Comparison of conventional, amplification and bio-assay detection methods for a chronic wasting disease inoculum pool

Longitudinal studies of chronic wasting disease (CWD) in the native host have provided considerable understanding of how this prion disease continues to efficiently spread among cervid species. These studies entail great cost in animal, time and financial support. A variety of methods have emerged including transgenic mouse bioassay, western blot, enzyme-linked immunoassay (ELISA), immunohistochemistry (IHC), serial protein misfolding cyclic amplification (sPMCA) and real time quaking-induced conversion (RT-QuIC), that deepen our understanding of this and other protein misfolding disorders. To further characterize an inoculum source used for ongoing CWD studies and to determine how the readouts from each of these assays compare, we assayed a CWD-positive brain pool homogenate (CBP6) and a mouse dilutional bioassay of this homogenate using the above detection methods. We demonstrate that: (i) amplification assays enhanced detection of amyloid seeding activity in the CWD+ cervid brain pool to levels beyond mouse LD50, (ii) conventional detection methods (IHC and western blot) performed well in identifying the presence of PrPSc in terminal brain tissue yet lack sufficient detection sensitivity to identify all CWD-infected mice, and (iii) the incorporation of amplification assays enhanced detection of CWD-infected mice near the LD50. This cross-platform analysis provides a basis to calibrate the relative sensitivities of CWD detection assays.


Introduction
Chronic wasting disease (CWD) is unique among the prion diseases due to its capacity to infect a wildlife population and for its unparalleled transmission efficiency. Infectious particles shed in several bodily fluids and excretions from infected, yet asymptomatic cervids likely contribute to this efficient transmissibility [1][2][3][4][5][6]. Additional concern is warranted for cervids and species sympatric with cervid populations as the geographical distribution, host range [7,8,9] and strain identification [10][11][12] of CWD continues to expand, undoubtedly resulting in increased environmental contamination. Although no cases of CWD transmission to humans have been reported, its zoonotic potential remains a concern [10,13] due to the prevalence of asymptomatic CWD in deer a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 are adequate to detect prions in tissues harvested postmortem [46][47][48][49]. As expected, we were able to demonstrate detection of PrP Sc by WB in the 10 −1 dilution of unamplified CWD+ cervid brain homogenate following proteinase K (PK) digestion. No amyloid was detected in further dilutions of CWD+ cervid brain CBP6 (10 −2 to 10 −7 ), nor in age-matched naïve deer brain (Fig 1, Table 1). BioRad ELISA detected PrP Sc in 10 −1 and 10 −2 dilutions (OD readings of 3.999 and 0.259 respectively) of unamplified CWD+ cervid brain pool CBP6. No amyloid was detected in further dilutions (10 −3 -10 −9 ), nor in age-matched naïve deer brain ( Table 1).
Detection of seeding activity by new generation assays: sPMCA, RT-QuIC or sPMCA/ RT-QuIC. We and others [50,51] have demonstrated that tissues harvested from infected hosts, especially those in the asymptomatic phase of disease, may contain low levels or variably PK resistant forms of the misfolded protein. The presence of very low concentrations of prion deposition diminishes ability to rely solely upon IHC, ELISA or WB for the detection of PrP Sc deposition associated with prion disease. Thus, we assessed and compared endpoint dilution of the same CWD+ cervid brain (CBP6) dilutional series using these assays-sPMCA, RT-QuIC and RT-QuIC readout of sPMCA 5 th round product (sPMCA/RT-QuIC). The conventional readout for prion seeding activity generated by sPMCA is WB (sPMCA/ WB). We subjected a dilutional series of naive and CBP6 CWD+ cervid brain homogenate (10 −2 to 10 −7 ) to 5 rounds of sPMCA. When assessed by WB, we revealed the presence of amyloid seeding activity to an endpoint dilution of 10 −6 (Fig 1, Table 1). Upon assessment of the same CWD+ cervid brain dilutional series by RT-QuIC we determined the endpoint for amyloid seeding activity to 10 −7 (Fig 2, Table 1). We have recently used RT-QuIC as a readout for sPMCA product [52] (sPMCA/RT-QuIC). Incorporation of RT-QuIC as the readout for sPMCA 5 th round product improved detection of amyloid seeding activity in CWD+ cervid brain homogenate to 10 −8 (Fig 2, Table 1). No amyloid seeding activity was detected at any dilution of the naïve brain homogenate post sPMCA/WB, RT-QuIC or sPMCA/RT-QuIC amplification.
sPMCA/RT-QuIC optimum seed input. RT-QuIC amyloid seeding activity can be inhibited at high seed concentrations and rescued by dilution [5,53,54]. To establish the optimal seed input for maximal amyloid seeding detection by sPMCA/RT-QuIC we assessed a Comparison of amyloid seeding activity in a dilutional series of CWD+ cervid brain by RT-QuIC and sPMCA/RT-QuIC. Amyloid seeding activity was detected by RT-QuIC in the 10 −3 -10 −7 dilutions of CWD+ cervid brain pool CBP6 (2 μg-200 pg seed material; blue circles and blue median lines). sPMCA/RT-QuIC assay (round 5 sPMCA product diluted 1:100) detected amyloid seeding activity in the 10 −2 -10 −8 dilutions of CBP6 (20 μg-20 pg seed material; red squares and red median lines). Neither assay detected significant amyloid seeding activity in the 10 −9 dilution of CBP6 (2 pg seed material) or in naïve cervid brain homogenate dilutions (100 replicates U ). 7-12 replicates are shown per assay dilution. Statistical significance between assays is indicated with an asterisk (� 0.0017). Statistical significance between samples and same dilution negative controls is indicated with Δ RT-QuIC and^sPMCA/ RT-QuIC (� 0.0001). U negative controls from all dilutions were combined as no dilutional effect was noted. dilutional series of CBP6 5 th round sPMCA product by RT-QuIC (Fig 3). We demonstrate that RT-QuIC amyloid seeding activity is rescued when 5 th round sPMCA CBP6 is diluted beyond a 1:10 dilution (Fig 3). We also reveal maximal readout for the combined assay when CBP6 sPMCA 5 th round product (initiated by 1:1000 dilution) is spiked at 1:100 dilution into RT-QUIC (Fig 3, asterisk). This combination of dilutions is used for sPMCA/RT-QuIC throughout the remainder of this publication. Negative controls run with each dilution remained below the threshold and were combined for ease of graphics (Fig 3).

Prion detection in brain tissue harvested from CWD+ CBP6 mouse bioassay
The use of conventional and amplification assays has contributed much to our current understanding of amyloid disorders, including prions. Yet, a limitation of these assays is their Detection of amyloid seeding activity by sPMCA/RT-QuIC in a CWD+ cervid brain dilutional series. Amyloid seeding activity was detected in 10 −2 -10 −8 dilutions of CWD+ cervid brain (CBP6) (20ug-20pg) using 1:100-1:100000 dilutions of sPMCA 5 th round product. Amyloid seeding activity was inhibited at the 1:10 dilution. Optimal prion detection was demonstrated at a 1:100 dilution (yellow bars) of sPMCA 5th round product. No significant amyloid seeding activity was initiated by 10 −2 -10 −9 dilutions of naïve cervid brain using 1:100-1:100000 dilutions of sPMCA 5 th round product. n = 7-8 RT-QuIC replicates/sample. n = 235 negative control dilution replicates. inability to directly measure infectivity. Bioassay remains the gold standard to assess infectivity. To determine the role in vitro analysis may have in providing insight to prion infectivity we compared results from five of these methods. To this end, brain tissue harvested from a dilutional mouse bioassay of CBP6 were assessed by conventional (IHC, WB) and amplification (sPMCA, RT-QuIC and sPMCA/RT-QuIC) assays. Insufficient quantities of mouse brain tissue negated inclusion of these samples for BioRad ELISA analysis.

RT-QuIC and sPMCA/QuIC amplification assays detect amyloid conversion beyond mouse bioassay LD 50
The incorporation of amplification assays, sPMCA and RT-QuIC, enhanced the detection of prions in a CWD+ cervid brain pool CBP6 to levels beyond the mouse bioassay LD 50 (Fig 4,  Table 1). This may be due to a threshold requirement (minimum infectious dose: MID) to initiate infectivity in vivo. The existence of a MID threshold for viral and bacterial infectious agents has been observed [55]. While a small dose of viral or bacterial antigen may initiate a vaccinating immune response (live attenuated vaccine strategies), to initiate infection and disease progression a larger dose of pathogen is required. For prions this may be reflected by the initial success within the endoplasmic reticulum (ER) to activate the unfolded protein response (UPR) in an attempt to restore normal protein folding and ER function [56, 57], resulting in prolonged development of sufficient accumulation of amyloid required to initiate prion disease. Alternatively, these findings may suggest that a fraction of conversion competent amyloid, as detected by in vitro amplification assays, is noninfectious. Previous studies have demonstrated that both infectious and noninfectious amyloid particles exist in prion inoculums [58]. It is also possible that this attests to the in vitro amplification process being more efficient at initiating amyloid conversion events than its in vivo counterpart.

sPMCA and RT-QuIC amyloid seeding activity is inhibited at high seed concentrations and rescued by dilution
We confirm earlier reports [36, 54, 59] that sPMCA and RT-QuIC amyloid seeding activity remains undetectable in tissue and biological fluid samples when spiked into either assay at high concentrations (10 −1 -10 −2 dilution) (Figs 2 and 3). Here, employing the CWD+ cervid brain pool CBP6, we demonstrate that amyloid detection in sPMCA product by RT-QuIC is rescued and provides peak detection when dilutions for both sPMCA and RT-QuIC are optimized (Fig 2). Observed inhibition may be related to components or matrix effects that occur in less dilute samples. Optimization may be required when assessing seeding activity by single or combined use of amplification assays to account for variation in amyloid conversion rates dependent upon tissue/fluid type, host species, prion strain or PRNP genotype.

Amplification assays detect prions in terminal brain tissue with similar efficacy as conventional IHC
We found that conventional IHC detected prions (PrP Sc deposition) in mouse bioassay terminal brain tissue with similar efficacy as the new generation of amplification assays (prion seeding activity) (Figs 4 and 5, S1-S3 Figs, and Table 2). These findings affirm the robust nature of conventional test assays and supports their continued use in the application of testing tissues harvested from terminal, end stage disease cases, when PrP Sc levels are high. The detection of high concentrations of prions by both conventional and amplification assays has been previously substantiated in tissues harvested during the late and terminal phases of prion disease [60][61][62][63][64]. Of increasing interest is the ability to detect very low concentrations of prions associated with early infection and silent carriers of disease. Conventional assays have been used to detect prions in lymphoid biopsies harvested from hosts lacking signs/symptoms of disease [65][66][67]. Yet, these methods present concerns regarding diagnostic accuracy due to limitations associated with the amount of tissue that can be analyzed, the low detection capacity inherent to these methods, and by the presence of protease sensitive prion isoforms that may be abolished by PK digestion [68].

New generation in vitro amplification assays provide improved detection rigor
We demonstrate that amplification assays can improve detection rigor as prion burden decreases (Figs 3 and 5). The ability to assess 8-16 replicates from a single sample improves readout reliability, and thus provides a screening mechanism for the detection of very low concentrations of prions present shortly after prion exposure, or for the assessment of nonclassical and asymptomatic disease states. These findings are supported by previous studies that have employed these techniques to demonstrate prions in samples harvested at very early time points post experimental infection or during the asymptomatic phase of disease [6,18,38,40,43,44,50,[69][70][71][72].

In vitro amplification assays provide enhanced detection of prion infection at the mouse bioassay LD 50
As mentioned above, our study finds that in vitro amplification by RT-QuIC and sPMCA/ RT-QuIC demonstrate enhanced detection of seeding activity in a CWD+ cervid brain pool homogenate than that reached by conventional test results from mouse bioassay of the same homogenate. Further analysis of terminal mouse brain tissue from the dilutional bioassay by RT-QuIC and sPMCA revealed 2 additional prion-infected mice at the mouse bioassay LD 50 (Fig 5). These findings further corroborate the published LD 50 established for this CWD + brain pool [36]. Interestingly, use of sPMCA/RT-QuIC revealed an increase in amyloid seeding rates for most of the bioassay mice (Fig 5). Both findings may have impact upon earlier detection and diagnosis of disease status, and prion detection in samples containing minute quantities of amyloid seeding activity. Our findings are isolated to the detection of prions in brain tissues and a single CWD+ brain pool, but are supported by a recent report [52] where the use of sPMCA/RT-QuIC demonstrate enhanced detection of seeding activity in biological fluids containing very low prion burdens. These findings may be especially important for antemortem prion diagnosis. Bioassay remains the gold standard and definitive readout for the detection of prion infectivity. Its use has provided tremendous insight to our understanding of prion pathogenesis and transmission dynamics [3,71,[73][74][75][76][77][78]. Yet, it is costly in money, time and animal use. Bioassay of prion concentrations near the LD 50 requires the use of 9-12 animals/cohort and between 400-500 days to reach a study endpoint. Conservatively, adding in animal costs, per diem and assessment of tissue, this equates to thousands of dollars and 1-2 years.

Future use for new generation amplification assays
This new generation of in vitro amplification assays have been reported to obtain sensitivity levels rivaling that of bioassay [27,79]. Our results confirm that sPMCA and RT-QuIC detect prions to bioassay endpoint (Fig 5, Table 2). We also show that RT-QuIC can detect additional seeding activity not demonstrated when sPMCA is analyzed by WB. Prion infectivity is known to consist of variably protease sensitive/resistant isoforms [80]. RT-QuIC avoids the use of proteinase K, and thus, may enhance detection of protease sensitive isoforms. While more testing is required, we propose that RT-QuIC and sPMCA/RT-QuIC may be useful to assess prion infectivity instead of bioassay.
Recent evidence suggests that other neurodegenerative diseases such as Parkinson's Disease (PD), Alzheimer's Disease (AD), and Huntington's disease produce prion-like aggregates [81]. Studies have utilized sPMCA [82][83][84] and RT-QuIC [30, 84] to detect α-synuclein proteins and β-amyloid oligomers in cerebral spinal fluid of AD and PD patients, respectively. Modified PMCA is able to detect polymers with high sensitivity and specificity, and has the capacity to distinguish these neurodegenerative diseases from others [82]. Similar to prion diseases, PD, AD and other neurodegenerative diseases have a long latency period during which time detection is limited. Current diagnostic methods for these human protein misfolding disorders rely upon imaging techniques, clinical examination and screening for potential biomarkers, however, a definitive diagnosis can only be made from post-mortem histology [85]. Earlier diagnosis will be key to continued development of therapeutics for all neurodegenerative diseases.
In summary, we provide evidence that the use of 'new generation' amplification assays enhance prion detection sensitivity and improve test validity and rigor for the analysis of biological samples containing very low concentrations of prions. Methods employing in vitro detection will never replace bioassay, but they can augment findings and become tools for the early diagnosis of human and animal protein misfolding disorders.

Ethics statement
All animals were handled in strict accordance with guidelines for animal care and use provided by the United States Department of Agriculture (USDA), National Institutes of Health (NIH) and the Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC), and all animal work was approved by Colorado State University Institutional Animal Care and Use Committee (IACUC protocol numbers 10-2189A, 12-3773A, and 13-4482A).
Cervids handling. Naïve deer brain homogenate: Naïve brain homogenate derived from pooling the brain tissue harvested from two naïve and IHC confirmed CWD-free white-tailed deer served as a negative control. Hereafter referred to as naïve cervid brain.
CWD-infected deer brain homogenates (CBP6): CWD positive inoculum was derived from pooling the brain tissue harvested from six terminal CWD-infected and IHC-confirmed positive white-tailed deer (CBP6) generated from previous experimental studies conducted at CSU [86,87]. The inoculum is hereafter referred to as CWD+ cervid brain CBP6. The deer were acquired from Warnell School of Forestry, University of Georgia-Athens. They were housed socially in 6×12 meter suites with sand and epoxy mixture sealed floors to afford normal wear on hooves. The sealed floors were covered in a layer of aspen chip bedding. A natural light equivalent was mimicked, humidity was maintained at 25-40%, and temperature ranged from 16-26 degrees Celsius. Enrichment consisted of cardboard boxes, ropes and plastic toys, and daily interaction with caretaker personnel. Deer received 50 g complete pelleted feed per kg/ day along with hay forage and water ad libitum. For sample collection (including prior to euthanasia), deer were anesthetized with IM injections of Ketamine (2-8 mg/kg) and Medetomidine (0.1-0.2 mg/kg). Euthanasia was performed by IV injection of pentobarbital sodium with phenytoin (1 ml per 4.5 kg).
Mice handling. Naïve mouse brain homogenate: Brain homogenate from transgenic mice overexpressing the elk prion protein, Tg(CerPrP) 5037, [88] inoculated with naïve cervid brain served as a negative control for all assays.
Bioassay mouse brain tissues: Tg(CerPrP) 5037 mice were used to establish an LD 50 of CBP6 CWD+ cervid brain [36]. Mice were housed socially in a commercial caging system in cages filled with aspen chips, provided nesting materials, checked daily, and provided commercial irradiated rodent chow and water ad libitum. Prior to inoculation, mice are given a 2 mg/kg dose of meloxicam analgesic. Briefly, mouse cohorts (n = 6/cohort) were inoculated intracranially in the right parietal lobe of the cerebral cortex with 30 μl of 10 −1 -10 −7 dilutions of CBP6 and were observed for signs of prion disease. Mice were terminated by inhalation of CO 2 upon development of prion disease or 500 days post infection (dpi). Brain tissue harvested from each inoculated mouse was divided in half with one half frozen at -80˚C and one half fixed in periodate-lysine-paraformaldehyde (PLP) fixative for a minimum of 48 h. After a minimum of 48 h fixed tissues were transferred to 1xPBS at room temperature until further analysis.

Western blot (WB)
Brain tissue from Tg(CerPrP) 5037-inoculated mice was prepared as a 10% (w/v) homogenate in 1xPBS and stored at 4˚C until further analysis. Homogenates were mixed with proteinase K (PK; Invitrogen) at 50 μg/mL, incubated at 37˚C for 30 min followed by 45˚C for 10 min. Samples were mixed with reducing agent (10x)-lithium dodecyl sulfate (LDS) sample buffer (4x) (Invitrogen) at a concentration of 1x, heated at 95˚C for 5 min and separated on NuPAGE 12% Bis-Tris gel at 125V for 1.5 h. Protein was transferred to a polyvinylidene fluoride (PVDF) membrane at 80 V for 1 h in transfer buffer (0.025 M Trizma base, 0.2 M glycine, 20% methanol, pH 8.3). The membrane was then incubated with 5% nonfat milk in 1xTris-buffered saline (TBS) with 0.1% Tween 20 (TBST) for 3 min and then for 12 min with BAR224-HRP (0.2 μg/ml final concentration; Cayman Chemical) diluted in TBST, followed by a 30 min wash with TBST. The membrane was then developed with ECL Plus Western blotting detection reagents (Pierce) and viewed on a Luminescent Image Analyzer LAS-4000 (Fujifilm).

Serial protein misfolding cyclic amplification (sPMCA)
Tg(CerPrP) 5037normal brain homogenate (NBH): NBH, the PrP C substrate used for sPMCA prion conversion, was prepared as follows: naïve Tg(CerPrP) 5037 mice <4 months of age were euthanized by CO 2 inhalation and perfused with 35 mL of 5 mM ethylenediaminetetraacetic acid tetrasodium salt (EDTA) in PBS via intracardiac catheterization. The brain was removed and flash frozen using liquid nitrogen. Brain homogenate was then prepared at a 10% (w/v) solution in sPMCA buffer (1% Triton-X 100 [v/v], 5 mM EDTA, and 150 mM NaCl) with the addition of Complete Protease Inhibitors (Roche Pharmaceuticals, Indianapolis, IN) in a homogenizer (Omni Bead Ruptor). Homogenates were then centrifuged for 1 min at 3000 rpm to remove bulk brain material, and the supernatant frozen in single-experiment aliquots at -80˚C in a prion-free room until use in sPMCA.
sPMCA: Thirty microliters (30 μl) of each sample (3 replicates for each sample, 2 investigators/sample; investigators were blinded to sample identity) was spiked into 50 μl of NBH and sonicated for 10 sec every 5 min for each 24 h round. After each 24 h round 30 μl of sonicated material was transferred into 50 μl of fresh NBH and subjected to another 24 h round for a total of 5 rounds. After 5 rounds material was analyzed by WB as described previously [71]. Naïve and CWD+ cervid brain inoculum (CBP6) served as positive and negative sPMCA controls.

Real-time quaking induced conversion assay (RT-QuIC)
Substrate for seeded RT-QuIC reactions was prepared by adding truncated Syrian Hamster recombinant protein encoding residues 90-231 (SHrPrP) prepared as previously reported [38] to RT-QuIC buffer (320 mM NaCl, 1.0 mM EDTA, 10 μM Thioflavin T [ThT, Sigma]) at a final concentration of 0.1 mg/ml. Ninety-eight microliters (98 μl) RT-QuIC substrate was added to optical bottom black 96-well plates (Nunc). Each mouse brain homogenate from the mouse CBP6 bioassay was diluted 10 −2 to 10 −7 , and 2 μl of each was run in quadruplicate on 2-3 plates by 2-3 investigators for a total of 7-12 replicates/sample. Positive and negative controls were included on all plates. Prepared plates were placed in a BMG Labtech Polarstar fluorometer and subjected to 700 rpm double-orbital shaking for 1 min ever other min for 15 min for 250 cycles. After each cycle ThT fluorescence was read at an excitation of 450 nm and emission of 480 nm. Gain was set at 1700 and read using orbital averaging with 20 flashes per well with the 4 mm setting. Fluorescent readings were recorded for all sample reactions for a total time of 62 h at a temperature of 42˚C. Samples were considered positive if they crossed a threshold (5 SD above the mean of the initial 5 readings). The inverse of the time when the reaction reached the threshold (1/time to threshold) was then used to determine the amyloid formation rate. Statistical analyses were run in Prism v6, GraphPad Software, La Jolla, CA. A Mann-Whitney test was used to generate (p-values less than 0.05 were considered significant) by comparing the sample rates to the rates of known negative control tissues. A Reed-Muench calculation was used to calculate the LD 50 for the titrated mouse bioassay as previously described [36].