Fate of Prions in Soil: A Review

Abstract

Prions are the etiological agents of transmissible spongiform encephalopathies (TSEs), a class of fatal neurodegenerative diseases affecting humans and other mammals. The pathogenic prion protein is a misfolded form of the host-encoded prion protein and represents the predominant, if not sole, component of the infectious agent. Environmental routes of TSE transmission are implicated in epizootics of sheep scrapie and chronic wasting disease (CWD) of deer, elk, and moose. Soil represents a plausible environmental reservoir of scrapie and CWD agents, which can persist in the environment for years. Attachment to soil particles likely influences the persistence and infectivity of prions in the environment. Effective methods to inactivate TSE agents in soil are currently lacking, and the effects of natural degradation mechanisms on TSE infectivity are largely unknown. An improved understanding of the processes affecting the mobility, persistence, and bioavailability of prions in soil is needed for the management of TSE-contaminated environments.

Transmissible spongiform encephalopathies (TSEs), or prion diseases, comprise a class of fatal, neurodegenerative diseases affecting a variety of mammals. These diseases include bovine spongiform encephalopathy (BSE) in cattle, scrapie in sheep and goats, chronic wasting disease (CWD) in North American members of the deer family (cervids), transmissible mink encephalopathy (TME) in farmed mink, and Creutzfeldt-Jakob disease and kuru in humans. The etiological agents, termed prions (proteinaceous infectious particles lacking nucleic acid; Prusiner, 1998), have not been fully characterized, but available evidence indicates that a misfolded form of the host-encoded prion protein (PrP) constitutes the primary, if not sole, component of the prion. The normal, benign cellular form is designated PrP^C; the pathologically misfolded form is denoted PrP^TSE. Prion diseases are characterized by long asymptomatic incubation periods, spongiform degeneration in the central nervous system, and accumulation of PrP^TSE.

Scrapie and CWD are the only TSEs that appear to be environmentally transmissible, which may reflect disease-specific differences in the levels of infectious agent released into the environment or higher susceptibility of sheep, goats, and cervids to infection by environmental routes (Pedersen and Somerville, 2011). Scrapie has been recognized in sheep for over 250 yr, and anecdotal evidence for environmental transmission is provided by several reports of naïve sheep contracting scrapie after grazing on fields once occupied by infected animals (Greig, 1940; Palsson, 1979; Georgsson et al., 2006). Chronic wasting disease was first recognized in the late 1960s in mule deer at a wildlife research facility in northern Colorado. The disease has been subsequently found in captive and free-ranging deer, elk and moose in the United States and Canada. A study exposing healthy animals to paddocks containing CWD-infected carcasses decomposed in situ or residual excreta from CWD-infected animals demonstrated environmental transmission of CWD in mule deer (Miller et al., 2004).

The relevant routes of environmental TSE transmission remain to be clarified, yet several lines of evidence suggest that soil may serve as a reservoir of TSE infectivity (Schramm et al., 2006). Prions may enter soil through urinary, salivary, or fecal shedding from infected animals or decomposition of diseased carcasses (Tamgüney et al., 2009; Gregori et al., 2008; Mathiason et al., 2006; Miller et al., 2004). Transmissible spongiform encephalopathy infectivity is known to persist in soil for years (Seidel et al., 2007; Brown and Gajdusek, 1991; Georgsson et al., 2006). Animals may ingest contaminated soils intentionally to supplement mineral intake, incidentally during feeding or grooming, or indirectly (e.g., clearance from lungs) (e.g., Weeks, 1978; Fries, 1996). Together, these studies suggest that soil represents a plausible route of environmental TSE transmission; however, several questions remain regarding the fate and persistence of prions in soil.

In this review, we discuss factors influencing the attachment of prion protein to soil constituents and potential implications for environmental TSE transmission. Association of prions with soil particles may impact TSE infectivity by affecting the bioavailability or persistence of these agents in the environment. To date, the concentration of prions in environmental media in areas where TSEs are endemic is largely unknown due to the limited ability to detect TSE agents in or extracted from environmental samples. Natural biotic and abiotic mechanisms of protein degradation may impact levels of TSE infectivity in the environment. Significant interest exists for evaluating these degradation pathways (e.g., degradation by soil microorganisms, extracellular proteases, or inorganic oxidants) as a means to reduce or eliminate TSE infectivity in contaminated soils.

Structure of the Pathogenic Prion Protein

The infectious agent in prion diseases consists primarily, if not solely, of a misfolded form of the prion protein. Disease propagation appears to occur via a templating mechanism in which PrP^TSE catalyzes the misfolding of PrP^C (Prusiner, 1991; Jarrett and Lansbury, 1993). The pathologic conformational isomer possesses the same primary sequence and covalent post-translational modifications as does the innocuous form; the only difference detected to date is the folding of the protein. The mature prion protein (i.e., after post-translational modification) consists of approximately 208 amino acids with two consenus sites for N-linked glycosylation at Asn 180 and Asn 196. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) of denatured PrP reveals three bands representing di-, mono-, and unglycosylated protein. The composition of the N-linked glycans is heterogeneous; at least 52 different sets of oligosaccharides have been identified (Stimson et al., 1999; Rudd et al., 1999; Rudd et al., 2001). When both Asn sites are occupied, more than 400 glycoforms are expected based on the diversity of known oligosaccharide structures (Endo et al., 1989). Sialic acid residues impart significant charge heterogeneity with individual PrP molecules exhibiting isoelectric points ranging from 4.6 to 7.9 (Bolton et al., 1985). The prion protein is usually post-translationally modified with a glycosyl-inositol phospholipid (GPI) membrane anchor attached at Ser 231 (Stahl et al., 1987). The GPI anchor can also be sialylated (Stahl et al., 1992).

The majority of PrP^TSE in diseased brain tissue is isolated as insoluble amyloid fibrils typically 50 to 300 nm in length and 10 nm in diameter (Fig. 1). Recent studies indicate that smaller subfibrillar aggregates of 14 to 28 PrP^TSE molecules possess the highest specific infectivity (i.e., infectivity per unit mass; Silveira et al., 2005). Both the folding and the aggregation states of PrP^TSE convey distinct biological and physicochemical properties to the protein. Prions exhibit an extraordinary resistance to inactivation by ultraviolet (UV) and ionizing radiation, exposure to proteases or chemical disinfectants, and heat treatments (Taylor, 1999, 2000; Taylor and Diprose, 1996; Brown et al., 2004). Treatment of PrP^TSE with proteinase K (PK) cleaves the N-terminal ~90 amino acid residues, leaving a truncated form of the protein designated PrP27-30. Prions also display distinct strains defined by differences in clinical symptoms, incubation times, and pathology (Bruce, 2003; Aguzzi et al., 2007; Morales et al., 2007). Strain differences appear to be enciphered in the conformation of PrP^TSE (Telling et al., 1996).

Fig. 1.

Transmission electron microscopy (TEM) image of PrP^TSE amyloid fibrils (HY strain of hamster-adapted transmissible mink encephalopathy) negatively stained with phosphotungstic acid. (Image kindly provided by Kurt H. Jacobson and Randy Massey.)

High-resolution structural information on PrP^TSE has not been obtained by nuclear magnetic resonance spectroscopy (NMR) or X-ray crystallography due to problems with insolubility, aggregation, and heterogeneity of purified samples. Circular dichroism and infrared spectroscopy data indicate that PrP^TSE has dramatically increased β-sheet content relative to PrP^C (Pan et al., 1993). Experimental information on PrP^TSE structure is limited to secondary structural analysis, electron crystallography of two-dimensional PrP^TSE crystals, antibody epitope mapping studies, H/D exchange experiments, covalent cross-linking studies, and X-ray fiber diffraction data. Two structural models of PrP^TSE determined in silico and constrained by low-resolution experimental data are consistent with most experimental data (Fig. 2; DeMarco et al., 2006; Govaerts et al., 2004). These models indicate the formation of trimers, which stack to form amyloid fibrils (Govaerts et al., 2004). Asn-linked glycans are predicted to lie at the periphery of the trimers and significantly impact accessibility of the polypeptide surface (DeMarco and Daggett, 2009). A third proposed PrP^TSE structure was recently published based on site-directed spin labeling and electron paramagnetic resonance spectroscopy studies of recombinant PrP amyloid fibrils (Cobb et al., 2007). Taken together, structural models provide an initial step toward understanding the conformational changes associated with TSEs, yet further investigation is needed to evaluate the contribution of specific structural elements to infectivity.

Fig. 2.

Model structures of PrP^TSE monomers and trimers. (A) Spiral model, derived from molecular dynamic simulations at low pH. (B) β-Helix model, derived by threading the PrP sequence through a known β-helix fold. Trimeric representations of the (C) spiral and (D) β-helix models. The (E) spiral and (F) β-helix model with the protein-only portion highlighted (gray, space-filling) and biologically relevant glycans (cyan, space-filling) attached to the appropriate glycosylation sites. Images C–F are on the same scale (bar = 50 Å). (Used by permission of the American Chemical Society, from DeMarco et al. [2006] Biochem. 45, Fig. 1, p. 15574.)

Sources of Prion Protein Used in Environmental Fate Studies

Isolation of PrP^TSE from Infected Animal Tissue

Discussion of the sources of PrP^TSE used in environmental fate studies is warranted before presenting the results of such investigations. Infected brain tissue is the most common source of bona fide PrP^TSE since it contains the highest concentrations of TSE agent (Kimberlin and Walker, 1977). Most studies have used rodent-adapted prion strains, which are largely similar to environmentally relevant TSE agents in terms of PrP primary and secondary structure, aggregation, and resistance to common inactivation techniques. However, strain-specific differences in PrP^TSE conformation, structural stability (Safar et al., 1998), and degree of glycosylation (Somerville, 1999) have been reported.

Methods for the partial purification of PrP^TSE from brain homogenate involve a series of detergent solubilization and centrifugation steps (Bolton et al., 1987; Caughey et al., 1991). The resulting prion preparations lack PrP^C but contain other copurifying biomolecules including carbohydrates, nucleic acids, proteins, and lipids (Stahl et al., 1987; Aiken et al., 1990; Kellings et al., 1993; Stahl et al., 1993; Moore et al., 2010). Purification increases PrP^TSE fibrillization producing protein aggregates that are larger than those present in infected brain tissue (McKinley et al., 1991). The effect of aggregation on PrP^TSE stability has not been formally evaluated, but changes in aggregation state may influence the extent to which the protein resists degradation. Experiments analyzing environmental persistence of prions should therefore be modeled with homogenized or intact infected tissues as opposed to purified PrP^TSE preparations. In contrast, examining molecular interactions between PrP^TSE and soil particle surfaces, such as the extent and rate of adsorption, requires the use of purified PrP^TSE to minimize contributions from other sample constituents.

The N-terminus of PrP^TSE is susceptible to proteolytic degradation. Pathogenic prion protein released into the environment from decomposing tissues might therefore be expected to exist primarily in the N-terminally truncated form (Saunders et al., 2008). However, species and/or strain differences in N-terminal truncation during tissue decomposition may exist. The N-terminus of PrP^TSE is not necessary for TSE infection (Prusiner et al., 1984) but may influence mechanisms of protein attachment to soil particle surfaces (Cooke et al., 2007).

Recombinant Prion Protein

Some studies intending to model the fate of prions in the environment have employed prokaryotically expressed recombinant prion protein (recPrP). Recombinant analogs expressed in prokaryotes lack glycosylation and the GPI anchor, which significantly affect the physicochemical characteristics of the protein. Nonglycosylated ovine recPrP exhibits an isoelectric point (IEP) between 9 and 10 (Vasina et al., 2005), whereas an average IEP of 4.6 was found for PrP^TSE aggregates (Ma et al., 2007). The N-linked glycans and GPI anchor of PrP expressed in mammalian systems contain sialic acid residues, resulting in the lower IEP of bona fide PrP^TSE (Bolton et al., 1985; Stahl et al., 1992). Under mildly acidic conditions, α-recPrP (proteolytically labile recPrP with a PrP^C–like conformation) folds into PrP^TSE–like β-sheet rich oligomers (termed β-recPrP) with increased resistance to PK digestion (Swietnicki et al., 2000). Nonetheless, β-recPrP is more labile than PrP^TSE and therefore has limited utility for understanding prion persistence in the environment.

PrP^TSE Detection

Methods for prion detection can be differentiated in terms of quantitation of prion protein and measurement of TSE infectivity. The amount of protease-resistant prion protein usually correlates with infectivity, although infectious agent may be present in the absence of detectable PrP^TSE (Lasmezas et al., 1997; Barron et al., 2007). Direct detection of prions is achieved only by using animal- or cell-based bioassays. Animal bioassays represent the most definitive test for TSE infectivity, and results are commonly expressed in terms of infectious units (IU₅₀). One IU₅₀ represents the amount of infectious agent required to infect half the population of test animals used in a study. Quantitation is achieved by either end-point titration or on the basis of the strong correlation between dose and length of the incubation period (Prusiner et al., 1980). End-point titration is more sensitive with a limit of detection of approximately 0.2 amol of PrP^TSE (i.e., 1 IU₅₀) (Lax et al., 1983). Practical limitations of animal bioassays include the long incubation periods required (e.g., 90–200 d for hamsters; 200–500 d for mice) and the expense of maintaining large cohorts of animals.

Prions have been detected using neuroblastoma cell-based culture assays (Klöhn et al., 2003; Mahal et al., 2007). These assays are significantly more rapid and less expensive than animal bioassays and achieve similar sensitivity. However, cell-line resistance to many important prion strains has hampered general application of these assays in TSE research. Cell-based infectivity measurements are compatible with prions bound to surfaces including soil particles (Genovesi et al., 2007) and stainless steel wires (Edgeworth et al., 2009). Measurement of prions in the presence of soil particles or natural organic matter (NOM) is critical for the development of methods to detect prions in or extracted from environmental matrices. Direct quantitation of prions in soil by animal bioassays is complicated by enhanced disease transmission by prions bound to at least some types of mineral, soil, and dietary particles (Johnson et al., 2007; Johnson et al., 2011).

Standard techniques for protein detection such as Western blotting and enzyme-linked immunosorbent assay (ELISA) are commonly used to measure PrP^TSE. Under denaturing conditions, immunodetection cannot discriminate between PrP^TSE and PrP^C. Samples are often treated with a protease (e.g., PK, thermolysin) to degrade PrP^C (and the protease-sensitive fraction of PrP^TSE) (Tzaban et al., 2002; Owen et al., 2007) before immunodetection. Although immunoassays are rapid and specific for PrP, they suffer from low sensitivity (i.e., detection limits of 2000–20,000 amol and 200 amol for Western blot and ELISA, compared with an estimated 0.2 amol for 1 IU₅₀) and fail to measure TSE infectivity. More sensitive PrP^TSE quantitation has been achieved by liquid chromatography–mass spectrometry (LC–MS) (Onisko et al., 2007). This method detects a signature PrP peptide following reduction, purification, and tryptic cleavage of PrP^TSE. Despite reported detection limits of 20 to 30 amol for pure standards, LC–MS requires considerable amounts of infectious brain material due to losses during protein purification. Increased sensitivity may be provided with signature peptides lacking methionine that can be easily oxidized and represents a potential source of error in quantitative mass spectrometry (Sturm et al., 2010). Further development is needed for application of LC–MS to complex biological and environmental samples.

Protein misfolding cyclic amplification (PMCA) relies on the structural transition of PrP^C to PrP^TSE catalyzed by small quantities of PrP^TSE in a manner conceptually similar to polymerase chain reaction (Saborio et al., 2001; Castilla et al., 2005). The amount of PrP^TSE is increased by repeated cycles of sonication (to disrupt PrP^TSE aggregates) and incubation using PrP^C from healthy brain homogenate as a substrate. Protein misfolding cyclic amplification has the lowest reported PrP^TSE detection limits corresponding to approximately 26 molecules or approximately 4 × 10⁻⁵ amol (assuming a molecular mass of 30 kDa for PrP^TSE) (Saá et al., 2006). Similar techniques have emerged using recPrP as a substrate including the amyloid seeding assay (Colby et al., 2007), quaking induced conversion (Atarashi et al., 2008) and real-time quaking induced conversion (Wilham et al., 2010).

Conversion assays represent a promising approach for sensitive detection of prions in naturally contaminated environmental media. Recently, detection of PrP^TSE by PMCA was reported in one of two stream water samples collected from a CWD endemic area (Nichols et al., 2009). Although the report is intriguing, further measurements are needed to substantiate this result due to the propensity for false positives with PMCA and the potential impact of dissolved organic matter on PMCA. Currently, only two studies have used PMCA to detect PrP^TSE in soils experimentally spiked with infected brain homogenate. Seidel et al. (2007) extracted hamster PrP^TSE from soil with 1% SDS and found that at least a fraction extracted remained catalytically active to seed PMCA. The authors did not report the exact level of amplification, which was presumably low. Nagaoka et al. (2010) used soil-bound PrP^TSE to seed PMCA, which provided 10³ to 10⁴ higher sensitivity than methods involving PrP^TSE extraction followed by Western blot detection. Effective amplification required small quantities of soil (8 μg), and no amplification was observed in the presence of more soil (80–800 μg). Further development of PMCA is needed for detection of PrP^TSE from environmentally relevant prion strains (e.g., CWD, scrapie). The impact of specific soil characteristics (e.g., texture, mineralogy, organic carbon content) on amplification should also be determined.

Protein extraction and sample cleanup complicate application of prion detection methods to environmental samples. Direct detection of prion protein in soil has been proposed to avoid the effects of inefficiencies in prion protein extraction on detection (Genovesi et al., 2007; Saunders et al., 2009a,b). This approach uses bovine serum albumin to saturate soil particle surfaces not occupied by PrP^TSE (or other proteins) and then visualizes soil-bound PrP^TSE by immunodetection. Quantitation is limited by accessibility of the antibody binding epitope, a narrow linear range and low sensitivity with PrP^TSE detection limits similar to those of Western blotting.

Attachment of PrP^TSE to Soil Particle Surfaces

Associations between PrP^TSE and soil particle surfaces are expected to impact the fate of pathogenic prion protein and TSE infectivity in the environment. The interaction of proteins with surfaces is commonly described as adsorption. We use this term when referring to the interaction of single protein molecules with surfaces. Aggregates of PrP^TSE are in the colloidal regime. We therefore use the term attachment to describe their interaction with particle surfaces.

Proteins exhibit a pronounced tendency to accumulate at solid–liquid interfaces under most conditions. This is due to the diverse properties of constituent amino acids and the rotational mobility along the polypeptide chain (for reviews, see Norde, 1986, 2003). Factors influencing protein adsorption to well-defined polymer and metallic substrates have been examined extensively in the context of biomaterials research. In contrast, relatively few studies have examined protein adsorption to soil constituents with the exception of investigations on the function of extracellular enzymes in soil and those on the persistence of Bacillus thuringiensis toxins (Cry proteins) released from transgenic crops. These studies have found that proteins bind strongly with most soil particle surfaces, and adsorption can induce conformational changes, alter enzyme activity, and protect proteins from degradation (Quiquampoix, 2000). Potential consequences of prion attachment to soil particle surfaces include: limited transport of PrP^TSE in porous media; transport with particles entrained in overland flow; protection from enzymatic, chemical, or physical degradation; altered bioavailability or uptake of soil-bound prions by animals; and changes in PrP^TSE conformation or aggregation state that may impact TSE infectivity. Below, we discuss the current state of knowledge concerning PrP^TSE attachment to soil particle surfaces and its implications for TSE infectivity in the environment.

Extent and Kinetics of PrP^TSE Attachment to Soil Particle and Mineral Surfaces

The majority of studies focused on determining the extent of PrP^TSE attachment to soil and mineral particles have used solution depletion experiments. Solution conditions including choice of buffer, pH, and ionic strength impact protein adsorption (Docoslis et al., 2001). The ionic strength of the interstitial water in many soils ranges from 0.003 to 0.016 M, and the dominant ions are typically Na⁺, K⁺, Mg²⁺, Ca²⁺, NH₄⁺, NO₃⁻, HCO₃⁻, Cl⁻, and SO₄²⁻ (Helmke, 1999). The soil solution is often approximated with 0.001 to 0.010 M in CaCl₂ or NaHCO₃ solutions. Investigation of the interaction of a protein with soil particles requires discrimination between particle-associated and free protein. The aggregation state of PrP^TSE complicates separation of attached from free prion protein because larger, unbound PrP^TSE aggregates may be separated with smaller soil particles (thereby overestimating the degree of attachment) if conditions are not found to minimize this. Most solution depletion experiments examining PrP^TSE attachment to soil and mineral particles have used centrifugation to separate soil-bound PrP^TSE from unbound protein without considering the potential cosedimentation of large, unbound PrP^TSE aggregates. Effective separation of clay particle-bound from unbound PrP^TSE has been achieved (to the limit of detection system used) by clarifying PrP^TSE preparations (to remove larger aggregates) and sucrose cushion centrifugation (Johnson et al., 2006). The attachment of PrP^TSE to soil particle surfaces has also been measured by modeling the transport of PrP^TSE in porous media; these experiments are described in the following section.

In light of the difficulty in separating particle-associated and free PrP^TSE in solution depletion experiments, interest exists for the application of in situ techniques for measuring PrP^TSE attachment. Vasina et al. (2005) used a flow cell and γ-counter to measure attachment of ¹²⁵I-radiolabeled β-recPrP to mica under laminar flow conditions. Disadvantages of this technique include the limited ability to radiolabel PrP^TSE aggregates and the potential impact of extrinsic labels on PrP attachment. Label-free, surface-sensitive techniques including optical waveguide lightmode spectroscopy, surface plasmon resonance, ellipsometry, and quartz crystal microbalance with dissipation monitoring have been used extensively to examine protein adsorption. To date, in situ methods have not been applied to measure PrP^TSE attachment to representative soil surfaces.

Solution depletion experiments have shown that PrP^TSE readily binds to soil minerals (viz. montmorillonite, kaolinite, and quartz) and whole soils varying in composition (Johnson et al., 2006; Leita et al., 2006; Cooke et al., 2007; Ma et al., 2007; Maddison et al., 2010). These studies found that PrP^TSE attachment occurs rapidly, with complete attachment reported to occur within 2 h (Johnson et al., 2006). When expressed on a surface area basis, montmorillonite and quartz microparticles exhibited the highest PrP^TSE adsorption capacities (~2–6 mg_protein m_mineral⁻²), whereas that of kaolinite was smaller by a factor of 25 (Table 1). These results suggest that mineral surface properties contribute to differences in PrP^TSE binding. Montmorillonite was also found to have a large sorption capacity for β-recPrP with complete adsorption observed at 1:1 (w/w) ratio of β-recPrP:montmorillonite (Rigou et al., 2006). Similar to PrP^TSE, β-recPrP adsorbed rapidly (within 1 h) to montmorillonite and muscovite mica surfaces (Rigou et al., 2006; Vasina et al., 2005).

Table 1.

Reported PrP^TSE adsorption capacities for common soil minerals (from Johnson et al. [2006] PLoS Pathog. 2, Table 1).

Mineral†	Estimated binding (sorbent mass basis)‡ capacity	Estimated binding capacity (sorbent surface area basis)‡
	μg protein mgmineral−1	mgprotein mmineral−2
Mte	87–174	2.8–5.7
Kte	1.7–2.6	0.15–0.22
Qtz	13.6–27.1	2.7–5.4

^†Kte, kaolinite; Mte, montmorillonite; Qtz, quartz.

^‡Protein concentration determined by Bradford assay; PrP^TSE concentration was taken as 87% of total protein (Silveira et al., 2005). Reported adsorption capacities represent upper estimates, as the fraction of PrP^TSE in clarified preparations may have been lower. For Mte, binding capacity was based on the external (N₂–accessible) surface area (see Johnson et al. [2006] for discussion).

A recent report of incomplete PrP^TSE attachment following protein–soil contact times of several weeks contrasts with all other published studies (Saunders et al., 2009b). This study utilized brain homogenates as the PrP^TSE source. Several experimental design problems hinder confident interpretation of the study’s results. Specifically, sedimentation conditions appear unlikely to have been sufficient to discriminate particle-associated from “free” PrP^TSE; PrP^C and PrP^TSE were not differentiated; controls to monitor changes in brain homogenate composition over time were lacking; and solution conditions (phosphate buffered saline [PBS], ionic strength = 0.15 M) deviated considerably from those encountered in most environments (Sposito, 1989). Furthermore, PBS can interfere with protein sorption (Docoslis et al., 2001).

Because prions probably enter the soil in biological fluids or tissues containing a mixture of different biomolecules (e.g., lipids, other proteins, nucleic acids), understanding the extent to which matrix components affect PrP^TSE adsorption is important for accurately modeling natural systems. Maddison et al. (2010) reported that attachment of PrP^TSE to soil was complete within 24 h in experiments using scrapie-infected sheep brain and BSE-infected bovine brain homogenates as prion sources. Similarly, Rigou et al. (2006) found that β-recPrP was completely attached even in the presence of a large excess of serum proteins (500:1 10% fetal calf serum: β-recPrP [w/w]). In contrast, Johnson et al. (2007) noted diminished PrP^TSE attachment to montmorillonite in the presence of brain homogenate which may have been attributable to competing macromolecules. The impact of potentially competing adsorbates on PrP^TSE attachment to soil constituents under well-controlled conditions warrants further research. However, considering the persistence of PrP^TSE and the relative excess of sorption sites in natural soils, competitive adsorption may have little effect on the long-term fate of prions in soil.

PrP^TSE Association with Natural Organic Matter in Soil

Natural organic matter (a heterogeneous mixture of organic molecules resulting primarily from the microbial decomposition of plant material) is expected to substantially impact PrP^TSE attachment to soil particle surfaces. Protein–NOM complex formation has been suggested on the basis of observations of simultaneous extraction of NOM and soil enzymes including urease, diphenyl oxidases, proteases, and hydrolases (Boyd and Mortland, 1990). These extracts contained enzymatic activity that in some cases displayed increased resistance to proteolytic or thermal degradation. The precise nature of associations between soil enzymes and organic matter remains unclear.

Relatively few studies have examined the interaction of proteins with NOM. Most used humic acid, an operationally defined fraction of NOM molecules that are soluble in base but insoluble under acidic conditions. Tan et al. (2008, 2009) examined the interactions between lysozyme and purified Aldrich humic acid and reported the formation of large protein-humic acid heteroaggregates (average hydrodynamic radius of 2000–3500 nm). Electrostatic attraction may promote this complexation because lysozyme (IEP = 10.4) and humic particles carry opposite net charge at neutral pH. Little information exists on the complexation of net negatively charged proteins with humic substances, yet this interaction may be mediated by positively charged patches on the protein surface or by hydrophobic interactions. Hsu and Hatcher (2005, 2006) used two-dimensional NMR to demonstrate noncovalent interactions as well as covalent bonding between synthetic ¹⁵N-labeled peptides and humic acids varying in aliphatic and aromatic carbon content, suggesting that protein–humic acid interactions may involve the formation of hydrogen or covalent bonds. Natural organic matter can interfere with common protein detection methods including total protein assays (i.e., Bradford assay) and Western blotting (Pedersen et al., 2009; Whiffen et al., 2007), complicating clear interpretation of protein–humic interactions (vide supra).

To date, the interaction of PrP^TSE with NOM has not been reported; three studies have examined the influence of NOM or NOM surrogates on recPrP adsorption. Pucci et al. (2008) used low temperature ashing to remove layers of NOM from whole soils and found reduced α-recPrP sorption to soils ashed in this manner, suggesting that NOM in soil provides sorption sites for α-recPrP. The amount of soil-bound recPrP was estimated from the intensities of the Amide I and II bands in photoacoustic-Fourier transform infrared spectra, or by measuring solution depletion of α-recPrP by a modified Lowry assay (controls showed no interference in protein signal from soil particles). Polano et al. (2008) used UV-visible spectroscopy to measure α-recPrP adsorption to clay, humic–clay complexes, and humic acid. The humic coating apparently enhanced α-recPrP binding, but the potential for humic substances to interfere with α-recPrP quantification at λ = 280 nm was not discussed. In a study using polymerized catechol as a surrogate for humic substances, the formation of α-recPrP-polycatechol complexes was suggested on the basis of the loss of characteristic protein peaks in UV-visible and Fourier transform infrared spectra of the supernatant following protein addition (Rao et al., 2007). These complexes were stable against extraction by 0.1 M phosphate pH 7 to 8.5, 1% sarkosyl, and 10% v/v propanol. Overall, studies using α-recPrP have found that organic matter significantly contributes to sorption. Further investigation is needed to assess the influence of NOM on PrP^TSE adsorption to soil. Differences may be expected due to the altered conformation, aggregation, glycosylation, glypiation, and lower IEP of the infectious protein.

Interactions between PrP^TSE and Soil Particle Surfaces

Prion protein attachment to soil particle surfaces is expected to involve electrostatic, hydrophobic, van der Waals contributions, and possibly conformational changes. Proteins can adsorb to mineral surfaces under globally repulsive electrostatic conditions, indicating local attractive electrostatic interactions (i.e., interaction with specific amino acid residues on the protein surface bearing charge opposite that of the global protein charge) or the contribution of hydrophobic effects (Arai and Norde, 1990; Norde, 2008). Structural rearrangements can constitute an important driving force for adsorption due to the entropy gain associated with protein unfolding on surfaces. Entropic effects also involve reorganization of water molecules surrounding the protein and surface; adsorption to hydrophobic sorbents may be driven largely by the resulting dehydration of the surface. At present, the mechanisms underlying PrP^TSE attachment to soil particle surfaces remain poorly understood. Insight into these mechanisms is expected to aid predictions of prion transport and bioavailability in natural and engineered environments, and the development of efficient extraction methods to recover PrP^TSE from soils for detection and decontamination efforts.

The strong attachment of PrP^TSE to soil particle surfaces has been evidenced by its limited desorption under a variety of conditions. Extraction of PrP^TSE from soil particles typically requires anionic detergents. Reported effective extraction conditions include SDS-PAGE sample buffer (100 mM Tris, 7.5 mM EDTA, 100 mM DTT, 350 mM SDS, pH 8.0) at 100°C (Johnson et al., 2006), 1% N-lauroyl sarcosine in 100 mM phosphate buffer at pH 7.0 with 40 μg mL⁻¹ PK (Cooke et al., 2007), and 1% SDS in water at room temperature (Seidel et al., 2007). Many treatments found effective to extract other proteins from mineral surfaces (e.g., pH extrema, PBS, guanidinium, urea; Docoslis et al., 2001) have been found ineffective in removing PrP^TSE from montmorillonite clay surfaces (Johnson et al., 2006). Attachment to particles in fine-textured soils and clay minerals often results in low PrP^TSE recoveries. The same is true of recPrP (Cooke et al., 2007; Rigou et al., 2006). Decreased recovery of PrP^TSE at longer contact times has been observed (Maddison et al., 2010) and may reflect protein degradation, increase in the strength of interaction with soil particles over time (Hatzinger and Alexander, 1995; van Oss et al., 2001), entrapment in particle-associated NOM, diffusion into small pores in the soil matrix, or the formation of hydrogen or covalent bonds with humic substances (Hsu and Hatcher, 2005, 2006). The extent to which these processes may contribute to the persistence of prions in soil is unclear at present.

Solution pH and ionic strength influence protein adsorption. Proteins typically exhibit maximal adsorption to negatively charged surfaces near the IEP of the protein. At pH values below IEP, decreased adsorption may result from the protein unfolding at the surface, thereby reducing the quantity of protein required for saturation. Above the protein IEP, the mechanism of attachment likely reflects a competition between repulsive electrostatic and attractive hydrophobic interactions between the protein and surface. Additionally, lateral electrostatic repulsion of like-charged protein molecules may reduce adsorption at pH values above and below the protein IEP (Quiquampoix et al., 1993; Norde and Lyklema, 1978; Suzawa and Shirahama, 1991). The influence of solution pH and ionic strength on PrP^TSE attachment to quartz sand has been reported (Ma et al., 2007). Quartz surfaces are negatively charged in solutions at environmentally relevant pH (i.e., pH 4–9). Maximal attachment to quartz was found at pH 4, near the average IEP for PrP^TSE aggregates. Attachment of PrP^TSE to quartz increased with increasing ionic strength, congruent with the shielding of electrostatic repulsion between the protein and quartz surface (Ma et al., 2007). Attachment under globally electrostatically repulsive conditions suggests the importance of oppositely charged patches on the protein surface.

Adsorption of β-recPrP to muscovite mica has similarly been studied as a function of solution pH (Vasina et al., 2005). The β-recPrP (IEP = 9.77 due to the absence of N-linked glycans and GPI anchor, which contain sialic acid residues in mammalian PrP) carried a net positive charge over the pH range examined (pH = 4–9), and its strong attachment to negatively charged mica surfaces under all solution conditions was attributable to electrostatic interactions. Revault et al. (2005) found that recPrP adsorbed to montmorillonite exhibited a gain in intramolecular β-sheet content and a concomitant loss of α-helix content in deuterated buffers with pD > 4. At pD > 7, the secondary structure of recPrP in solution resembled that of the adsorbed protein. Spectra of adsorbed β-conformers differed from those of soluble β-oligomers (precursors to a fibril state) suggesting adsorption-induced conformational change on montmorillonite surfaces would not lead to a pathogenic (i.e., PrP^TSE) conformation.

The N-terminal domain (approximately residues 23–90) of PrP^TSE appears important for sorption, at least to some clay mineral surfaces. Both recPrP and PrP^TSE desorbed from montmorillonite and clay-rich soils were cleaved at the N-terminus (Johnson et al., 2006; Rigou et al., 2006; Cooke et al., 2007; Maddison et al., 2010). Cooke et al. (2007) found that PK digestion increased PrP^TSE desorption from a sandy clay loam but not from the loamy sand soils examined. The N-terminus of PrP^TSE and recPrP is flexible and contains 10 basic amino acid residues. These positively charged residues may interact strongly with negatively charged clay surfaces, even when global electrostatics are repulsive. The N-terminus may intercalate into the interlayer spaces of montmorillonite. Although X-ray diffraction failed to provide evidence of penetration of the inter-layer spaces by PrP^TSE (Johnson et al., 2006), intercalation of other proteins in the interlayer spaces of smectite clays has been reported (Violante et al., 1995; Harter and Stotzky, 1973).

Predicting the attachment of PrP^TSE to soil particles is complicated by the limited information available about its structure and the variable number and composition of oligosaccharides. The N-linked glycans are expected to significantly influence PrP^TSE adsorption since they occupy nearly the same volume as the polypeptide and extend toward the exterior of structured aggregates (DeMarco and Daggett, 2009). Few studies have investigated the effect of glycosylation on protein adsorption. A recent study using 3 ermomyces lanuginosus lipase found no difference in adsorption to hydrophobic surfaces among the unglycosylated, monoglycosylated, and pentaglycosylated protein but adsorption to hydrophilic surfaces of only the pentaglycosylated form (Pinholt et al., 2010). The glycans in this study were composed of uncharged sugar residues. In general, the size and charge of oligosaccharides likely determines their relative contribution to protein adsorption. The GPI anchor at the C terminus of PrP^TSE may also influence its adsorption to soil particle surfaces. Because the phosphatidylinositol group is lipophilic, it is expected to interact strongly with hydrophobic surfaces. However, the specific effect of membrane anchors on protein adsorption has not been discussed in the literature.

Mobility of PrP^TSE in Soils

The transport of prions in soil has important implications for the disposal of potentially infected materials, management of TSE-contaminated pastures, delivery of PrP^TSE to surface waters, and the bioaccessibility of TSE agents in natural environments. In saturated column experiments, the potential for partially purified PrP^TSE to migrate through soils and porous landfill materials has been investigated (Jacobson et al., 2009; Jacobson et al., 2010). The migration of PrP^TSE through fine quartz sand, shredded municipal solid waste (fresh and aged), and daily cover materials (natural soils and green waste residual) was examined using landfill leachate as the eluant (Jacobson et al., 2009). All detectable PrP^TSE was retained near the point of introduction in columns packed with soil or fine quartz sand, but 0.3 to 28% of the added PrP^TSE eluted from columns containing green residual waste (i.e., composted materials) and municipal solid waste. A similar study using protease-digested brain homogenate as the prion source assessed the potential for PrP^TSE to migrate through natural soils (Jacobson et al., 2010). All detectable PrP^TSE was retained in columns of five different soils of high sand or silt content when artificial rainwater was used as the eluant (Jacobson et al., 2010). A study using α-recPrP also reported limited transport in column experiments monitored for 6 to 9 mo under conditions of normal and reduced microbial activity and varying water content (simulating a fluctuating water table) (Cooke and Shaw, 2007). Overall, limited prion mobility in soil columns is consistent with the strong attachment of PrP^TSE to soil particles. Dissolved organic carbon may contribute to PrP^TSE mobility in soil and therefore warrants further study. The studies conducted to date suggest that in the absence of preferential flow paths or facilitated transport, prion mobility in most soils and subsurface environments should be quite limited. This suggests maintenance of TSE agents near the soil surface where they would be more accessible to grazing animals. Prions at the soil surface would also be more available for entrainment into overland flow and delivery to water bodies (Nichols et al., 2009).

Infectivity of Soil Particle-Bound PrP^TSE

For soil to represent a plausible reservoir of TSE agents in the environment, soil particle-bound prions must remain bioavailable and infectious to susceptible animals. In laboratory experiments, PrP^TSE bound to soil and mineral particles retained undiminished infectivity toward hamsters inoculated intracerebrally (Johnson et al., 2006). Oral exposure of hamsters to montmorillonite– and soil–prion complexes demonstrated transmission by this route of uptake (Johnson et al., 2007). Surprisingly, PrP^TSE binding to montmorillonite dramatically increased disease penetrance (Fig. 3) with an effective titer higher than that of the unbound agent by a factor of 680. Similarly, two of the three soils tested increased disease transmission, as did quartz microparticles and kaolinite (Johnson et al., 2007; Johnson et al., 2011). Because the digestive physiology of rodents differs significantly from that of animals affected by environmental TSE transmission (deer, elk, goats, sheep), the extent to which soil impacts oral TSE transmission in ruminants warrants research.

Fig. 3.

Oral inoculation of hamster with PrP^TSE (HY strain of hamster-adapted transmissible mink encephalopathy) and montmorillonite (Mte) dramatically increases disease penetrance and shortens incubation period relative to that of PrP^TSE alone. Hamsters dosed with Mte alone remained healthy throughout the course of the experiment. (From Johnson et al. [2007] PLoS Pathog. 3, Fig. 3B, p. 877.)

Several possible mechanisms may account for increased transmission of soil-bound prions. Attachment to soil particles may alter PrP^TSE conformation or aggregation state in a manner that increases infectivity. Surface-induced conformational change is well established for adsorbed proteins, and the extent of this transition depends on surface polarity and the conformational lability of the protein. Several studies have reported the conformational alteration of prtoeins (including α-recPrP) upon adsorption to clay minerals (Baron et al., 1999; Servagent-Noinville et al., 2000; Revault et al., 2005). As discussed above, conformational alteration of α-recPrP on adsorption to montmorillonite has been reported (Revault et al., 2005). Research examining potential changes to the secondary structure of mineral particle-bound PrP^TSE is needed. Slight alteration of PrP^TSE structure on binding to soil particles may have a considerable effect on TSE infectivity.

Efficient transmission of soil-bound prions also depends on the effect of particle binding on in vivo mechanisms of uptake. Soil particles may enhance disease transmission by protecting prions from inactivation in the gastrointentinal tract and/or by prolonging PrP^TSE residence times. Detachment from mineral particles may not be required for PrP^TSE uptake. Peyer’s patches (gut-associated lymphatic tissue in the small intestine) are involved in the early stages of CWD and scrapie transmission (Sigurdson et al., 1999; Andréoletti et al., 2000; Miller and Williams, 2002). Peyer’s patches sample the gut lumen for microparticles similar in size to montmorillonite, taking them up by endocytotic mechanisms (des Rieux et al., 2006).

Biotic and Abiotic Inactivation of Prions

Efforts to limit the spread of horizontally transmitted TSEs are complicated by the known persistence of prions in soil (Fig. 4; Seidel et al., 2007; Brown and Gajdusek, 1991). Currently, effective methods to remediate naturally contaminated areas are lacking, and the extent to which natural degradation processes in soil may inactivate prions remains unclear. Potential PrP^TSE inactivation pathways include microbial degradation, degradation by extracellular enzymes (proteases and oxidases), oxidation by inorganic oxidants, and denaturation due to repeated soil freeze–thaw and wetting–drying events. Insight into the potential degradation of prions in soil may advance the understanding of environmental TSE transmission and result in practical methods for land decontamination. As discussed above, analytical methods capable of detecting prions in naturally contaminated soils are currently lacking. Studies examining prion degradation have therefore used samples experimentally amended with prions. These experiments often use rodent-adapted prion strains that lack environmental relevance. While useful, inactivation of prions needs to be confirmed for the strain and species of interest. For example, the stability in brain homogenate of rodent-adapted prion strains may vary from those of scrapie and CWD agents (Saunders et al., 2008). Many studies use immunodetection methods to measure declines in PrP^TSE immunoreactivity; definitive results on prion inactivation require measurement of TSE infectivity by bioassay. Below we discuss initial research on biotic and abiotic mechanisms of prion degradation in soil with an emphasis on future research needs.

Fig. 4.

PrP^TSE persists in soil for at least 29 mo. Western blot detection of PrP^TSE extracted from experimentally contaminated soils after different time periods. Lane 1: proteinase K–digested hamster brain homogenate (positive control); lanes 2–10: PrP^TSE extracted at time point 0 (lane 2), after 1 mo (lane 3), after 3 mo (lane 4), after 6 mo (lane 5), after 12 mo (lane 6), after 18 mo (lane 7), after 21 mo (lane 8), after 26 mo (lane 9), and after 29 mo (lane 10). (From Seidel et al. [2007] PLoS ONE 2, Fig. 2A, p. 435.)

Degradation of PrP^TSE by Microorganisms

Considering the diversity of bacterial and fungal organisms in the soil environment, significant interest exists in identifying microorganisms capable of degrading PrP^TSE. Several studies have examined PrP^TSE degradation in samples with enhanced microbial activity. Scherbel et al. (2006) found that rumen and colonic microorganisms were competent to degrade PrP^TSE associated with hamster-adapted scrapie (263K strain) under anaerobic conditions in the presence of soluble carbohydrates. Degradation in this study was evidenced by complete loss of PrP^TSE immunoreactivity; however, further analysis by animal bioassay revealed significant residual TSE infectivity (Scherbel et al., 2007). Compost represents a promising matrix for PrP^TSE degradation due to the combination of high temperatures (>55°C), pH fluctuations (5.5–8.5), and intense microbial activity. Huang et al. (2007) found that composting scrapie-infected sheep tissue for 3 to 4 mo reduced or eliminated PrP^TSE immunoreactivity. The effect of composting on any concomitant reduction in TSE infectivity awaits verification by animal bioassay. The microbial consortia responsible for activated sludge and anaerobic digestion in wastewater treatment plants were also examined for possible PrP^TSE–degrading activity. Hinckley et al. (2008) found that PrP^TSE (HY strain of hamster-adapted TME) partitioned to activated sludge solids (floc), indicating that the protein will be subject to biosolids treatment processes. PrP^TSE was not significantly degraded during incubation with activated sludge consortia or simulated mesophilic anaerobic digestion. Similarly, Kirchmayr et al. (2006) found little reduction in PrP^TSE levels following incubation with mesophilic anaerobic digester sludge.

Other microbial consortia that have been studied include bacteria used in food production and earthworm gut microflora (Müller-Hellwig et al., 2006; Nechitaylo et al., 2010). Six strains of cheese-ripening bacteria were found to reduce PrP^TSE (263K) levels following 24-h digestion at 30°C. Nechitaylo et al. (2010) found little degradation of α-recPrP by aqueous extracts of casts from two earthworm species and concluded that earthworm digestive enzymes and gut microflora are not expected to significantly degrade PrP^TSE.

To our knowledge, only one study to date has investigated soil microorganisms as a means of prion inactivation. Rapp et al. (2006) demonstrated that buried animal carcasses stimulate the proteolytic activity of the surrounding soil and that proteases extracted from the stimulated soil degraded β-recPrP within 8 d. The use of β-recPrP in this study limits the conclusions that can be drawn about PrP^TSE degradation under similar conditions.

Prion Degradation by Isolated Enzymes

Proteases (protein-degrading enzymes) are grouped into five distinct classes on the basis of the nature of their catalytic site: serine, cysteine, aspartate, metallo, and threonine proteases. Metallo and aspartate proteases use an activated water molecule to attack the peptide backbone of the protein, whereas serine, cysteine, and threonine proteases are named for the nucleophilic residue in the active site of the enzyme (Puente et al., 2003). A number of proteases have been examined for their ability to degrade PrP^TSE. For optimal activity, most require conditions that partially denature the protein such as high temperature, high pH, and/or the presence of detergents.

Subtilisin-like enzymes constitute one of the major subfamilies of serine proteases and have received the most study with respect to potential PrP^TSE degradation. In most cases, these subtilisin-like proteases are applied under harsh conditions (temperature ≥ 50°C, pH ≥ 9). Exposure to the subtilisin enzyme Properase effected a 10³-fold reduction in mouse-adapted BSE infectivity at 60°C and pH 12 (McLeod et al., 2004). Amino acid substitutions increased the thermostability of the enzyme, which is now marketed as Prionzyme (Dickinson et al., 2009). Exposure (30-min) to Prionzyme has achieved a ≥10⁷-fold reduction in mouse-adapted BSE infectivity after treatment at 60°C and pH 12. Treatment of PrP^TSE with subtilisin enzyme 309-v resulted in a 52% increase in survivorship of mice inoculated with mouse-adapted scrapie agent (Pilon et al., 2009).

Keratinases often exhibit sequence homology with subtilisin enzymes (Macedo et al., 2005; Evans et al., 2000) and have attracted attention because like PrP^TSE their natural substrate, keratin, possesses high β-sheet content. The feather-degrading keratinase produced by Bacillus licheniformis strain PWD-1 effected the extensive breakdown of PrP^TSE (BSE and sheep scrapie), but only in the presence of 2% sarkosyl and at temperatures exceeding 100°C (Langeveld et al., 2003).

Alkaline serine proteases (ASP) are serine proteases that function optimally at an alkaline pH. At high temperature and pH, PrP^TSE (263K) was digested by treatment with an ASP (Hui et al., 2004b) and by an ASP-like enzyme isolated from Streptomyces strain 99-GP-2D-5 (Hui et al., 2004a). Incubation of PrP^TSE (263K) with purified NAPase, a keratinolytic alkaline protease from Nodocardiopsis sp. TOA-1, exhibited optimal activity at 60°C and pH 11 (Mitsuiki et al., 2006); however, controls were lacking to demonstrate that the alkali solution (pH 11) used for enzyme purification did not degrade PrP^TSE.

Disaggregation and partial denaturation of PrP^TSE may be promoted at high pH, elevated temperatures, and in the presence of detergents as inter- and intramolecular hydrogen bonds are broken, access to the peptide backbone is facilitated, and the protein is solubilized. While these conditions enhance proteolytic degradation of PrP^TSE, they do not represent conditions prevailing in natural soils and are impractical for treating in-place soils. To date, three studies have reported PrP^TSE proteolysis under milder, more environmentally relevant conditions. Prionzyme was shown to degrade PrP^TSE (elk CWD and HY strain of TME) at neutral pH and moderate temperatures (Saunders et al., 2010). Proteases produced by the mesophilic soil bacterium Streptomyces galbus var. achromogenes 695-206 were capable of degrading β-recPrP at 30°C but were less effective at degrading bona fide PrP^TSE (mouse-adapted BSE strain 6PB1) (Tsiroulnikov et al., 2004). Proteolytically active extracts from whole lichen tissue reduced PrP^TSE (Hy, 263K, and CWD) levels at least 100-fold at 37°C (Johnson, personal communication, 2010). The efficacy of the above treatments in reducing TSE infectivity remains to be verified.

Additional Factors Affecting Prion Degradation in Soil

The attachment of PrP^TSE to soil particles may confer partial protection from denaturation and proteolysis (Johnson et al., 2006). Mineral-bound Cry proteins are protected from proteolysis by soil microorganisms (Koskella and Stotzky, 1997; Chevallier et al., 2003). Glomalins, glycoproteins associated with arbuscular mycorrhizal fungi, represent a significant fraction of the carbon immobilized in soil due to their resistance to microbial hydrolysis (Wright and Upadhyaya, 1996, 1998). This persistence is probably due to their association with soil minerals including iron and aluminum oxyhydroxides (Quiquampoix and Burns, 2007). The extraordinary persistence of prions in the environment may likewise be due, at least in part, to their strong attachment to soil particles.

Physical processes that prions may be exposed to in the environment such as freeze–thaw cycles or wetting and drying events may also contribute to the degradation of prions. These processes disrupt soil structure (Hillel, 1998) and may partially denature PrP^TSE to the extent that enzymes gain access to the peptide backbone or infectivity is lost. To date, effects of freeze–thaw and wetting–drying cycles on PrP^TSE and prion infectivity have not been formally investigated.

Abiotic Degradation of Prions

Transition metal oxides found in soils (viz. manganese, iron and aluminum oxides) can mediate the transformation of organic compounds including aromatic amines (Laha and Luthy, 1990), phenolic compounds (McBride, 1989), estrogens (Xu et al., 2008), herbicides (Barrett and McBride, 2005; Cheney et al., 1998), and antibiotics (Zhang et al., 2008; Rubert and Pedersen, 2006; Zhang and Huang, 2007; Chen and Huang, 2010). Aqueous suspensions of a synthetic manganese oxide similar to vernadite (δ-MnO₂) were found to effect PrP^TSE degradation (HY strain of hamster-adapted TME) at pH ≤ 5 (Russo et al., 2009). At pH 4 and the highest MnO₂ concentration used, PrP^C-to-PrP^TSE converting ability (as measured by PMCA) was reduced by ≥4 orders of magnitude.

Lignolytic fungi produce reactive oxygen species in their initial assault on woody plant debris (Cohen et al., 2002), and a methoxyhydroquinone-driven Fenton reaction has been used to simulate the extracellular chemistry of such fungi in vitro (Metz et al., 2009). The Fenton reaction can induce radical-mediated damage to both the peptide backbone and side chains of proteins through hydroxylation or direct oxidation of amino acid residues (Stadtman and Berlett, 1997) and may represent a pathway of PrP^TSE inactivation in the environment. Fenton reagent induced degradation of PrP^TSE (263K, Sc237) in vitro, significantly delayed onset of clinical symptoms (Park et al., 2008) and appeared to effect a 10⁶-fold reduction in PrP^C-to-PrP^TSE converting ability (based on PMCA) (Suyama et al., 2007). In the latter study, untreated positive controls also failed to efficiently amplify, limiting confidence in the PMCA results. Overall, these studies were conducted at hydroxyl radical concentrations considerably higher than those expected in soil environments.

Conclusions

The pathogenic prion protein readily attaches to soil particle surfaces, and the interaction between PrP^TSE and some clay mineral surfaces is particularly avid. Prions exhibit limited mobility in soil columns, suggesting that in the absence of preferential flow paths or facilitated transport, TSE agents released into soil would remain near the soil surface where they are more accessible to grazing animals. Particle-associated PrP^TSE molecules may migrate from locations of deposition via transport processes affecting soil particles, including entrainment in and movement with air and overland flow. The mechanisms of PrP^TSE attachment to soil particle surfaces remain to be elucidated. The extent of PrP^TSE interaction with natural organic matter and the influence of N-linked glycosylation and glypiation on PrP^TSE attachment to soil particle surfaces warrant research. Soil and mineral particle-bound prions exhibit increased disease transmission by the oral route relative to that of the unbound agent. Enhanced disease transmission by particle-bound PrP^TSE may result from altered bioavailability of soil-bound prions by animals, changes in PrP^TSE conformation or aggregation state that impact TSE infectivity, or both. Research is needed to elucidate these mechanisms and evaluate the extent to which soil impacts oral TSE transmission in ruminants.

Several reports have shown that prions can persist in soil for several years. Significant interest remains in developing methods that could be applied to degrade PrP^TSE in naturally contaminated soils. Preliminary research suggests that serine proteases and the microbial consortia in stimulated soils and compost may partially degrade PrP^TSE. Transition metal oxides in soil (viz. manganese oxide) may also mediate prion inactivation. Overall, the effect of prion attachment to soil particles on its persistence in the environment is not well understood, and additional study is needed to determine its implications on the environmental transmission of scrapie and CWD.

Abbreviations

α-recPrP

proteolytically labile recombinant prion protein

β-recPrP

recombinant prion protein folded into β-sheet rich oligomers

ASP

alkaline serine protease

BSE

bovine spongiform encephalopathy

CWD

chronic wasting disease

ELISA

enzyme-linked immunosorbent assay

GPI

glycosylinositol phospholipid

IEP

isoelectric point

IU

infectious unit

LC–MS

liquid chromatography–mass spectrometry

NAP

keratinolytic alkaline protease from Nodocardiopsis sp. TOA-1

NMR

nuclear magnetic resonance spectroscopy

NOM

natural organic matter

PAGE

polyacrylamide gel electrophoresis

PBS

phosphate buffered saline

pD

negative logarithmic value of the deuterium ion concentration

PK

proteinase K

PMCA

protein misfolding cyclic amplification

PrP

prion protein

PrP^C

normal, benign cellular form of the prion protein

PrP^TSE

pathologically misfolded form of prion protein

recPrP

recombinant prion protein

SDS

sodium dodecyl sulfate

TME

transmissible mink encephalopathy

TSE

transmissible spongiform encephalopathy

UV

ultraviolet