Since recent studies could demonstrate that scrapie infectivity is present in various tissues and body fluids of infected animals [38,39], it has been assumed that the persistent prion protein enters the environment by contaminated excrements, birth-related tissues such as placenta, or even whole carcasses. While a similar excretion pattern appears conceivable for the CWD agent [40] which has also been found in saliva [15], the present knowledge about the BSE pathology in cattle does not argue for a significant shedding of the infectivity via faeces, urine, or during birth [41,42]. However, small ruminants infected with BSE could supposedly spread the BSE agent throughout the environment in a similar manner as known from scrapie-affected sheep or CWD-affected animals [43]. Moreover, unconventional conditions like for example the burial of animal carcasses at larger numbers as practiced for example during the foot and mouth disease outbreak in the UK in 2001 [44], may have fostered a dissemination of BSE agent in the soil or ground water. Once deposited there, all three TSE agents, BSE, scrapie, and CWD, must be assumed to persist in an infectious state for long periods of time.
So far, the oral transmission efficacy of long-term prion contaminations in soil have not been investigated. Therefore, we have studied the persistence of PrP Sc in the environment over time and measured its oral transmissibility by bioassay in Syrian hamsters. With outdoor lysimeter experiments we simulated the situation on pastures using soil spiked with scrapie-infected hamster brain homogenate over a period of 29 months and analyzed the fate of the prion proteins by sensitive Western blotting and, in part, also by protein misfolding cyclic amplification (PMCA). The infectivity of such contaminated soil samples and the respective aqueous soil extracts was tested in the hamster bioassay.

Extraction and Recovery
In the first phase of the study several buffers -known as standard buffers for protein extraction from mammalian cells -and other solutions were tested in order to find out the optimal extraction method for prion protein from contaminated soil (see Materials and Methods). The obtained results lead us to use 1% SDS (sodium dodecyl sulphate) in sterilized water (figure 1a), which allowed detection of PrP Sc in samples of German standard soil containing an amount of PrP Sc that corresponded to 1.25 mg of infectious brain material (figure 1b).

Soil Incubation Experiments
The findings from our lysimeter experiments indicated a remarkable persistence of PrP Sc in soil by clearly showing that -even after an incubation for 29 months -PrP Sc could be still extracted from soil and detected by Western blotting. These results were confirmed by using two alternative highly sensitive Western blotting techniques. As shown in figure 2a, PrP Sc , in the form of its protease-resistant core of the pathological prion protein PrP Sc (PrP27-30), is still detectable in soil after 29 months. However, a decrease in the extractable concentration is obvious. The strongest decrease can be seen during the first month of incubation in the soil (figure 2a, lane 2 and 3). After three months the extractable amount of PrP27-30 remained almost constant until month 21, with small variations at month 12 and month 18 (figure 2a, lane 6 and 7). In Fig. 2a, a slightly stronger signal for PrP27-30 was found for 21 months (lane 8) as compared to 18 months (lane 7). Both samples were taken from the same soil bag, however, the location from where the sampling was performed may have differed. Thus, the finding possibly reflects minor inhomogenities either in the original load of contamination or in the degradation of PrP Sc depending on the microenvironment of the examined sample. On the other hand, it has to be noted that the extraction yield and Western blotting efficiency inevitably may also vary in certain ranges from sample to sample. A further decrease was observed after 26 and 29 months (figure 1a, lane 9 and 10), however, the PrP27-30 specific bands remain clearly visible. Where performed, analyses of samples from different soil bags produced consistent results. In order to verify that the detected bands originated from PrP Sc a control experiment was performed: After deglycosylation with PNGase F the PrP27-30 bands extracted from soil after 21 and 18 month showed an electrophoretic shift towards a single band at about 20 kDa, the molecular mass to be expected for deglycosylated PrP27-30 from 263K hamster scrapie (figures 2 b, lanes 3 and 4, respectively).

Amplification of PrP Sc extracted from soil
In order to check whether PrP Sc extracted from soil can be used for the amplification of protease resistant prion protein by the protein misfolding cyclic amplification (PMCA) reaction [45,46], soil extracts from month 21 were used as test seeds. As shown in figure 2c, a strong increase of the signal for PrP Sc was obtained with this method: while no signal for PrP Sc can be detected in Western blot detection of PrP27-30, the proteinase K-resistant core of PrP Sc , extracted by using 1% SDS from soil contaminated with 263 K scrapie brain homogenate from hamsters after 1 h of incubation (dilution series). PrP Sc could be detected in soil samples containing 1.25 mg or higher amounts of scrapie brain tissue after extraction with 1% SDS-solution. Samples were digested with proteinase K prior to Western blotting.

Presence of PrP Sc in surrounding soil samples
For analyzing the fate of the prion protein in more detail, soil samples in the vicinity of the gauze bag, as well as the gauze bag itself were analyzed by Western blotting. As shown in figure 3a, no PrP Sc specific signal could be detected in the surrounding soil samples. This is indicative of PrP Sc being immobilized after binding to soil compartments. In some approaches, a very faint PrP Sc specific signal was visible when analyzing aliquots of SDS solution in which the gauze bag had been washed, however, it could be shown that this positive prion signal was a result of adherent soil particles (figure 3, lane 3, arrow).

Bioassay Experiments
For a detailed risk assessment of scrapie-contaminated soil it was of major importance to analyze whether the detectable PrP Sc in the soil extracts still exhibited oral infectivity after incubation times up to 29 months. Therefore, a bioassay with Syrian hamsters was performed by feeding the animals with contaminated soil or aqueous soil extracts that had been collected after soil incubation for 26 and 29 months, respectively. Hamsters fed with contaminated soil exhibited first scrapie-associated symptoms at 13166 days [mean6SD] after the first application. The hamsters reached the terminal stage of scrapie at 162612 days after the first feeding (tables 1 & 2). This indicates substantial amounts of persistent infectivity in soil that had been incubated for 26 and 29 months. After reaching the terminal stage of scrapie the animals were sacrificed and analyzed for the presence of PrP Sc in their brains. As shown in figure 4a (lanes 2 and 3 depicting two hamsters exemplarily), the brains of all animals in this group contained high amounts of PrP Sc . In addition, comparison of the electrophoretic and glycosylation profiles of the brain samples from these hamsters (   were maintained in the animals fed with soil that had been incubated with 263K scrapie brain homogenate. Remarkably, not only the hamsters fed with contaminated soil but also four hamsters of the group fed with aqueous soil extracts developed terminal scrapie at 256641 days after the first application (with incubation times of individual animals ranging from 201 to 321 days). Since this bioassay is ongoing, further animals of this group might still develop scrapie in the future.

DISCUSSION
The results of this research project show for the first time that the scrapie strain 263K remains persistent in soil over a period of at least 29 months and remains highly infectious after oral application to Syrian hamsters. It has to be pointed out that the key results of our time-course study on the fate of PrP Sc in soil have been validated, in part by examining blinded samples, at independent laboratories.
Only a few studies have addressed the question of a persistence of prions in soil so far [24][25][26]47], and the results from these studies are in principle in accordance with our observations. A pioneering study was published by Brown and Gajdusek in 1991 [24] showing that an aqueous extract from scrapie-contaminated soil remains infectious even after an incubation period of three years as confirmed by hamster bioassay. However, the infectivity studies were conducted by intracerebral injection and not by oral application. Furthermore, the PrP Sc concentration was not analyzed in this study, so that no data are available about the proteins absorption behavior to soil particles and about the corresponding degradation kinetics. Most recently, PrP Sc has been shown to bind to soil minerals [25] but only short-time incubation experiments of maximal one week were conducted and, again, bioassays were performed by the intracerebral route.
In this study we show by Western blotting a strong decrease in the amount of extractable PrP Sc over an incubation period of 29 months in soil. It is not yet clear whether this decrease resulted from a molecular degradation of PrP Sc or a tighter binding to soil particles. Stronger binding of molecules to soil particles with increasing incubation time is a well-known phenomenon in soil chemistry -the so called ''aging'' -and influences bioavailability and re-mobilization significantly [48,49].
Upon feeding hamsters with scrapie contaminated soil which had been incubated for over two years in outdoor lysimeters all animals developed terminal scrapie after relatively short in-  cubation times (162 dpi). In other studies it has been well established that pure 10% (w/v) brain homogenates from 263K scrapie hamsters cause terminal scrapie in perorally challenged hamsters after mean incubation times of about 155-165 days with an attack rate of 100% [50][51][52][53]. This indicates that scrapiecontaminated soil may represent a potential TSE hazard for ruminants in the environment. While a considerable excretion of infectivity has to be assumed for scrapie or BSE infected sheep and CWD infected deer [40,43,54], it is generally acknowledged that the potential environmental contamination risk represented by BSE infected cattle is marginal, if at all present [41,42]. On the other hand, the burial of bovine carcasses [44] might have accidentally led to a spill of BSE prions into the environment. Furthermore, the fact that even feeding of aqueous extracts from scrapie-contaminated soil induced a terminal scrapie infection in four hamsters so far suggests that surface water or groundwater from pastures of scrapie-affected flocks may provide a potential source of scrapie infectivity. However, the relevance of the results obtained in this study for the field situation should be interpreted with some caution, since only one soil type was used and only a limited number of animals were challenged in the bioassay. Therefore, other soil types and a larger number of animals have to be tested in future studies to allow for a robust risk assessment. Furthermore the exact binding properties and degradation kinetics of PrP Sc should be subject to further research. In addition, all published studies addressing the persistence of prion infectivity in soil were performed with scrapie prions while TSE agents causing BSE and especially CWD have not been analyzed so far.
An intensified monitoring of PrP Sc (and possibly also prion infectivity) in the soil appears mandatory for a more precise assessment of the risks emanating for humans and animals from prions in the environment. As shown in this report, PrP Sc extracted from soil can be used as a catalytically active seed in the protein misfolding cyclic amplification (PMCA) reaction. This opens a promising perspective for considerably improving the detectability of prions in the environment.

Biological Safety
Assays with scrapie-infected hamster brain were performed under laboratory conditions according to bio safety level 3** and in protected outdoor lysimeters, respectively.
Initially non-infectious pork brain was mixed with German standard soil and the efficiency of PrP C extraction using the above mentioned buffers was tested. In a second step, the optimal extraction procedure was verified by applying the procedure to soil samples mixed with 263K scrapie agent and monitoring the PrP Sc retrieval.

Incubation Experiments and Sampling Scheme
The outdoor experiments and all other experiments were performed using brains of terminally-diseased hamsters challenged with scrapie strain 263K.
For each approach, 1 g infectious hamster brain material was homogenized in 10 ml PBS (phosphate buffered saline) and added to 20 g of standard German sandy loom soil (Borstel). The soil/ brain mixture was filled into gauze bags and buried in lysimeters filled with the same soil at a depth between 15 and 20 cm. The gauze had a mesh size of 250 mm, which enables the contact with microorganisms and meso fauna with the soil/brain mixture but avoids the contact with macro fauna derived/related organisms. To protect the gauze bags from mice, the bags were put into steel cages. At defined time points (after 0, 1, 3, 6, 12, 18, 21, 26, 29 months) the steel cages were dug out and the gauze bags as well as the surrounding soil were taken for sampling and analyzing. The analyses on the presence of residual PrP27-30 in soil samples were performed with a different number of soil bags, depending on the incubation times to be tested. For ''short'' incubation times of ,12 months, for which detectability of residual PrP appeared most likely when the experiments were designed, samples from three different soil bags were examined. For ''intermediate'' incubation times of 12-18 months samples from two different bags were tested, and for ''long'' incubation times of .18 months one soil bag was yet available for sampling. All bags were buried in close vicinity.

Sample Preparation and Western Blot Analysis of Soil and Hamster Brain Samples
Remark: Western blot experiments were performed in independent runs at different laboratories, with each laboratory using its established techniques and procedures. The protocols are listed below. The different analytical protocols produced consistent results.

Western Protocol I
For the protein extraction from the contaminated soil as well as from surrounding soil samples, 20 ml of a 1%-SDS-solution (SDS; Sigma-Aldrich, Steinheim, Germany) were added to 20 g of testing material in a 50 ml tube. The suspension was vigorously shaken on a horizontal shaker for approx. 1 h, followed by a centrifugation step at 5,000 rpm for 20 min. 200 ml of the clear supernatant was incubated with proteinase K (50 mg/ml; 37uC; 1 h, Carl Roth GmbH, Karlsruhe, Germany) to eliminate nonresistant proteins. After digestion, the supernatants were boiled for 5 min in Laemmli's sampling buffer (50 mM Tris (pH 6.8), 2% SDS, 10% glycerol, 50 mM ß-mercaptoethanol and 0.001% bromphenol blue) in a 1.5 ml tube and analyzed by Western blotting. The prepared samples were stored at 280uC.
For confirmation of these results, highly sensitive Western blotting was performed in independent experiments at different laboratories using their established protocols.

Western Protocol II
The protein extractions from contaminated soil samples were carried out by adding 2 ml 1%-SDS-solution to 2 g of testing material in a 15 ml tube. The suspension was vigorously shaken for 2 min, followed by a centrifugation step at 7,000 rpm for 10 min. 50 ml of the clear supernatant was used for proteinase K digestion (100 mg/ml; 37uC; 1 h). After digestion the supernatants were mixed with an equal volume of 26 sample loading buffer (50 mM Tris (pH 6.8), 2% SDS, 10% glycerol, 50 mM ßmercaptoethanol and 0.001% bromphenol blue) and heated to 100uC for 5 min for PrP-Western blotting.
Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blot analyzes of samples from hamsters were performed as described elsewhere [58] with recently published modifications [39]. In short: Following SDS-PAGE, proteins were transferred to a polyvinylidene difluoride (PVDF) membrane (Immobilon, Millipore, Billerica, USA) using a semidry blotting system. The membrane was blocked by incubation for 30 min in TBS containing 3% (w/v) nonfat milk powder (NFMP) and 0.05% (w/v) Tween-20 (NFMP-TBST). Blots were incubated overnight in primary antibody solution (monoclonal anti-PrP antibody 3F4 [mAb 3F4] from cell culture, diluted 1:2,000 in NFMP-TBST) at 4uC. After washing 5 times for at least 20 min with NFMP-TBST, blots were incubated in secondary antibody solution (alkaline phosphatase conjugated goat anti-mouse IgG; Dako, Glostrup, Denmark; 1:5,000 diluted in NFMP-TBST) for 90 min at room temperature. After washing 5 times with NFMP-TBST over a total period of at least 1.5 h, the membranes were pre-incubated 265 min in assay-buffer (100 mM Tris, 100 mM NaCl, pH 9.5) and finally developed with CDP-star solution (Tropics, Applied Biosystems, Beford, USA) for 5 min according to the instructions of the manufacturer. PrP-signals were visualized on a X-OMAT AR (Kodak, Sigma-Aldrich, Steinheim, Germany) film. Films were exposed for 5-30 min.
For deglycosylation, 500 ml of the soil extract solution were dialyzed in PBS for 16 h in order to remove the detergents, and 20 ml of the aliquot was digested using PNGase F (New England Biolabs, Ipswich, USA) according to the instructions of the manufacturer prior to Western blotting.

Western Protocol III
Sample preparation and extraction of the prion protein from soil samples as well as proteinase K digestion was performed as described in protocol I. After electrophoresis using 16% bisacrylamide gels, proteins were transferred on a PVDF-membrane in a semi-dry chamber. The membranes were blocked in 5% dry fat milk in PBS (phosphate buffered saline) containing 0.1% Tween 20 (Merck, Darmstadt, Germany) (PBS-T) for 30 min and subsequently incubated with the PrP-specific monoclonal detection antibody 3F4 (Chemicon International, Inc., California) in a dilution of 1:3,000 in 5% dry fat milk in PBS-T for 1 h 30 min. The membranes were washed 3 times for 10 min with PBS-T and then incubated with a secondary antibody bound to alkaline phosphatase in a dilution of 1:2,000 in PBS-T (goat-antimouse-AP, Dianova, Hamburg, Germany) for 1 h. After again washing three times for 10 min, the membranes were incubated 2 times for 2 min in assay buffer containing 200 mM Tris-HCl and 10 mM MgCl 2 (pH 9.8). Finally, the chemiluminescence substrate CDP-Star (Tropix, Bedford, USA) was applied and incubated on the membrane for 5 min before the light signals were detected in a camera using the analysis software Quantity One (Bio-Rad, Munich, Germany).

Bioassay in Syrian hamsters
Forty two female Syrian hamsters, approx. 90 days old, were obtained from Charles River Laboratories, Germany. They were handled according to the regulations of the local authorities (Bezirksregierung Arnsberg, reference number 50.8735.1 Nr. 108/ 1) in a biosafety Level 3 containment facility.
12 animals were fed weekly over a period of 12 weeks; 11 times with 100 mg soil/brain mixture each and one time with 50 mg soil/brain mixture taken from the outdoor lysimeters after an incubation period of 26 and 29 months (table 1). Oral application was achieved by mixing the soil (1,150 mg in total) with commercial hamster feed (ssniff, Soest, Germany).
To analyze the remaining infectivity in the aqueous extract, 10 g contaminated soil was mixed with 10 ml water and was vigorously shaken on a horizontal shaker for approx. 1 h. The mixture was centrifuged with 5,000 rpm for 5 min and the supernatant was used for the bioassay. Oral application was performed weekly over a period of 12 weeks (1,150 ml in total) by mixing the extract with commercial hamster feed (ssniff, Soest, Germany). For this purpose, 12 additional animals were fed 11 times with 100 ml and 1 time with 50 ml aqueous extract from the soil/brain mixture taken from the outdoor lysimeters after an incubation period of 26 and 29 months.
The hamsters of both groups were monitored at least twice a week for the development of clinical signs of scrapie. Hamsters diseased with 263K scrapie showed head bobbing, ataxia of gait and generalized tremor. Such animals were frequently and persistently in motion, easily irritated by noise and touch, upon which they often twitch, and had difficulties maintaining balance and rising from a supine position. These clinical symptoms of hamsters are entirely consistent with those previously reported for the 263K scrapie agent [59]. At the occurrence of first neurological symptoms, the animals were observed daily and at appropriate end-points, clinically affected hamsters were killed. Brains from these hamsters were examined and analyzed for the presence of PrP Sc .
As a control, six animals were fed with non-contaminated standard soil and 6 animals were fed with aqueous extract from non-contaminated standard soil over a period of 12 weeks (1,150 mg or 1,150 ml in total). Additional 6 hamsters were fed only with commercial hamster feed (ssniff, Soest, Germany).

PMCA-Reaction
The PMCA method was carried out as reported previously [45,46] with modifications: In brief, normal hamster brains were homogenized at 10% (w/v) in PBS containing complete protease inhibitor cocktail (Boehringer-Ingelheim, Mannheim, Germany), 4 mM EDTA and 1% Triton-X-100. The homogenate was briefly centrifuged at 2,000 g and the supernatant was used as the PrP C source. Amplification was performed using an automatic ultrasonification apparatus (Sonicator 300 from Misonix, New York, USA) with a water-tank.
The PrP C substrate was mixed with 1/10 volume of the soil extract previously dialyzed in PBS, resulting in a total volume of 100 ml. PMCA amplification was performed by 40 cycles of sonication (40 sec. each) followed by incubation at 37uC for 1 h in the water-tank. The amplified product of the first round of amplification was diluted 1:5 with normal brain homogenate and the second round was performed. This process was repeated 4 times to obtain 160 cycles of PMCA. From each PMCAamplification round, aliquots of 50 ml were taken and digested with proteinase K (180 mg/ml) for 1 h at 56uC. Finally, the same volume of 26 sample buffer was added and heated for 10 min at 100uC prior to SDS-PAGE.