Antemortem Detection of Chronic Wasting Disease Prions in Nasal Brush Collections and Rectal Biopsy Specimens from White-Tailed Deer by Real-Time Quaking-Induced Conversion

Abstract

Chronic wasting disease (CWD), a transmissible spongiform encephalopathy of cervids, was first documented nearly 50 years ago in Colorado and Wyoming and has since spread to cervids in 23 states, two Canadian provinces, and the Republic of Korea. The expansion of this disease makes the development of sensitive diagnostic assays and antemortem sampling techniques crucial for the mitigation of its spread; this is especially true in cases of relocation/reintroduction of farmed or free-ranging deer and elk or surveillance studies of private or protected herds, where depopulation is contraindicated. This study sought to evaluate the sensitivity of the real-time quaking-induced conversion (RT-QuIC) assay by using recto-anal mucosa-associated lymphoid tissue (RAMALT) biopsy specimens and nasal brush samples collected antemortem from farmed white-tailed deer (n = 409). Antemortem findings were then compared to results from ante- and postmortem samples (RAMALT, brainstem, and medial retropharyngeal lymph nodes) evaluated by using the current gold standard in vitro assay, immunohistochemistry (IHC) analysis. We hypothesized that the sensitivity of RT-QuIC would be comparable to IHC analysis in antemortem tissues and would correlate with both the genotype and the stage of clinical disease. Our results showed that RAMALT testing by RT-QuIC assay had the highest sensitivity (69.8%) compared to that of postmortem testing, with a specificity of >93.9%. These data suggest that RT-QuIC, like IHC analysis, is an effective assay for detection of PrP^CWD in rectal biopsy specimens and other antemortem samples and, with further research to identify more sensitive tissues, bodily fluids, or experimental conditions, has potential for large-scale and rapid automated testing for CWD diagnosis.

INTRODUCTION

Chronic wasting disease (CWD) is an efficiently transmitted spongiform encephalopathy of cervids (e.g., deer, elk, and moose) and is the only known prion disease affecting both farmed and free-ranging, nondomestic animals. It is the only prion disease of animals the control and eradication of which, through movement restrictions, genotypic breeding schemes, or herd reduction/depopulation efforts, for example, are problematic (1, 2). While the origins of CWD are uncertain, the disease has been present in cervid populations of northern Colorado and southern Wyoming for nearly 50 years (3, 4) and has now been identified in both captive and free-ranging cervids in 23 states, two Canadian provinces, and the Republic of Korea (5, 6). With intensified national and international surveillance efforts, CWD continues to be identified in areas previously considered free of infection, including recent discoveries in Iowa, Texas, Pennsylvania, and Ohio (7, 8, 9, 10). The prevalence of CWD varies from 0 to 30% among free-ranging populations (11, 12) but may approach 80% in cervid farm operations under quarantine (13).

The expanding distribution of CWD across North America can be considered to have followed two nearly distinct pathways: (i) gradual proliferation of the disease among free-ranging cervids, with an often low rate of diffusion and stable or slowly increasing prevalence, and (ii) interstate and international dissemination among farmed cervid herds, with a potential for erratic geographic manifestation and rapidly escalating prevalence (5). Infrequently, though with potentially calamitous results, these pathways may intersect—wherein infection may spill over from one to the other—though little has been reported to substantiate the role of captive cervid operations in the expansion of the CWD range in wild deer and elk or its converse. Epidemiologic investigations are necessary to demonstrate and further substantiate the frequency of these events. A reliable and sensitive postmortem or, more importantly, antemortem testing strategy for farmed cervids may have a role in impeding the broadening geographic distribution of CWD among captive animals and the potential for its local transmission between farmed and free-ranging deer and elk.

Postmortem testing is currently the standard means of identifying CWD-infected cervids by evaluating the brainstem at the level of the obex and medial retropharyngeal lymph node (RLN) by either immunohistochemistry (IHC) analysis or enzyme-linked immunosorbent assay (ELISA). IHC analysis, the gold standard for CWD regulatory testing in the United States, has identified prion infection in the deer RLN as early as 3 to 6 months into the course of the disease and in the brainstem as soon as 6 to 9 months postexposure (14). Antemortem testing of peripheral lymphoid tissues, including tonsil and recto-anal mucosa-associated lymphoid tissue (RAMALT), by IHC analysis has demonstrated relatively high sensitivity in the context of postmortem testing (15, 16, 17). It is generally acknowledged that conventional assays, including IHC analysis and ELISA, underestimate the level of protease-resistant prion protein (PrP^res) in a given sample because of the necessity of harsh chemical pretreatments (18, 19, 20). In some cases, this suspicion has been confirmed by bioassay of IHC analysis-negative tissues. This shortcoming has led to the development of assays that utilize the amplification of PrP^res (e.g., serial protein misfolding cyclic amplification [20, 21]), fluorometric quantitation of seeding activity (e.g., real-time quaking-induced conversion [RT-QuIC] [22, 23, 24]), or other assays that avoid harsh proteolytic treatments, (e.g., the conformation-dependent immunoassay [25]). Such assays have the potential for increased sensitivity and earlier detection of CWD-positive animals (19, 20, 26, 27), with RT-QuIC previously demonstrating practicality in the detection of PrP^CWD in a number of tissues and bodily fluids, including cerebrospinal fluid (CSF), urine, saliva, blood, brain tissue, and lymph node tissue (22, 26–33).

In the present study, we hypothesized that RT-QuIC analysis of RAMALT and nasal brush samples collected antemortem would be as sensitive as conventional IHC analysis of RAMALT biopsy specimens, with results that would have a high correlation with postmortem IHC analysis of RLN and obex tissues. We used a standardized RT-QuIC assay to blindly examine antemortem samples collected from 409 farmed white-tailed deer (Odocoileus virginianus) from three sites across the United States. RT-QuIC results were then correlated with ante- and postmortem results of IHC analysis of RAMALT, RLN, and brainstem samples (including obex scoring), as well as the prion genotype (PRNP), to estimate the utility of this approach in large-scale antemortem CWD surveillance of farmed cervids.

MATERIALS AND METHODS

Study populations.

The study populations consisted of two groups of farmed cervids in the United States under quarantine with a history of CWD (herd 1, n = 14; herd 2, n = 356). A separate herd outside the zone where CWD is endemic and with no history of CWD was included as a negative-control group for antemortem collections (herd 3, n = 39). Study animals included both males and females from a range of age groups and management styles. Antemortem samples included RAMALT biopsy specimens and nasal brushings collected in accordance with Institutional Animal Care and Use Committee (IACUC) protocols (KSU3503). DNA extracted from a section of rectal tissue was used to determine the PRNP genotype in herds 1 and 2 {specifically, positions 95 [glutamine (Q) or [histidine (H)], 96 [glycine (G) or serine (S)], 116 [alanine (A) or glycine]; and 226 [glutamine or lysine]} as described by O'Rourke et al. (34). Postmortem samples collected from animals in positive herds included RLN and brainstem samples at the level of the obex.

Tissue collection and processing.

Samples were collected from the negative-control group while the animals were under minimal restraint in a conventional deer chute handling system. Those from positive groups were collected while the animals were under sedation with a standard deer immobilization mixture of butorphanol (27.3 mg), azaperone (9.1 mg), and medetomidine (10.9 mg) given intramuscularly. RAMALT biopsy specimens were collected according to IACUC-approved protocols as follows. With sterile, single-use sampling instruments, RAMALT samples were collected by removing a 1.0-by-0.5-cm section of mucosa from the wall of the rectum approximately 1.0 cm anterior to the mucocutaneous junction of the anus and perpendicular to the oral/aboral axis of the rectum. The sample was divided into two equal pieces, one frozen and maintained at −80°C and the remainder placed in 10% neutral buffered formalin prior to IHC analysis. Frozen RAMALT biopsy specimens were later prepared as an ∼2% homogenate in phosphate-buffered saline (PBS) with a TissueLyser II (Qiagen) with a single 5-mm stainless steel bead and 2-ml conical snap cap tubes, by using two 2-min cycles of homogenization at a power setting of 20. Homogenates were then maintained at −80°C until analysis by RT-QuIC assay and DNA extraction for genotyping.

Nasal brush samples were cleanly collected from the right nasal cavity concurrently with RAMALT biopsy specimens as follows. A sterile 20-cm uterine cytology brush (Cancer Diagnostics no. SPH500) was gently inserted into the right nasal vestibule, directed dorso-caudally through the dorsal nasal meatus, and fed in approximately 12 to 15 cm until located directly rostral to the ethmoid turbinate (Fig. 1) The brush was spun gently to collect turbinate epithelial tissue and then retracted from the nasal cavity, placed in PBS, and refrigerated at 4°C between collection and processing. Samples were processed by vigorous vortexing in PBS to remove and suspend cellular matter present on the brush. The cellular suspension was then centrifuged at 3,000 × g for 10 min at 4°C. The supernatant from the cellular suspensions was poured off, and the cellular pellet was resuspended in 0.5 ml of PBS and homogenized as described above. Homogenates were then maintained at −80°C until analysis by RT-QuIC assay.

FIG 1.

Collection of nasal brush samples from farmed white-tailed deer. (A) The 20-cm brush was inserted into the right nostril and directed dorso-caudally to the level of the ethmoid turbinates. The brush was then gently twisted to collect epithelium, removed, and placed in PBS. (B) Cadaver specimen showing the anatomic location of sample collection from white-tailed deer.

Following euthanasia, samples of the brainstem at the level of the obex and medial RLN were collected and preserved in 10% neutral buffered formalin.

IHC analysis.

RAMALT biopsy specimens were assayed for PrP^CWD by IHC analysis as previously described (35, 36). Briefly, tissue was preserved in 10% neutral buffered formalin and then embedded in paraffin blocks. Tissue sections 5 μm thick were mounted on glass slides and deparaffinized before treatment with 99% formic acid. IHC staining for PrP^CWD was performed with the primary antibody anti-prion 99 (Ventana Medical Systems, Tucson, AZ) and then counterstained with hematoxylin. Biopsy specimens were considered positive if follicles stained with coarse, bright red, granular material (36). The numbers of staining and nonstaining follicles in each rectal biopsy specimen were documented. Samples not demonstrating IHC staining were considered “not detected.” A similar protocol was used for postmortem obex and RLN analysis, with the amount and location of IHC staining in the obex semiquantitatively evaluated to estimate disease progression (17). The obex sections were scored from 0 to 4 on the basis of the following criteria: grade 0, no IHC staining within the obex; grade 1, IHC staining only within the dorsal motor nucleus of the vagus (DMNV); grade 2, IHC staining within the DMNV and area postrema with or without focal staining in the nucleus of the solitary tract (NST) and adjacent white matter; grade 3, IHC staining in the DMNV and NST with light to moderate staining extending into other nuclei and white matter; grade 4, heavy IHC staining of the DMNV, multiple other nuclei, and white matter throughout the obex.

RT-QuIC preparation and procedure.

RT-QuIC assays were performed with a truncated form of recombinant Syrian hamster PrP (SHrPrP; residues 90 to 231) in pET41b and expressed and purified as previously described (26). In brief, 1-liter cultures of lysogeny broth (LB) containing autoinduction supplements (EMD Biosciences) were inoculated with SHrPrP-expressing Rosetta strain Escherichia coli, which was grown overnight and harvested when an optical density at 600 nm of ∼3 was reached. Cells were lysed with BugBuster reagent with supplemented Lysonase (EMD Biosciences), and inclusion bodies (IB) were harvested by centrifugation of the lysate at 15,000 × g. IB pellets were washed twice and solubilized overnight in 8 M guanidine hydrochloride (GuHCl) in 100 mM NaPO₄ and 10 mM Tris (pH 8.0), clarified by centrifugation at 15,000 × g for 15 min, and added to Superflow nickel-nitrilotriacetic acid (Ni-NTA) resin (Qiagen) preequilibrated with denaturing buffer (6.0 M GuHCl, 100 mM NaPO₄, 10 mM Tris, pH 8.0). Denatured SHrPrP and Ni-NTA resin were incubated by rotation at room temperature for 1 h and then added to an XK fast protein liquid chromatography column (GE Healthcare). Refolding was achieved on a column with a linear refolding gradient of denaturing buffer to refolding buffer (100 mM NaPO₄, 10 mM Tris, pH 8.0) over 3 h at 1.5 ml/min. SHrPrP was eluted with a linear gradient of refolding buffer to elution buffer (100 mM NaPO4, 10 mM Tris [pH 8.0], 500 mM imidazole [pH 5.5]) over 40 min at 2.0 ml/min. Peak UV 280-nm fractions were pooled and dialyzed overnight against two changes of 4.0 liters of dialysis buffer (20 mM NaPO₄, pH 5.5). Recovered SHrPrP was adjusted to a final concentration of ∼0.5 mg/ml and stored at 4°C for up to 45 days.

Nasal brush preparations were diluted 1:10 in RT-QuIC dilution buffer, while rectal biopsy specimen homogenates were diluted 1:100 in RT-QuIC dilution buffer. Five microliters of this 10⁻¹ or 10⁻² dilution was added to 95 μl of RT-QuIC reaction buffer, consisting of 50 mM NaPO₄, 350 mM NaCl, 1.0 mM EDTA tetrasodium salt, 10 μM thioflavin T (ThT), and 0.1 mg/ml truncated SHrPrP^C, to yield a final volume of 100 μl. Each test sample was prepared in triplicate on a single plate in two separate experiments. Positive controls, consisting of 5 μl of a 1 × 10⁻³% homogenate (∼5 μg) of pooled brain tissue from six white-tailed deer experimentally infected with CWD (cervid brain pool 6) spiked into 95 μl of RT-QuIC reaction buffer, were included in triplicate in each experiment. Negative controls, also prepared in triplicate, consisted of three RAMALT biopsy specimens or nasal brushings from deer from a CWD-free captive facility that were known to be negative, as well as untreated RT-QuIC reaction buffer spiked with 5 μl of RT-QuIC dilution buffer. Reactions were prepared in a black 96-well, optical-bottom plate that was then sealed and incubated in a BMG Labtech Polarstar fluorimeter at 42°C for 24 h (96 15-min cycles) with intermittent shaking cycles, specifically, 1-min shaking periods (700 rpm, double orbital pattern) alternating with 1-min rest periods. ThT fluorescence measurements (450-nm excitation and 480-nm emission wavelengths) were taken every 15 min with the gain set at 1,200. The relative fluorescence (in relative fluorescence units [RFU]) of each triplicate sample was progressively monitored against time with orbital averaging and 20 flashes/well at the 4-mm setting.

A replicate well was considered positive when the relative fluorescence crossed a predefined threshold, calculated as 10 standard deviations above the mean fluorescence of all of the samples across cycles 2 through 8. Times at which a sample crossed the positive threshold (threshold cycle) were recorded; samples remaining below the threshold were considered negative. Because of the rare potential for spontaneous amplification in true-negative samples, only samples amplifying in ≥2 of 6 replicates were considered positive.

Correlation of RT-QuIC results with CWD IHC analysis, obex scoring, and the PRNP genotype.

We sought to examine if RT-QuIC results from RAMALT and nasal brush collections could be associated with a number of predictor variables, including genotype and both ante- and postmortem IHC analysis results. We used Spearman correlations to assess the relationship and direction of relationship between RT-QuIC results obtained with RAMALT and nasal brush samples to results obtained with previously used invasive measures of detecting CWD.

RESULTS

Population data and IHC analysis for CWD.

In herd 1, five CWD-positive animals were identified (35.7%) through postmortem testing. Three of these were positive by IHC analysis of brainstem and RLN samples, and each of these was also positive by RAMALT IHC analysis (60%); the remaining two were RLN positive only. The mean follicle count in RAMALT biopsy specimens was 5.5 (range, 0 to 15); 6 or 7 follicles were observed in the two RAMALT biopsy specimen-negative/RLN positive animals. All positive animals were considered wild type homozygous for PrP codons 95Q, 96G, 116A, and 226Q (QGAQ), though there were a range of PRNP alleles present in the population (Tables 1 and 2).

TABLE 1.

Summary of study populations, including sex, allelic frequency, samples collected, and postmortem CWD status

Herd	No. of males	No. of females	Allelic frequency				No. RAMALT sampled	No. nasal brush sampled	No. CWD positive postmortem	No. CWD negative postmortem
			QGAQ	GSAQ	HGAQ	QGAK
1	3	11	0.71	0.21	0.036	0.036	14	14	5	9
2	183	173	0.63	0.36	0.011	0.0042	355	356	284	72
3	10	29	NDa	ND	ND	ND	39	39	NDb	NDb

^aND, not determined.

^bThese animals were not tested postmortem, as they were considered a CWD-negative herd and not subject to quarantine and euthanasia.

Among the herd 2 animals, 284 (79.8%) were CWD positive. Two hundred three of these were positive by IHC analysis of brainstem and RLN samples; of these, 119 were also RAMALT positive by IHC analysis. Seven were positive by RLN and RAMALT analysis only, while an additional 74 were positive by RLN analysis alone. The mean biopsy specimen follicle count was 3.1; of the 165 RAMALT biopsy specimen-negative, RLN positive animals, 136 (82.4%) had <5 follicles observed. Follicles were absent from 11 of these deer. From these results, the sensitivity of RAMALT IHC analysis was 44.4%, while its specificity was 100%. As in herd 1, a range of PRNP alleles were represented, including the wild type and 96G/S, 95Q/H, and 226Q/K variants (Tables 1 and 2).

TABLE 2.

Summary of PRNP genotype, sex, CWD status, obex score, and RT-QuIC results of the two CWD-positive herds^a

Herd and allele	No. negative		No. positive		Total	No. IHC positive with:			No. with obex score of:					No. RT-QuIC positive with:
	M	F	M	F		RLN	Obex	RAMALT	0	1	2	3	4	RAMALT	NB
1
QGAQ/QGAQ	1	1	2	3	7	5	3	3	2	0	2	1	0	3	0
QGAQ/QSAQ	0	5	0	0	5	0	0	0	0	0	0	0	0	0	0
QSAQ/HGAQ	0	1	0	0	1	0	0	0	0	0	0	0	0	0	0
QGAQ/QGAK	0	1	0	0	1	0	0	0	0	0	0	0	0	0	0
2b
QGAQ/QGAQ	5	2	62	61	130	123	113	94	10	5	12	82	14	118	28
QGAQ/QSAQ	19	18	68	64	174	132	86	31	46	24	37	21	4	68	10
QSAQ/QSAQ	9	14	13	6	42	19	0	0	19	0	0	0	0	6	0
QGAQ/HGAQ	0	3	2	2	7	5	1	0	4	1	0	0	0	0	0
QGAQ/QGAK	0	0	0	2	2	2	0	0	2	0	0	0	0	0	0
QSAQ/QGAK	1	0	0	0	1	0	0	0	0	0	0	0	0	0	0

^aM, male; F, female; NB, nasal brushings.

^bThe PRNP genotypes of four animals in herd 2 were not available.

In herd 3, the negative-control herd, all RAMALT biopsy specimens were negative by IHC analysis. Animals were examined postmortem when they expired of natural causes, and the herd retains its CWD-negative status.

RT-QuIC analysis of RAMALT biopsy specimens.

In the two positive herds, RT-QuIC analysis of RAMALT biopsy specimens identified 3/5 positives from herd 1 (each also RAMALT IHC analysis positive, 60%) and 193/284 positive animals from herd 2 (68%). Specificity was 100% and 93.9% in herds 1 and 2, respectively. In CWD-negative herd 3, RT-QuIC analyses of RAMALT from all of the animals were considered negative.

RT-QuIC analysis of nasal brushings.

In the two positive herds, RT-QuIC analysis of nasal brushings identified 0/5 animals from herd 1 and only 44/284 positive animals from herd 2 (15.5%). The specificities for the respective herds were 100% and 90.2%. Nasal brushings collected from herd 3 were all negative by RT-QuIC analysis.

Correlations of RT-QuIC results with RAMALT, RLN, and obex IHC analysis; obex scores; and PrP genotype.

There was a positive correlation (62%) between RT-QuIC results obtained with RAMALT and IHC analysis of RAMALT. Under the specific assay conditions used, with approximately equal input sample sizes, RT-QuIC was a significant improvement over IHC analysis when evaluating RAMALT biopsy specimens (P < 0.001, Table 3). RT-QuIC results obtained with RAMALT were positively correlated with RLN and obex IHC analysis at 55% and 72%, respectively. RT-QuIC results obtained with RAMALT were negatively correlated with an obex score of 0 (−72%), but positive correlations were found with obex scores of 3 (57%) and 4 (21%). RT-QuIC results obtained with RAMALT were not correlated with PRNP positions 95 and 226, were positively correlated with PRNP position 96 for homozygous glycine, (47%), and were negatively correlated with heterozygous glycine-serine (−28%) and homozygous serine (−28%). RT-QuIC results obtained with nasal brushings were correlated with obex IHC analysis at 21% but not correlated with RLN analysis at 9%. Specifically, RT-QuIC results obtained with nasal brushings were negatively correlated (−21%) with an obex score of 0 but positively correlated (51%) with an obex score of 4 (Table 3; Fig. 2).

TABLE 3.

Spearmen correlations of herd variables^a

Correlation	Obex	RLN	RAMALT		NB RT-QuIC	GG96	GS96	SS96	Obex score of:
			IHC analysis	RT-QuIC					0	1	2	3	4
Positiveb	0.59 (<0.001)	1.0 (<0.001)	0.38 (<0.001)	0.55 (<0.001)	0.09 (0.08)	0.26 (<0.001)	−0.07 (0.16)	−0.29 (<0.001)	−0.59 (<0.001)	0.16 (<0.001)	0.21 (<0.001)	0.33 (<0.001)	0.12 (0.04)
RAMALT IHC analysis	0.57 (<0.001)	0.38 (<0.001)	1.0 (<0.001)	0.62 (<0.001)	0.25 (<0.001)	0.52 (<0.001)	−0.36 (<0.001)	−0.25 (<0.001)	−0.57 (<0.001)	−0.07 (1.0)	0.00 (1.0)	0.55 (<0.001)	0.26 (<0.001)
GG96	0.39 (<0.001)	0.26 (<0.001)	0.52 (<0.001)	0.47 (<0.001)	0.18 (<0.001)	1.0 (<0.001)	−0.79 (<0.001)	−0.30 (<0.001)	−0.39 (<0.001)	−0.12 (1.0)	−0.11 (1.0)	0.51 (<0.001)	0.17 (0.09)
GS96	−0.13 (0.01)	−0.07 (0.16)	−0.36 (<0.001)	−0.28 (<0.001)	−0.11 (0.03)	−0.79 (<0.001)	1.0 (<0.001)	−0.35 (<0.001)	0.13 (1.0)	0.19 (0.03)	0.20 (0.02)	−0.35 (<0.001)	−0.12 (1.0)
SS96	−0.40 (<0.001)	−0.29 (<0.001)	−0.25 (<0.001)	−0.28 (<0.001)	−0.10 (0.05)	−0.30 (<0.001)	−0.35 (<0.001)	1.0 (<0.001)	0.40 (<0.001)	−0.11 (1.0)	−0.14 (0.53)	−0.23 (<0.001)	−0.08 (1.0)

^aIncluding CWD status (as identified through brainstem and RLN IHC analysis), RAMALT IHC analysis, and the PRNP 96 codon. NB, nasal brushings.

^bApplies to tissue correlated with either an obex or an RLN sample that was positive for CWD by IHC analysis.

FIG 2.

Association of antemortem testing results with obex scores. As obex scores increased, the sensitivity of both nasal brush and rectal biopsy specimen (RAMALT) testing improved. A positive association was also observed between the obex score and the 96G allele. NB-R, nasal brush testing by RT-QuIC assay; RB-I, RAMALT testing by IHC analysis; RBR, RAMALT testing by RT-QuIC assay; 96G, frequency of the 96G allele.

Obex results were not correlated with PRNP positions 95 and 226 and were negatively correlated with PRNP position 96 for homozygous glycine at obex scores of 0 and 2 at −39% and −11%, respectively. Obex results were positively correlated with 96G/G for scores of 3 and 4 at 51% and 17%, respectively. Obex scores were positively correlated with 96G/S at obex scores of 0, 1, and 2 at 13%,19%, and 20%, respectively, but negatively correlated at obex scores of 3 and 4 at −35% and −12%, respectively. An obex score of 0 was positively correlated with 96S/S (40%) but negatively correlated with obex scores of 2 and 3 at −14%, respectively, but not correlated with an obex score of 4 (Fig. 3).

FIG 3.

Association of test results with the PRNP 96G/S alleles. Antemortem test results were strongly correlated with the 96G allele, with greater sensitivity of both nasal brush and rectal biopsy specimen (RAMALT) analyses by RT-QuIC assay. Postmortem diagnosis of CWD also highly correlated with the 96G allele. NB-R, nasal brush testing via RT-QuIC assay; RB-I, RAMALT testing by IHC analysis; RB-R, RAMALT testing by RT-QuIC assay; PM, postmortem CWD status via brainstem and medial RLN testing by IHC analysis.

DISCUSSION

As novel CWD foci arise across North America, the development of rapid postmortem—and especially antemortem—tests with both high sensitivity and specificity becomes critical to preventing its incursion into new cervid populations. IHC analysis of RAMALT has been under evaluation as a potential diagnostic approach for nearly a decade, and much has been learned about the shortcomings in sensitivity with regard to both the genotype and the clinical stage of the disease. In contrast, RT-QuIC analysis of nasal brush collections from both humans and elk, with mixed success, has only recently been reported (37, 38).

In the present study, we sought to initially compare an amplification-based assay on antemortem RAMALT biopsy specimens to IHC analysis and ultimately to standard postmortem analysis. In our analysis of >400 white-tailed deer, we found this assay to have high sensitivity (∼69%), an apparent improvement over RAMALT IHC analysis. As in previous reports, both the obex score and the PRNP genotype were positive predictors of antemortem test results. The apparent shortfall in sensitivity found with RAMALT IHC analysis relative to RT-QuIC was likely dependent on the size of the biopsy specimens collected (∼0.25 cm² was used for each assay) and the resultant low number of follicles observed per biopsy specimen. Though no standard diagnostic sample exists for IHC analysis screening of RAMALT biopsy specimens for CWD infection, historically, samples of ≥1.5 cm² with more than five follicles have been considered ideal (39). Samples meeting these requirements have yielded sensitivities ranging from 68 to 91% under various circumstances (13, 15, 17, 39, 40)—which are very similar to our estimated RT-QuIC sensitivity in the current study and similar even to recognized tests for other regulated diseases of cervids, e.g., tuberculosis (41). Our previous work found that when sample sizes were ideal for each assay, the sensitivity of RT-QuIC was comparable to that of IHC analysis, and it seems reasonable to assume that this would have also been true in the present study. The consistency reported in RAMALT biopsy specimen sensitivity seems to imply that 70 to 90% may be the theoretical upper limit in large-scale studies employing this testing strategy. As a result, further studies are needed to detail the pathogenesis of CWD with regard to the timing of the appearance of CWD prions in peripheral central (e.g., RAMALT and nasal epithelium versus proximal intestinal tissues) and central tissues (e.g., RLN versus the brainstem), which would assist efforts to identify additional accessible tissues or encourage alternate approaches for sampling central lymphoreticular tissues to identify CWD-positive animals in earlier stages of disease, especially prior to the onset of shedding.

As with our previous report, RT-QuIC evaluations of nasal brush collections were much less sensitive than those of RAMALT biopsy specimens, though positive results again correlated well with both the susceptible (96GG) genotype and higher obex scores. With advancements in sample preparation techniques and assay design, it is possible that the sensitivity of nasal brush collections may improve in the future, though our present findings again seem to indicate that screening of human patients at risk for Creutzfeldt-Jakob disease or other naturally occurring or iatrogenic prion diseases may suffer from reduced sensitivity in preclinical cases.

The specificity of the RT-QuIC assay in the present study (∼90 to 94%) was lower than anticipated in both RAMALT biopsy specimens and nasal brush collections and lower than that in our previous analyses of elk. As a result, we may conclude that while RT-QuIC offers numerous advantages over IHC analysis, including rapidity of sample processing and analysis and real-time data acquisition akin to those of real-time PCR, it is prone to the same obstacles—including the potential for cross-contamination during sample collection, processing, and analysis. While great care was taken to prevent cross-contamination, the overwhelming number and prevalence of CWD-positive animals (compared to our previous study) may have contributed to the relatively low specificity in the present study via environmental (e.g., dust [42]) or laboratory contamination. Future studies with RT-QuIC should take this into consideration and develop not only a priori sample-handling strategies but analytical strategies as well to address the potential for false-positive amplification through spontaneous seeding or cross-contamination.

Because of the consistent linkage between antemortem test results and both the genotype and the stage of clinical disease, two notable questions arise. (i) Can a test be developed that is highly sensitive for CWD—and other prion diseases—regardless of the stage of disease and the PRNP background? (ii) Short of reaching this goal, is there a testing strategy that can be used with assays that represent the state of the field to date? The answer to the first question is likely yes, though that test, potentially involving blood analyses (43, 44) and/or continued modification and development of the RT-QuIC assay, may yet be some time off. In that case, the answer to the second question lies in understanding the limitations of biopsy specimen testing and potentially requiring repeated testing in combination with genetic screening to allow higher confidence levels in a CWD-negative result.

This promise should lead to a discussion of the role of genetics in CWD resistance in the farmed cervid industry. If genetic testing to identify the PRNP alleles present in a population may be used to estimate our confidence in an antemortem test result, perhaps they could be used more proactively to foster resistance in farmed or otherwise managed herds—an approach that has already been used with success in sheep herds affected by scrapie (45–47). The caveat here is that, unlike with scrapie, a white-tailed deer PRNP allele has not yet been identified that completely prevents infection with the agent of CWD—though a number of putatively resistant alleles have been identified, including the 95H, 96S, 116G, and 226K alleles reevaluated here. Apart from the well-characterized resistance imparted by the 96S allele, which manifests itself, at minimum, as a delay in disease progression, little is known about the level of resistance that may be offered by the other alleles, especially in homozygous form, because the alleles themselves are quite rare in the population (<5%) (48). For this reason, further studies on CWD resistance in cervids are necessary not only to develop a more acute understanding of what constitutes resistance to prion disease but also to help promote breeding schemes for resistant animals that could preclude the necessity of an antemortem test altogether.

In summary, we present further data supporting the practicality of RAMALT biopsy specimens for the antemortem screening of cervids for CWD by RT-QuIC assay. Compared to nasal brush sampling, RAMALT testing by RT-QuIC assay had a more pronounced sensitivity in the detection of true infection, though the same pitfalls affecting nucleic acid amplification could be at play with abnormally folded protein amplification, given our unexpected, though minor, loss of specificity. Our findings should prove useful in the continued development and wide-scale application of antemortem tests for prion diseases and may be combined with genetic testing to provide greater confidence in assay results.