Inactivation of Template-Directed Misfolding of Infectious Prion Protein by Ozone

Abstract

Misfolded prions (PrP^Sc) are well known for their resistance to conventional decontamination processes. The potential risk of contamination of the water environment, as a result of disposal of specified risk materials (SRM), has raised public concerns. Ozone is commonly utilized in the water industry for inactivation of microbial contaminants and was tested in this study for its ability to inactivate prions (263K hamster scrapie = PrP^Sc). Treatment variables included initial ozone dose (7.6 to 25.7 mg/liter), contact time (5 s and 5 min), temperature (4°C and 20°C), and pH (pH 4.4, 6.0, and 8.0). Exposure of dilute suspensions of the infected 263K hamster brain homogenates (IBH) (0.01%) to ozone resulted in the in vitro destruction of the templating properties of PrP^Sc, as measured by the protein misfolding cyclic amplification (PMCA) assay. The highest levels of prion inactivation (≥4 log₁₀) were observed with ozone doses of 13.0 mg/liter, at pH 4.4 and 20°C, resulting in a CT (the product of residual ozone concentration and contact time) value as low as 0.59 mg · liter⁻¹ min. A comparison of ozone CT requirements among various pathogens suggests that prions are more susceptible to ozone degradation than some model bacteria and protozoa and that ozone treatment may be an effective solution for inactivating prions in water and wastewater.

INTRODUCTION

The normal cellular prion protein (PrP^C) is present in most tissues of humans and other animals (15) and is expressed at particularly high levels in neuronal tissues (20, 40). PrP^C can transform into a protease-resistant, infectious, misfolded pathological isoform (PrP^Sc) which can subsequently act as an efficient template for conformational misfolding of PrP^C. Template-directed misfolding of PrP^C by PrP^Sc represents the fundamental mechanism for disease progression in prion-related transmissible spongiform encephalopathies (TSEs) such as scrapie, chronic wasting disease (CWD), and bovine spongiform encephalopathy (BSE) (28). Moreover, the associated change in conformational structure results in the protein (PrP^Sc) becoming resistant to most treatment methods used for the inactivation of microbial pathogens (37).

Tissues associated with the nervous and immune systems are considered specified risk materials (SRMs) in the food animal production industry. Due to the resistant nature of PrP^Sc and the potentially high levels of PrP^Sc present in SRM tissues, the liquid and solid wastes generated from the disposal of SRM pose serious concerns due to the possible release and bioaccumulation of misfolded infectious prion proteins in the environment. PrP^Sc has been found to be extremely recalcitrant in the environment. Results reported by Brown and Gajdusek (5) demonstrated that PrP^Sc infectivity remained in soil for 3 years, and both soil and the aqueous extracts from contaminated soil were infectious in animal models (33). Recently, the prion agent of CWD in cervids was detected in one surface water sample from an area in the United States where CWD is endemic (23). The authors of that study suggested that the persistence and accumulation of prions in the environment may promote the transmission of CWD (23). Water and wastewater may act as transport agents for prions from liquid wastes from slaughtering houses, rendering plants, agricultural digesters, and some septic systems (27). PrP^Sc appears to be resistant to conventional municipal water and wastewater treatment regimens, such as chlorination (usually ∼1 mg/liter available Cl₂ in water) (38), UV irradiation (11), and mesophilic anaerobic sludge digestion (12). The high resistivity of PrP^Sc to conventional inactivation in water and wastewater intensifies concerns about prion contamination of the environment; thus, effective approaches for prion decontamination in aqueous environment are desirable.

Ozone as an advanced oxidation technology is widely used in water and wastewater treatment processes for inactivation of bacteria (44), viruses (19, 39), and protozoa (18). Although certain advanced oxidation processes have been shown to inactivate infectious prions (14, 25, 26, 34, 35), several major knowledge gaps exist; in particular, knowledge concerning how interactions between oxidant reaction conditions (e.g., temperature, pH, duration of exposure, ozone dose, and organic load) affect inactivation levels of prions is lacking. Consequently, the derivation of a concentration-time (CT) disinfection value for prions has never been reported. The CT value is defined as the product of the residual disinfectant concentration (e.g., in milligrams per liter) and the contact time (e.g., in minutes) required to achieve a certain level of inactivation of a particular target organism (i.e., the CT₉₉ is the CT product required for 99% [or 2-log₁₀] inactivation). Maintaining and measuring the residual oxidant dose is critical during the disinfection process, since advanced oxidant products are short-lived and consumed rapidly by nontarget biomolecules (i.e., first-order decay), and a sustained residual dose is required for continuous reaction against the more resistant biomolecular structures (e.g., aggregates of PrP^Sc). The U.S. Environmental Protection Agency (USEPA) routinely uses a CT concept for characterizing disinfection requirements for microbial pathogens under a given set of reaction conditions (pH and temperature) and for which a comparative assessment of the susceptibility of the pathogens can be made (41, 42). Consequently, the CT value is commonly used as an engineering target for inactivation of pathogens in water matrices, as a regulatory standard in water treatment, and for modeling the inactivation kinetics of physicochemical disinfectants (19, 39).

Until recently, a major challenge associated with studying physicochemical inactivation of prions had been the lack of a cost-effective, high-volume assay for detection and quantification of PrP^Sc, and animal bioassays therefore remain the gold standard for assessing prion infectivity. However, the protein misfolding cyclic amplification (PMCA) assay has recently emerged as a powerful in vitro tool for detection and quantification for PrP^Sc (32). Protein misfolding cyclic amplification results in template-directed misfolding of PrP^C (naturally present in normal brain homogenates [NBH]) as a product of seeding a small quantity of infectious PrP^Sc into the in vitro reaction, thereby mimicking the pathological process of disease progression in vivo (6). The overall amount of amplification obtained in the PMCA is contingent on the amount of infectious seed used to initiate template-directed misfolding (32), thus allowing quantification of PrP^Sc found in the original infectious seed. The PMCA assay has a wide dynamic range of sensitivity (31). This assay considerably reduces the time required for generating results (i.e., 3 days) compared to animal bioassay models (several months to more than a year), and studies examining heat sterilization of PrP^Sc demonstrated that results obtained by PMCA correlated with animal infectivity (22, 36).

The objective of this study was to determine the effectiveness of ozone for inactivating template-directed misfolding properties of PrP^Sc (263K scrapie), as determined by PMCA, under a variety of experimental conditions (ozone dose, contact time, pH, and temperature) for which CT values for ozone could be derived.

MATERIALS AND METHODS

Buffers.

Phosphate-buffered saline (PBS; 130 mM sodium chloride, 20 mM potassium chloride, 7 mM sodium phosphate, 3 mM potassium phosphate, Milli-Q water) was prepared and used to make brain homogenates and ozone stock solution in inactivation experiments. An alternative PBS buffer (0.66×) was used as a component of a conversion buffer for the PMCA assay. The 0.66× PBS was prepared using PBS tablets (BioBasic Inc., Markham, Ontario, Canada) by dissolving 1 tablet per 150 ml of Milli-Q water.

PMCA conversion buffer was prepared using a final concentration of 0.15 M sodium chloride (Fluka/Sigma-Aldrich, Toronto, Canada), 5 mM EDTA (Gibco Invitrogen Canada Inc., Burlington, Ontario, Canada), and 1% Triton (MP Biochemicals, Salon, OH) in 0.66× PBS and adding 1× complete protease inhibitor cocktail (Roche Diagnostics, Laval, Quebec, Canada) according to manufacturers' instructions.

Animals.

Three- to 6-week-old female Syrian golden hamsters (Charles River Laboratories International, Inc., Wilmington, MA) were used to prepare infectious brain homogenates (IBH) and normal brain homogenates (NBH). The hamster handling protocol used in this study adhered to the Canadian Council of Animal Care (CCAC, Canada) guidelines.

Preparation of hamster brain homogenates. (i) Infectious prion brain homogenates (IBH).

Female Syrian Gold hamsters were exposed either orally (100 μl) or by intraperitoneal injection (50 μl) to an inoculum of 263K scrapie-positive brain homogenates at the Canadian Food Inspection Agency (CFIA) Transmissible Spongiform Encephalopathy (TSE) Laboratory in Nepean, Ontario, Canada. Hamsters displaying clinical signs of scrapie, typically 95 to 110 days postinoculation, were euthanized with carbon dioxide and the brains harvested in as short a time as possible. All infected hamsters were confirmed positive by routine diagnosis (enzyme-linked immunosorbent assay and immunohistochemistry) at the CFIA TSE Laboratory. The infectious dose of brain homogenates was subsequently determined to be 10^9.94 ID₅₀ per gram of brain tissue as confirmed by hamster infectivity endpoint titration assays. IBH samples (10% [wt/vol] in 1× PBS) were manually disrupted (15 to 20 strokes) on ice using a Potter glass tissue grinder/homogenizer and allowed to stand on ice for 30 min followed by 1 min of centrifugation at 1,000 × g. The clarified 10% IBH supernatant was used as a stock solution for ozone inactivation and PMCA experiments.

(ii) Normal brain homogenates (NBH).

Hamsters were sacrificed by exposure to excess carbon dioxide (dry ice in a kill box). Upon confirmed death, cold 1× PBS with 5 mM EDTA (Gibco Invitrogen Canada Inc., Burlington, Ontario, Canada) was perfused through the hamster circulatory system with the assistance of a peristaltic pump attached to a syringe with the needle puncture to the left ventricle of the heart. A 10% NBH solution was prepared by adding 1 g of perfused brain suspended in 8 ml of the conversion buffer and 1 ml of 15 USP sodium heparin solution (BD Vacutainer, Franklin Lakes, NJ). The sodium heparin solution was originally prepared by the addition of 10 ml of 0.66× PBS to a precoated 150 USP sodium heparin Vacutainer (BD Vacutainer) and subsequently separated into aliquots and frozen at −20°C. Normal brain samples in the conversion buffer/sodium heparin solution were manually disrupted (15 to 20 strokes) on ice using a Potter glass tissue grinder/homogenizer and allowed to stand on ice for 30 min followed by 1 min of centrifugation at 1,000 × g. The clarified 10% NBH supernatant was used for PMCA assays and experimentation.

Ozone inactivation of PrP^Sc. (i) Preparation of ozone-demand-free (ODF) reagents and reactors.

The 1× PBS was used to make ozone stock solution. The pH was adjusted to pH 4.4, 6.0, and 8.0 with 1 M HCl or 1 M NaOH. The PBS buffer, reaction tubes, pipette tips, and magnetic stir bars were made ozone demand free prior to use. The ODF PBS buffer was prepared by bubbling ozone gas (see ozone gas preparation description below) through a glass bottle for 20 min. The bottle was then placed at room temperature with the lid loose until ozone residuals were completely dissipated, as confirmed by the Indigo method (1). The reaction tubes, pipette tips, and stir bars were soaked in an ozone solution in deionized water (initial concentration > 10 mg/liter) overnight, followed by drying at room temperature for 3 days.

(ii) Ozone inactivation experiment.

Ozone stock solutions (in 1× PBS at pH 4.4, 6.0, and 8.0) were generated using ultrapure oxygen and an ozone generator (G30; PCI Wedeco). Concentrated ozone stock solutions in PBS buffer were prepared by bubbling ozone gas through 1 liter of PBS buffer chilled at 4°C for around 30 min, and the ozone dose within the stock solution was determined using the Indigo method (1).

For ozone inactivation experiments, shell vial reaction tubes (Fisher Scientific, Canada) were mounted on top of a stir plate with a Teflon-coated magnetic stir bar in each tube to ensure even mixing. The temperature was controlled by submerging the reaction tubes in ice water (4°C) or at room temperature (20°C). Separate reaction tubes were set up to withdraw samples at predetermined reaction times. The experiment was carried out by adding ODF PBS buffer (same pH as the stock solution) and diluted IBH into reaction tubes followed by ozone stock solution, with a final volume of 1 ml. During the reaction, reaction tubes were covered with plastic lids. Samples were withdrawn at predetermined reaction times for residual ozone concentration determinations, immediately followed by addition of 20 μl of 1 M sodium thiosulfate to neutralize residual ozone in the reaction tubes. The ozonated samples were then frozen at −80°C until the PMCA assay was performed. The nonozonated control samples were treated in the same manner except that ozone was completely neutralized prior to addition of diluted IBH. The residual ozone was determined by the Indigo method (1) with a UV-visible spectrophotometer (Biospec Mini 1240; Shimadzu, Japan) immediately after the experiment. Absorbance measurements were performed at 600 nm in a 1-cm path length quartz cell. The presence of diluted IBH and the applied sodium thiosulfate in the control samples had a negligible effect on Indigo method (data not shown). The absorbance of the control samples was set up as the reference for calculation of the residual ozone concentration at predetermined reaction time points.

Protein misfolding cyclic amplification (PMCA) assay.

The PMCA assay was used to measure inactivation of the templating properties of ozone-treated PrP^Sc samples. Control and ozonated samples were serial diluted 10-fold in 10% NBH prior to the PMCA assay. Subsequently, two replicates of 8-μl aliquot samples of each 10-fold dilution series were mixed with 72 μl of 10% NBH in 200-μl flat-cap, thin-wall PCR tubes (Axygen, Union City, CA) by inversion. As a negative control for PMCA, 80 μl of a 10% NBH sample was also prepared in the thin-wall PCR tubes. PCR tubes were randomly placed in a Misonix model 4000MXP sonicator (Misonix Inc., Farmingdale, NY) with the sonicator microplate cup horn housed within the acoustic enclosure (provided with the instrument) and the water reservoir temperature set to 37°C. PMCA was performed for 19 h, with 40 s of sonication followed by 29-min and 20-s incubation periods within each cycle (total number of cycles = 38) at a potency of 90%. Samples were subsequently frozen at −80°C, and PrP^Sc was detected by Western blot analysis.

PK digestion, SDS-PAGE, and Western blot analysis.

Proteinase K (PK) digestion (200 μg of PK/ml) was carried out on all PMCA samples and non-PMCA controls (1% IBH–NBH). The digestion was performed at 37°C for 20 min and stopped by the addition of an equivalent volume of 2× Laemmli buffer (28.5 ml of Laemmli sample buffer [Bio-Rad Laboratories, Mississauga, Ontario, Canada], 0.6 g of sodium dodecyl sulfate [SDS; Sigma-Aldrich Canada Ltd., Oakville, Ontario, Canada], 1.5 ml of β-mercaptoethanol [Sigma-Aldrich Canada Ltd., Oakville, Ontario, Canada]) and incubated at 100 ± 5°C for 5 min. Denatured samples (25 μl) were fractionated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) (Pierce precise 12% precast polyacrylamide gels; Thermo Scientific, Rockford, IL) at 100 V for 1 h and transferred onto a polyvinylidene difluoride (PVDF) membrane (Bio-Rad Laboratories, Mississauga, Ontario, Canada) at 20 V overnight on ice. The blots were blocked in 5% skim milk–1× PBS (Bio-Rad Laboratories, Mississauga, Ontario, Canada) with 0.1% Tween 20 (Bio-Rad Laboratories, Mississauga, Ontario, Canada) for 1 h at room temperature. The blots were then probed with primary anti-prion protein 3F4 antibody (Millipore, Billerica, MA) at 1:20,000 in 1× PBS (10 mM sodium phosphate and 150 mM sodium chloride) containing 0.1% Tween 20 for 1 h at room temperature, followed by washing three times (10 min each time) in 1× PBS–0.1% Tween 20. The conjugated secondary antibody, goat anti-mouse horseradish peroxidase (HRP; Bio-Rad) (1:10,000 in 1× PBS–0.1% Tween 20), was subsequently added to bind to the primary antibody and incubated for 1 h at room temperature, followed by washing three times (10 min each time) in 1× PBS–0.1% Tween 20 and washing twice (5 min each time) in 1× PBS without Tween 20. Immunoreactive bands were then visualized using ECL reagent and an ImageQuant RT ECL Imager (Amersham, GE Life Sciences, Canada).

Quantitative analysis of Western blot images.

The blot images of control and ozonated samples were analyzed using ImageQuant TL software (GE Healthcare). The net intensity of each blot was obtained from the software. The signal intensity of all dilutions of ozonated samples and 2 to 6 log₁₀ dilutions of the nonozonated control samples were normalized as a percentage of the average signal intensity (saturated) of two replicates of a 1 log₁₀ dilution of the nonozonated control samples. The background intensity was subtracted before normalization. By assuming that the normalized intensity (as a percentage) versus fold dilution followed an exponential relationship before signal intensity saturation (32), the inactivation of PrP^Sc by ozonation was calculated according to equation 1:

$$\text{Inactivation}\left({\text{log}}_{\text{10}}\frac{N_0}{N}\right)={\text{log}}_{\text{10}}\left(\frac{{\log }_{10}\left(\text{normalized intensity 1}\right)\times \left(\text{dilution fold 1}\right)}{{\log }_{10}\left(\text{normalized intensity 2}\right)\times \left(\text{dilution fold 2}\right)}\right)$$ (1)

where N₀ is the concentration of PrP^Sc template in PMCA nonozonated control samples, N is the concentration of PrP^Sc template in PMCA ozonated samples, normalized intensity 1 is the intensity of the highest dilution of nonozonated sample divided by the intensity of the 1 log₁₀ dilution of nonozonated sample, dilution fold 1 is the dilution (in the form of 10ⁿ, where n is an integer) of the highest-diluted lane of the nonozonated control sample, i.e., 10⁴ and 10⁵, normalized intensity 2 is the intensity of the highest dilution of the ozonated sample divided by the intensity of the 1 log₁₀ dilution of the nonozonated sample, and dilution fold 2 is the dilution (in the form of 10ⁿ, where n is an integer) of the highest-diluted lane of the ozonated control sample, i.e., 10¹, 10², 10³, 10⁴.

Estimation of CT values.

The CT values were estimated for the purpose of assessing the degree of ozone inactivation of the template-directed misfiling of prions under specific conditions. Since ozone decomposition follows first-order kinetics after the initial ozone demand, the CT values at a contact time of 5 min were estimated by determining the area under the ozone decay curve at the specific time, using equation 2:

$$CT=\int C(t)dt=\frac{C_0}{k\prime }[1-\exp (-k\prime t)]$$ (2)

where C₀ is the ozone concentration at time 5 s (closest to time 0), k′ is the first-order ozone decomposition rate (per minute⁻¹), and t is contact time (in minutes). To be conservative, the CT for a contact time of 5 s was calculated by multiplying instantaneous ozone concentration by contact time.

RESULTS

Evaluation of ozone demand.

Prior to investigating the effect of ozone inactivation on PrP^Sc, preliminary experiments were carried out to determine the ozone demand associated with inactivation of PrP^Sc in experimental brain homogenates. At an ozone dose of ∼12 mg/liter, ozone was completely consumed by a 0.1% IBH suspension within 5 s, while a residual ozone concentration was maintained for up to 5 min when a 0.01% IBH sample was used (Table 1). For this reason, a 0.01% IBH sample was used for all subsequent ozonation experiments for determining the inactivation efficiency of PrP^Sc by ozone and for calculation of an ozone CT product for PrP^Sc.

Table 1.

Ozone demand of PrP^Sc at pH 6.0 and 4°C

Ozone dose (mg/liter)	Concn of IBH (%)	Contact time	Residual ozone (mg/liter)
12.9	0.1	5 s	0
		30 s	0
		2 min	0
		5 min	0
12.5	0.01	5 s	8.3
		30 s	8.2
		2 min	6.0
		5 min	5.0

Detection of PrP^Sc by PMCA assay.

A 0.1% suspension of IBH represented the lowest concentration of IBH for which PrP^Res could be detected by Western blot analysis alone (Fig. 1A), indicating that Western blot analysis was insufficient as a detection tool for characterizing ozone inactivation of PrP^Sc. Incorporation of PMCA upstream of the Western blot analysis increased the sensitivity of detection of PrP^Sc to approximately 6 to 7 log₁₀ compared to Western blot analysis alone (Fig. 1). A 10% IBH sample could be diluted 100-million-fold (10⁸) and still be detected after only a single round of PMCA (38 cycles) (Fig. 1B). This expanded range of sensitivity allowed detection and quantification of ozone inactivation of PrP^Sc under the conditions necessary to maintain an ozone residual (i.e., 0.01% IBH suspension) and consequently for derivation of a range of CT values based on approximately 4 orders of magnitude of inactivation (i.e., up to CT_99.99 [PMCA detection of a prion signal between 0.01% IBH and 0.000001% IBH]).

Fig 1.

A comparison between traditional Western blot methodologies and PMCA for detection of 263K scrapie. (A) Western blot analysis of a 10-fold serially diluted IBH sample (10%) without PMCA. Lanes labeled 10 to 0.001 represent percentages of IBH. (B) Western blot analysis of a 10-fold serially diluted IBH (10%) with PMCA. Lanes labeled 1 to 0.00000001 represent percentages of IBH. Lanes labeled “PK−” represent 1% IBH not treated with PK. Lane labeled “−” represent 10% NBH treated with PK. Lanes labeled “+” represent 1% IBH treated with PK. Molecular weight markers (lanes labeled MW) at 50, 37, 25, and 20 are indicated.

Ozone inactivation of PrP^Sc.

Ozone inactivation experiments were performed under various conditions of initial ozone doses (7.6 mg/liter to 25.7 mg/liter), contact times (5 s and 5 min), pH (4.4, 6.0, and 8.0), and temperatures (4 and 20°C) to assess optimal conditions for inactivating the templating properties of PrP^Sc. In preliminary low-dose (7.6 mg/liter) ozone reactions carried out at pH 8.0 and 20°C using PMCA, a PrP^Sc-reactive band was observed in control blots (i.e., samples in which ozone was neutralized by sodium thiosulfate) at as low as a 5 log₁₀ dilution of a 0.01% IBH sample (Fig. 2C). However, after only 5 s of exposure to those conditions of ozone dose, pH, and temperature, a loss of 1 log₁₀ in PrP^Sc Western blot signal intensity was observed (Fig. 2C), and exposure to ozone at this dose for 5 min resulted in a 3 log₁₀ loss in observable PrP^Sc signal intensity (Fig. 2C). In fact, exposure of an IBH (0.01%) sample to ozone under any of the conditions tested always resulted in a measureable loss of signal intensity by PMCA compared to controls (Fig. 2). At pH 4.4 and 20°C, both low-dose ozone (13.0 mg/liter) and high-dose ozone (23.5 mg/liter) readily inactivated PrP^Sc, as assessed by the inability of ozone-treated IBH to act as a seeding template for conformational misfolding of PrP^C by PMCA (Fig. 2A). Exposure to ozone for as little as 5 s at low pH (4.4) and 20°C appeared to completely inactivate the templating properties of the PrP^Sc present in a 0.01% IBH sample (Fig. 2A). In this context, PMCA represented a valuable tool for examining ozone inactivation of the templating properties of PrP^Sc.

Fig 2.

Western blot analysis of 0.01% IBH samples treated with ozone at 20°C and amplified by PMCA. The images in each panel, from left to right, represent non-ozone-treated control samples, samples treated with ozone for 5 s, and samples treated with ozone for 5 min. The applied ozone doses are provided to the left of each panel. (A) pH 4.4; (B) pH 6.0; (C) pH 8.0. The numbers at the top of each image represent log₁₀ dilutions of 0.01% IBH. Lanes labeled “−” represent 10% NBH treated with PK. Lanes labeled “+” represent 1% IBH treated with PK. Molecular weight markers (lanes labeled MW) at 50, 37, 25, and 20 are indicated.

To generate a more accurate quantitative estimate of ozone inactivation, Western blot images obtained from PMCA reactions were analyzed by densitometry (Fig. 3). Densitometric analyses of Western blots were normalized against the saturated signal intensity of PMCA nonozonated control samples, and the normalized intensity was used to estimate log₁₀ reductions in signal intensity of ozone-treated samples by the use of equation 1 (Fig. 3).

Fig 3.

Densitometry analysis of Western blot images in Fig. 2A and B. (A) Normalized intensity of images in Fig. 2A, lower panel (pH 4.4, ozone dose of 23.5 mg/liter). (B) Normalized intensity of images in Fig. 2B, lower panel (pH 6.0, ozone dose of 20.7 mg/liter). The bars show the ranges of two replicates.

A summary of the log estimates of the levels of ozone inactivation of PrP^Sc based on PMCA densitometric analysis of Western blots under the various experimental conditions is provided in Table 2. Ozone dose, contact time, pH, and temperature were all shown to affect ozone inactivation of PrP^Sc. Higher ozone doses and longer contact times resulted in greater PrP^Sc inactivation at any given pH and temperature. For example, at pH 6.0 and 20°C, low-dose (11.9 mg/liter) ozone exposure resulted in 1.9 log₁₀ inactivation of the templating properties of PrP^Sc after 5 s, while a higher inactivation of 3.6 log₁₀ was achieved at 5 min of exposure (Table 2). At this same pH and temperature, but with a higher ozone dose (20.7 mg/liter), greater inactivation was achieved (2.2 log₁₀ and ≥4 log₁₀ after exposure of PrP^Sc to ozone for 5 s and 5 min, respectively [Table 2]).

Table 2.

Summary of ozone inactivation of PrP^Sc under various conditions

pH	Temperature (°C)	Ozone dose (mg/liter)	Contact time	CT (mg liter−1 min)	Inactivation (log10) (N0/N)
4.4	4	13.7	5 s	0.59	2.8
			5 min	31.6	≥4
		25.7	5 s	1.52	≥4
			5 min	67.8	≥4
	20	13.0	5 s	0.59	≥4
			5 min	28.6	≥4
		23.5	5 s	1.17	≥4
			5 min	56.9	≥4
6.0	4	12.5	5 s	0.69	1.9
			5 min	32.2	3.6
		20.7	5 s	1.33	2.2
			5 min	56.5	≥4
	20	11.9	5 s	0.66	2.4
			5 min	26.0	4.4
		20.7	5 s	1.15	2.9
			5 min	41.5	≥4
8.0	4	9.4	5 s	0.40	0.2
			5 min	14.7	2.4
		14.1	5 s	0.72	1.1
			5 min	25.5	2.9
	20	7.6	5 s	0.02	0.9
			5 min	NA	2.9a
		11.3	5 s	0.36	1.9
			5 min	6.51	3.0

^aNo residual ozone maintained.

In addition to ozone dose and contact time dependency, ozone inactivation of PrP^Sc was pH and temperature dependent. A reaction pH of 4.4 provided greater levels of inactivation of PrP^Sc than a higher pH (6.0 or 8.0) at the same temperature and ozone dose. For example, when the pH of the reaction was increased from 4.4 to 6.0 at 20°C, low-dose ozone (11.9 mg/liter) did not completely inactivate the templating properties of the infectious IBH seed after 5 s or 5 min of contact time (Fig. 2B). Increasing the ozone dose to 20.7 mg/liter at this same pH and temperature inactivated the templating properties of the IBH seed after a 5 min exposure but not after a 5 s exposure (Fig. 2B). This is in contrast to experiments carried out at pH 4.4, where a dose of 13.0 mg/liter of ozone completely inactivated the templating properties of the PrP^Sc in the IBH after only 5 s of exposure (Fig. 2A). When the pH of the reaction was further increased to pH 8.0, ozone doses of 11.3 mg/liter did not completely inactivate the templating properties of PrP^Sc in the IBH after 5 min (Fig. 2C). Greater inactivation of PrP^Sc was observed at higher temperatures. For example, at pH 4.4, an ozone dose of 13.0 mg/liter, and contact time of 5 s, PrP^Sc inactivation increased from 2.8 to ≥4 log₁₀ as the temperature was increased from 4 to 20°C (Table 2). Overall, the ideal reaction conditions for PrP^Sc inactivation were at pH 4.4 and 20°C (Table 2).

Ozone inactivation CT values for PrP^Sc under various treatment conditions were generated using equation 2 and are presented in Table 2. With a range of initial ozone doses of between 11.3 and 14.1 mg/liter, at 4°C, a CT of between 0.59 and 0.72 mg · liter⁻¹ min resulted in 2.8 log₁₀ inactivation of PrP^Sc at pH 4.4, followed by 1.9 log₁₀ at pH 6.0 and 1.1 log₁₀ at pH 8.0. At 20°C, a CT of between 0.36 and 0.66 mg · liter⁻¹ min resulted in ≥4 log₁₀ inactivation at pH 4.4, followed by 2.4 log₁₀ at pH 6.0 and 1.9 log₁₀ at pH 8.0. At a CT value between 25.5 and 32.2 mg · liter⁻¹ min, PrP^Sc inactivation was ≥4 log₁₀ at pH 4.4 and both temperatures and at pH 6.0 and 20°C, while the inactivation was 3.6 and 2.9 log₁₀ at 4°C and pH 6.0 and 8.0, respectively (Table 2).

CT values of PrP^Sc were compared to those determined for other well-studied waterborne pathogens (Table 3). In general, PrP^Sc was found to be less resistant to ozone than some encysted waterborne protozoa (i.e., Cryptosporidium) or spore-forming bacteria (Bacillus subtilis) under similar experimental conditions. For example, inactivation of Cryptosporidium oocysts at pH 6 to 7 at 5°C required an ozone CT value of 32 mg liter⁻¹ min for 2 log₁₀ inactivation, whereas PrP^Res at pH 6 and 4°C required 1.33 mg liter⁻¹ min⁻¹ for 2.4 log₁₀ inactivation (i.e., Cryptosporidium was >24 times more resistant to ozone than PrP^Sc).

Table 3.

Summary of the efficacies of ozone inactivation of microorganisms in water and PrP^Sc

Microorganism or PrPSc	pH	Temperature (°C)	CT (mg liter−1 min)	Inactivation (log10)	Reference or source
E. coli	6–7	5	0.02	2	13
Rotavirus	6–7	5	0.006–0.06	2	13
Adenovirus 40	7	5	0.01–0.02	2	39
	7	5	0.07–0.60	4
Poliovirus	7.2	5	0.60	2	41
			1.20a	4a
		20	0.25	2
			0.50	4
Giardia lamblia cysts	7	5	1.30	2	41
			1.90	3
		20	0.48	2
			0.72	3
Cryptosporidium parvum oocysts	6–7	5	32	2	42
			47	3
		20	7.8	2
			12	3
PrPSc (263K scrapie)	4.4	4	0.59	2.8	This study
		20	0.59	≥4
	6.0	4	0.69	1.9
		20	0.66	2.4
	8.0	4	0.72	1.1
		20	0.36	1.9

^aExtrapolated by applying first-order kinetics with a safety factor of 3.

DISCUSSION

This study investigated the effectiveness of ozone inactivation for infectious prion protein (263K hamster scrapie, PrP^Sc) in aqueous solution, and this report is the first to provide ozone inactivation CT values for infectious prion protein (263K hamster scrapie; PrP^Sc). The current paper sets a base for understanding the conditions affecting ozone inactivation of prion proteins and consequently lays the foundation for modeling the kinetics of ozone inactivation of PrP^Sc (and other chemical disinfectants). A detailed understanding of the kinetics of ozone inactivation of PrP^Sc is instrumental for assessing the applicability and efficacy of new or existing technologies for mitigating prion contamination risks in water (i.e., SRM-generated wastewater).

Our data suggest that PrP^Sc is highly susceptible to inactivation by ozone. The ozone CT values derived for PrP^Sc in this study were considerably lower than those described for certain waterborne pathogens (i.e., Cryptosporidium) and spore-forming bacteria (i.e., B. subtilis) at comparable temperatures and pHs (Table 3). Although the applied ozone dose in this study was higher than is normally used for ozone applications in drinking water disinfection (due to the high ozone demand of the IBH), CT values provide a normalized approach to characterizing susceptibility of a particular microbial contaminant to ozone in an aqueous matrix. Ozone is used extensively for inactivation of pathogens and the degradation of various noxious chemicals in large-scale municipal drinking water and municipal/industrial wastewater treatment systems (10, 24, 29, 30). Application of ozone to these water matrices requires that ozone be (i) of a sufficient concentration to satisfy all oxidation (i.e., ozone) demands associated with organic loads (lipids, proteins, carbohydrates, etc.) and (ii) delivered efficiently under conditions of continuous flow in order to achieve the large-scale treatment volumes necessary in the water and wastewater treatment industry. At an industrial scale, continuous-flow reactors currently enable ozone doses (i.e., dissolved ozone concentrations) to reach 10 mg/liter, with a treatment capacity of >500 megaliters of water per day (i.e., >5,000 kg of ozone production per day). It has been recently reported that the use of a continuous flow reactor with an impinging bubble column contactor significantly increases the mass transfer rate of ozone into solution, achieving cumulative ozone doses > 300 mg/liter within 20 min of ozonation of a pulp mill effluent (8), and the mass transfer rate of ozone may be made higher by using multijet ozone contactors (3). Meanwhile, ozone engineering solutions of this magnitude, for the water treatment industry (specifically for disinfection purposes), are fundamentally based on the USEPA CT concept of ozone inactivation against target organisms or chemicals, emphasizing the importance of the present manuscript for derivation of a CT value for ozone inactivation of PrP^Sc. Understanding the resistivity of PrP^Sc to ozone based on a CT value, and comparing this value to those determined for other target microbes, provides critical insights into the practicality of applying existing industrial-scale ozonation systems to the control of PrP^Sc in wastewater produced by rendering facilities. Our data suggest that PrP^Sc is extremely susceptible to inactivation by ozone (compared to other microbes). Consequently, ozone technology solutions that currently exist in the water and wastewater industries may hold promise for control of prions present in wastewater from SRM-rendering facilities. Similarly, the data also suggest that ozone may be extremely valuable for other disinfection purposes, such as sterilization of medical instruments in hospitals.

Ozone inactivation was shown to be dose, contact time, pH, and temperature dependent. The pH conditions of 4.4, 6.0, and 8.0 were chosen to be representative of moderate acidic, slightly acidic, and slightly alkaline conditions for disinfection; higher pH was not selected due to the high ozone decomposition rate at alkaline conditions (16). Temperatures of 4°C and 20°C were chosen to represent typical disinfection temperatures associated with temperate climate conditions. Inactivation of PrP^Sc was greatest at a pH of 4.4 and lowest at the highest pH (8.0). The pH effect alone does not appear to act directly on the conformational stability of the infectious property of the prion agent, as pH by itself has been shown to have little effect on scrapie infectivity over the range of pH 2 to 10 (21). Since ozone decomposition has been shown to be more rapid at a higher pH (16), acidic conditions may favor the sustained and direct attack of molecular ozone on the prion protein itself or on biomolecular targets affecting the stability of the misfolded conformer (i.e., lipids) (2). Direct oxidation of protein targets by molecular ozone predominates reaction kinetics at lower pH, as indirect oxidative byproducts (i.e., hydroxyl radical generation) comprise a minor component of the oxidative potential at low pH (7). High inactivation rates of ozone at more acidic pHs have also been observed for Escherichia coli (35), norovirus (19) and helminth eggs (43). We are currently addressing what effects molecular ozone and its oxidative derivatives (OH., H₂O₂, etc.) have on inactivation of PrP^Sc as a means of characterizing the molecular mechanisms responsible for inactivation. Furthermore, optimization of the experimental methods and approaches to prion inactivation, as described here, lays the foundation for a more thorough examination of the kinetics of ozone inactivation of prions.

The effect of temperature on the inactivation of microorganisms and the degradation of organic pollutants by ozone have also been investigated by us and others (9, 18). In general, elevated temperatures result in a more rapid inactivation of microbial pollutants by ozone, an outcome also observed in the present study. Interestingly, temperature adjustments have opposing effects on ozone solubility and disinfection rates (4).

A recent study by Johnson et al. (14) demonstrated that UV-ozone treatment of hamster-adapted transmissible mink encephalopathy prions induced inactivation of PrP^Sc at ≥5 log₁₀. Ozone was generated by UV at a wavelength of 185 nm and then decomposed by UV at another wavelength (254 nm) to produce hydroxyl radicals. However, the UV-ozone system studied by Johnson et al. (14) produced limited ozone, thus generating limited hydroxyl radicals, requiring exposure times up to several weeks, and questioning the practicality of using UV-generated ozone as a decontamination/sterilizing approach to prion inactivation. Prion inactivation has also been studied with other advanced oxidation methods, such as the use of copper and hydrogen peroxide (17, 34), iron and hydrogen peroxide (35), photo-Fenton treatment (25), and titanium dioxide photocatalysis (26). Hydrogen peroxide inactivation (100 mmol/liter) in the presence of copper (0.5 mmol/liter) was reported to achieve ≥5.2 log₁₀ inactivation of 263K scrapie prion with a contact time of 2 h at room temperature (34). Increasing the concentration of hydrogen peroxide to 2.2 mol/liter reduced the contact time to 30 min for the same level of inactivation (17). Hydrogen peroxide inactivation (1.5 mol/liter) in the presence of Fe²⁺ (15.7 mmol/liter) and heating at 50°C for 22 h was able to produce an approximately 6 log₁₀ reduction of prion infectivity (35). PrP^Sc could also be degraded by ≥2.4 log₁₀ by photo-Fenton treatment (147 mmol/liter H₂O₂, 8.9 mmol/liter Fe³⁺) after 5 h of UV-A exposure (25) and be degraded by ≥2 log₁₀ by titanium dioxide photocatalysis (25 mmol/liter titanium dioxide, 118 mmol/liter H₂O₂) after 12 h of UV-A exposure (26). Due to the low sensitivity of the PrP^Sc detection methods used in the last two studies cited above, inactivation of more than 2.4 log₁₀ was not achieved. In these advanced oxidation studies, hydroxyl radicals as a sole component to inactivate PrP^Sc demonstrated their capabilities; however, prolonged exposure times (from 30 min to 22 h) were essential to continuously generate potent hydroxyl radicals sufficient for the inactivation. In contrast, ozone at a pH of 4.4 and 20°C in this study caused a very rapid ≥4 log₁₀ inactivation of PrP^Sc after 5 s of exposure, suggesting that ozone treatment might be more efficient for PrP^Sc inactivation than treatment with other advanced oxidants.

Conclusions.

PMCA is an extremely sensitive tool for detection and quantification of PrP^Sc and for measuring ozone inactivation of the template-directed misfolding properties of PrP^Sc. Ozone inactivation of scrapie 263K was shown to be dose, contact time, temperature, and pH dependent. In addition, ozone was found to be extremely effective with respect to inactivation of 263K scrapie, with more than 4 log₁₀ inactivation observed at a CT of 0.59 mg · liter⁻¹ min at pH 4.4 and 20°C. The derived ozone CT product for PrP^Sc was similar to that of poliovirus and considerably less than that of encysted protozoan parasites such as Cryptosporidium spp. and Giardia spp.