Mineral licks as environmental reservoirs of chronic wasting disease prions

Chronic wasting disease (CWD) is a fatal neurodegenerative disease of deer, elk, moose, and reindeer (cervids) caused by misfolded prion proteins. The disease has been reported across North America and recently discovered in northern Europe. Transmission of CWD in wild cervid populations can occur through environmental routes, but limited ability to detect prions in environmental samples has prevented the identification of potential transmission “hot spots”. We establish widespread CWD prion contamination of mineral licks used by free-ranging cervids in an enzootic area in Wisconsin, USA. We show mineral licks can serve as reservoirs of CWD prions and thus facilitate disease transmission. Furthermore, mineral licks attract livestock and other wildlife that also obtain mineral nutrients via soil and water consumption. Exposure to CWD prions at mineral licks provides potential for cross-species transmission to wildlife, domestic animals, and humans. Managing deer use of mineral licks warrants further consideration to help control outbreaks of CWD.


Introduction
Chronic wasting disease (CWD) was first observed in 1967 [1] and long thought to be a disease of minor scientific curiosity affecting mule deer (Odocoileus hemionus) and confined to the Rocky Mountains in northern Colorado and southern Wyoming, USA. Subsequently the disease was found in white-tailed deer (O. virginianus) and elk (Cervus canadensis). The geographic range of CWD has also expanded dramatically since 2000 [2] and is now present in 25 U.S. states, two Canadian provinces (http://www.nwhc.usgs.gov/disease_information/ chronic_wasting_disease/index.jsp), South Korea, Norway [3], and Finland (https://yle.fi/ uutiset/osasto/news/first_case_in_finland_elk_dies_due_to_chronic_wasting_disease/ 10108115) and has been found in moose (Alces alces) and reindeer (Rangifer tarandus) [2,4]. In addition, CWD prevalence has continued to increase with some free-ranging herds a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Here, we test the hypothesis that mineral licks used by deer harbor CWD prions, thus serving as potential environmental reservoirs for these infectious agents. During 2012-2015 we collected soil and water samples from 11 mineral licks (10 man-made and one natural) frequented by free-ranging white-tailed deer in a large CWD enzootic zone west of Madison, Wisconsin, USA [6] (Fig 2). We adapted a 96-well microplate variant of PMCA that incorporates Teflon 1 beads (mb-PMCA) [44] to detect CWD prions in soil and water samples. We optimized conditions to extract CWD prions from soils to enable reliable detection by mb-PMCA (S1 Text and S2-S4 Figs). We also tested deer feces collected in proximity to a mineral lick as a potential source of CWD prions. We previously detected CWD prions in fecal and urine samples from experimentally infected cervids by bead-assisted PMCA [22].

Ethics statement
Field studies conducted on public lands were done with the permission of the Wisconsin Department of Natural Resources. Field studies on private lands were done with the permission of the landowners. Field studies did not involve endangered or protected species.

Sample collection and preparation
Eleven mineral licks in the CWD-affected zone of south-central Wisconsin were located with the assistance of Wisconsin Department of Natural Resources personnel during 2013 (Fig 2). Deer visiting these mineral licks consume soil to supplement their mineral intake. During rain events, rainwater often accumulates at the lick sites. Deer visiting rainwater-filled mineral licks often stand in the water and suspend sediment as they drink. We therefore collected soil samples from dry mineral licks and, after a rainfall event, water samples before and after disturbing the underlying sediment. We collected soil at each mineral lick a single time as follows: six soil samples (2.54 cm diameter, 2.54 cm depth) were collected using a 1" diameter galvanized steel LaMotte soil sampling probe when water was not present. For each mineral lick site, the upper and lower 1.27 cm halves of the soil samples were pooled separately to yield a single pooled sample for the upper soil layer and one for the lower soil layer. Soil from Site 2 was too wet to reliably separate into upper and lower layers and was therefore pooled and analyzed as one sample. Pooled soil samples were freeze-dried, homogenized, and sieved through #18 U.S. standard testing sieve (VWR 57334-450) to remove water and non-soil components. We returned to each lick after a rainfall event and collected one water sample prior to and a second after disturbing the underlying sediment with a stick found in the immediate vicinity of each lick. Water collected prior to disturbance was clear of sediment and water collected after disturbance was turbid. We opportunistically collected fecal samples in the proximity of one heavily used lick to determine whether prions were shed in feces by deer using this site.

Extraction of prions from environmental samples
We evaluated a series of extractant solutions for their ability to recover prions from amended soils and allow amplification by mb-PMCA (S1 Text and S1 Fig). Of the extraction solutions evaluated 0.1 M sodium phosphate (pH 7.4) with 1% (w/v) N-lauroylsarcosine sodium salt (sarkosyl) followed by precipitation with sodium phosphotungstic acid (NaPTA) was selected.
From each pooled upper soil sample and each pooled lower soil sample, two 25 mg soil subsamples were rinsed twice with ultrapure water (18 MO cm; Barnstad GenPure Pro) before extraction of soil-bound prions. Each subsample extract was used to seed four mb-PMCA replicates. The upper or lower soil sample from each mineral lick was therefore represented by four analytical replicates for each of two subsamples (8 replicates for each soil layer, 16 soil replicates per lick). Water samples (50 mL) were centrifuged (2 min, 2000g) to separate particulate matter before 100 μL was removed and diluted 1:1 (v/v) with 3% sarkosyl in PBS. Following, NaPTA precipitation, detailed below, each sample was used to seed four mb-PMCA replicates (4 replicates for each water sample, 8 water replicates per lick).
Prions were extracted from two independent 100 mg replicates of each deer fecal sample using a protocol for ovine feces [45] modified to include additional rounds and a longer duration of centrifuging to fully clarify samples. Each fecal sample was diluted 1:9 (w/v) in ultrapure water containing Roche complete ultra mini EDTA-free protease inhibitor (Fisher Scientific 50-100-3269) and homogenized twice for 40 s at maximum speed (6 m s -1 ) in beadbeater tubes containing glass beads and silicon carbide particles (MP Biomedical #116916100). Sodium dodecyl sulfate was added to a final concentration of 1% (v/v) before three further homogenizations by bead beating. Samples were rotated and incubated (60 min, room temperature), then centrifuged (60 min, 15000g, 10˚C) to clear the particulate matter. Further clarification was achieved by transferring the supernatant to a fresh tube and centrifuging under the same conditions. Supernatant was then transferred to another tube and diluted 1:1 (v/v) with phosphate-buffered saline (PBS) containing 4% sarkosyl. Pierce Universal Nuclease for Cell Lysis (Thermo Scientific #88701) was added before heating the samples to 50˚C for 30 min.
All mb-PMCA experiments included eight replicates of unspiked NBH as a negative control for spontaneous formation of PrP res during mb-PMCA, a total of 42 replicates in 5.5 plates (2 replicates had incomplete PK digestion and, thus, were inconclusive). Because no extraction was performed on the water, unspiked NBH also served as a negative control for the water. In addition, to control for the possible effect of soil extracts on promoting formation of PrP res , we extracted Elliot silt loam (International Humic Substances Society, St. Paul, Minnesota, United States) and treated the extract in the same manner as the extracts from the mineral licks; two concurrent independent extracts were used to seed 4 replicates apiece and put through 2 rounds of mb-PMCA. Feces from a white-tailed deer negative for CWD were put through the fecal extract protocol and two rounds of mb-PMCA to serve as a negative control and validate the fecal extraction methods; two concurrent independent extracts were used to seed 4 replicates apiece and put through 2 rounds of mb-PMCA. For positive controls, we included the 1:3.1×10 4 and 1:1.2×10 17 dilutions of a 10% (w/v in PBS) brain homogenate of an orally inoculated, clinically affected CWD-positive white-tailed deer (96 GG) on every plate. We reliably detected prions in the 1:1.6×10 5 dilution of the CWD-positive brain homogenate after one round of mb-PMCA and in the 1:7.5×10 19 dilution after two rounds.

Contamination controls
We took several precautions to prevent inadvertent introduction of CWD prions into the field samples during sample collection and continuing through mb-PMCA testing. These precautions included changing gloves between collection of fecal samples; changing gloves between mineral licks; wiping the LaMotte soil sampling tool with 10% bleach before and after each mineral lick; discarding plastic bags used for soil and fecal collection after a single use; discarding 50 mL water collection tubes after a single use; double bagging of samples at the site; aliquoting stock solutions to single-use volumes; changing pipette tips after each use; wiping lab tools with 10% bleach before and after each use; using fresh disposable paper bench covering for every task; handling and loading plates with only one sample at a time and changing gloves between samples; loading plates with all wells capped except those being seeded with that replicate; and handling samples and loading plates in the following order: no seed negative controls, matrix extract negative controls, experimental samples, followed by positive controls.

Results
We detected CWD prions in soil samples, water samples, or both collected from nine of the 11 mineral licks following two rounds of amplification with mb-PMCA (Table 1). We limited the number of mb-PMCA rounds to two for soil and three for feces to minimize the possibility of de novo prion generation [47]. Soil from seven licks contained CWD prions. No pattern was apparent in the presence of prions in the upper vs. lower 1.27 cm of soil. We sampled water from nine mineral licks. Prions were detectable in the undisturbed water from four of the sites and in disturbed water from two of these sites. Two mineral licks (Sites 2 and 6) contained detectable CWD prions in both water and soil samples. The amounts of prions amplified from these soil and water samples was near the limit of detection for two rounds of mb-PCMA, likely due in part to co-extracted inhibitors of the PMCA reaction and incomplete extraction from soil particles. The detection of prions at 9 of 11 sites sampled, however, demonstrates widespread contamination of mineral licks in the CWD outbreak zone. The generally higher detection frequency of CWD prions in water samples relative to the corresponding soil samples suggests that either prion concentrations are higher in the water samples or that coextracted constituents from soil inhibited amplification by mb-PMCA.
At the mineral lick site with the highest detection of CWD prions in environmental samples (Site 6), we opportunistically sampled white-tailed deer fecal pellets. We detected CWD prions in six of the 10 fecal samples after three rounds of amplification by mb-PMCA. Of eight replicates tested for each fecal sample, one fecal sample had four positive replicates, three had two positive replicates, and two had a single positive replicate. Importantly, no false positives were produced in any of our negative control samples. No PrP res was detected after two or three rounds of amplification by mb-PMCA in negative control samples conducted when testing lick samples, which included at least one no-seed control, one negative soil extract control, and one no-seed fecal extract control from a CWD-negative white-tailed deer.

Discussion
Our results demonstrate that CWD-infected white-tailed deer deposit prions at mineral licks they visit. Although the mechanism of prion deposition is unknown, we suspect deposition of saliva by infected deer during ingestion of soil and water at mineral licks has the highest potential to facilitate indirect transmission to susceptible deer. Saliva from white-tailed deer infected with CWD contains on the order of 1-5 infectious doses (ID 50 ) per 10 mL as quantified by real-time quaking-induced conversion, where an ID 50 is the dose of CWD prions capable of infecting half of the transgenic mice expressing cervid prion protein [48]. Frequent visitation by infected cervids could allow mineral licks to become potential "hot spots" for indirect transmission of CWD [49]. Currently, little is known about the relative importance of direct contact and environmental routes of CWD transmission in free-ranging cervids [10]. Thus, how artificial and natural mineral licks contribute to current and future CWD infection in cervids and whether licks should be managed to control cervid use are important questions for further research.
Despite the relatively recent detection of CWD in Wisconsin (2001) and the moderate incidence of infection (6-19% prevalence in adult deer in the area sampled at the time of sample collection), our results suggest contamination of mineral licks in the CWD outbreak zone is widespread. This finding suggests that mineral licks may serve as reservoirs of CWD prions that contribute to disease transmission to susceptible animals. Although the levels of CWD prions in the samples analyzed appears low, we note that the association of prions with clay minerals often present at mineral licks can dramatically enhance disease transmission via the oral route of exposure [30][31]. For hamster-adapted scrapie prions binding to montmorillonite clay particles enhanced transmission by a factor of 680, however, an upper bound on the enhancement factor could not be assigned [30][31]. At present, the degree to which binding to clay mineral particles enhances CWD transmission to deer via the oral (or nasal) route of exposure is not known. Furthermore, repeated oral exposure to prions is associated with increased likelihood of disease transmission [50]. Differences in the sialyation status of Nlinked glycans between brain-derived and secreted/excreted PrP CWD may impact oral infectivity [51]. Cervid species that avoid interspecific contact make use of the same mineral lick sites [49], potentially leading to interspecies transmission. Mineral licks also attract livestock and other wildlife that supplement mineral intake via soil and water consumption, exposing these animals to CWD prions. Exposure of predators and scavengers to CWD prions via consumption of infected tissue has been previously documented [23]; our results suggest that environmental exposure of non-cervid animal groups can also occur via environmental routes.
We also detected CWD prions in fecal samples collected in proximity to a mineral lick, indicating that fecal excretion represents a route of CWD deposition into the environment with potential transmission to susceptible cervids [19]. Deposition of fecal pellets by whitetailed deer near bait sites increases with higher deer visitation [52] and similar patterns probably occur at mineral licks. Thus, increased local fecal deposition by CWD-infected deer likely contributes to increased environmental concentrations of prions in and around mineral licks.
Deer generally avoid consumption of feces [52]; however, the apparent long-term duration of prion infectivity in the environment [27][28][29], the enhanced disease transmission by soil-bound prions combined with the repeated visitation, long-term existence of and multi-generational use of mineral licks suggest the impact of concentrated environmental contamination on the dynamics of disease transmission warrants further investigation. Recent laboratory research indicates plants grown in prion-contaminated soil can accumulate prions [53]. Our data suggest that plants growing near contaminated mineral licks may warrant investigation as a source of prions for foraging animals. Areas where cervids congregate for mineral consumption, feeding and baiting sites, winter yarding, wallows [54] or other activities where CWD prions are deposited in the environment may also provide potential long-term reservoirs for transmission to cervid and non-cervid species.

Conclusions
We used mb-PMCA to detect CWD in soil and water from mineral licks naturally contaminated with prions and used by free-ranging deer, livestock, and non-cervid wildlife species. Detection of prions in environmental reservoirs represents an important first step in understanding the contribution of environmental transmission to CWD epizootics and potential for cross-species transmission. The present study characterized an environmental prion reservoir by (1) identifying an apparent "hot spot" of deposition and potential exposure to both cervid and non-cervid species; (2) indicating CWD prions shed by free-ranging cervids are present in areas of frequent use leading to environmental contamination and potentially plant uptake; and (3) motivating investigation of the exposure and susceptibility of non-cervid species to CWD contaminated soil, water, and plant materials. Future research should be directed at quantifying CWD prion concentrations at mineral licks and other areas where cervids congregate, determining the persistence of prion infectivity at these sites, delineating spatial-temporal patterns of environmental prion deposition and accumulation, and assessing consumption by susceptible animals. Identifying additional environmental reservoirs of CWD prions and determining the contributions of direct and indirect transmission over the course of CWD outbreaks represent key aims in advancing understanding of long-term CWD infection dynamics. The indicated extraction solution was added (200 μL) to each of the soil pellets and vortexed (2 h, 1,200 RPM, room temperature). Soil particles were sedimented (1000g, 10 min), and the supernatants were retained. Extractions were done in triplicate, and 100 μL of each extract and the water rinses were used to assay the absorbance at λ = 465. Absorbances were compared to dilution series of Elliot soil humic acid in each of the above extraction solutions, and the concentration of NOM was estimated. Shown are the mean NOM concentrations for the three replicates with the standard deviations. (B) Extraction of PrP CWD from soil. PK-treated 10% brain homogenate from a CWD-positive white-tailed deer (40 μL) was adsorbed to Elliot soil (25 mg) in ultrapure water (100 μL, 24 h) followed by a 2-h desorption step in 100 μL water (to remove any non-adsorbed unbound PrP CWD ). The sorbed PrP CWD was extracted at room temperature with 200 μL of the indicated extraction solution and analyzed by immunoblotting. Abbreviations: A-G, extraction solutions (see descriptions above); M, molecular mass marker; mAb, monoclonal antibody; rinse, water rinse; S, supernatant from binding experiment. were mixed with 2 μL from the second 5-fold dilution of 10% brain homogenate (BH) from an end-staged CWD-positive wt/wt deer. Normal brain homogenate (NBH; 90 μL) was added with two Teflon 1 beads in a 200-μL thin-walled PCR tube. Samples were sonicated for 96 cycles (30 s sonication with 27:30 incubation at 37˚C between sonications). (B) Elliot soil or Defore soil (25 mg) was extracted using 100 μL of the indicated extraction solution. An aliquot (8 μL) of these extracts were mixed with 2 μL from the second 5-fold dilution of 10% BH from an end-stage CWD-positive wt/wt deer. NBH (90 μL) was added with two Teflon 1 beads in a 200-μL thin-walled PCR tube. Samples were sonicated for 96 cycles (30 s sonication with 27:30 incubation at 37˚C between sonications). Extraction solutions are described in the text. Proteinase K (PK) resistant prion protein was detected using Western blot with antibodies 8G8 and BAR224. (TIF)

S3 Fig. Effect of humic and fulvic acid detection of PrP CWD by PMCAb.
Prions (10% CWD brain homogenate) were diluted using four serial fivefold dilutions in normal brain homogenate (NBH) to obtain the seed dilution used in this experiment. The seed dilution (2 μL) was transferred into 36 μL NBH with saponin, and 2 μL humic or fulvic acid (dissolved in ultrapure water) was added to achieve a total mass of 5, 2.5, 1. Samples, normal brain homogenate (NBH; 90 μL), and two Teflon 1 beads were added to a 200-μL thin-walled PCR tubes containing NBH. Samples were sonicated for 96 cycles (0.5 min sonication with 27.5 min incubation at 37˚C between sonications). Proteinase K (PK) resistant prion protein was detected using immunoblotting with antibodies 8G8 and BAR224. (TIF)