Comparative analysis of prions in nervous and lymphoid tissues of chronic wasting disease-infected cervids

Abstract

The prevalence, host range and geographical bounds of chronic wasting disease (CWD), the prion disease of cervids, are expanding. Horizontal transmission likely contributes the majority of new CWD cases, but the mechanism by which prions are transmitted among CWD-affected cervids remains unclear. To address the extent to which prion amplification in peripheral tissues contributes to contagious transmission, we assessed the prion levels in central nervous and lymphoreticular system tissues in white-tailed deer (Odocoileus virginianus), red deer (Cervus elaphus elaphus) and elk (Cervus canadensis). Using real-time quaking-induced conversion, cervid prion cell assay and transgenic mouse bioassay, we found that the retropharyngeal lymph nodes of red deer, white-tailed deer and elk contained similar prion titres to brain from the same individuals. We propose that marked lymphotropism is essential for the horizontal transmission of prion diseases and postulate that shed CWD prions are produced in the periphery.

Transmissible spongiform encephalopathies (TSEs) are fatal, incurable, infectious disorders of mammals that are caused by prions. Zoonotic transmission of bovine spongiform encephalopathy (BSE) caused variant Creutzfeldt–Jakob disease (vCJD) in young adults and teenagers [1]. vCJD not only illustrates the ability of TSEs to occur as unpredictable epidemics, but also raises concerns about the risk from additional animal prion diseases, including sheep scrapie and cervid chronic wasting disease (CWD). TSE prion replication results from corruption of the host-encoded cellular prion protein (PrP^C) by its infective counterpart (PrP^Sc) through conformational templating [2]. The contagious transmission of CWD in free-ranging and captive animals is unique among prion diseases. Prevalence can reach 13 % in certain wild populations [3], while >80 % of captive deer housed in paddocks that previously contained CWD-infected animals ultimately developed disease [4]. By contrast, bovine spongiform encephalopathy incidence in dairy cattle herds was less than 3 % [5]. While the facile spread of CWD in wild populations is likely a result of both direct animal-to-animal contact and exposure to contaminated environments, the biological underpinnings of this process are not completely understood, and it is not entirely clear why horizontal transmissibility varies among TSEs.

Early, widespread and progressive accumulation of PrP^Sc in the lymphoid system is a feature of CWD in deer, and PrP^Sc or prion infectivity has been detected in many additional peripheral tissues [6–18]. While similar patterns of early lymphoid prion deposition and eventual widespread prion deposition are well established in sheep scrapie [19–24], PrP^Sc is restricted almost entirely to nervous tissue in BSE-affected cattle [25–28].

Despite significant involvement of the lymphoreticular system, studies of CWD infectivity have focused primarily on prions in the central nervous system (CNS). We therefore compared the properties of prions in retropharyngeal lymph nodes with prions in the CNS of the same individual for three cervid subspecies: red deer (Cervus elaphus elaphus) inoculated with CWD-affected brain homogenate, allowed to progress until terminal disease [16]; naturally infected elk (Cervus canadensis); and naturally or experimentallyinfected white-tailed deer (WTD) (Odocoileus virginianus) [29]. We used complementary approaches, including real-time quaking-induced conversion (RT-QuIC) [30], cervid prion cell assay (CPCA) [31] and bioassay in CWD-susceptible transgenic (Tg) mice [32]. In RT-QuIC with truncated Syrian hamster recombinant PrP (the standard RT-QuIC substrate [33]), the rate of amyloid formation (QuIC rate) is the inverse of the lag phase observed between the addition of a prion-containing seed and the detection of amyloid by thioflavin-T fluorescence. The rate of amyloid formation for prions from the retropharyngeal lymph node (RPLN) and the CNS were indistinguishable in natural CWD infections of WTD and elk (Fig. 1). Because RT-QuIC relies on the seeded conversion of recombinant PrP to an amyloid form, and the correlation of seeding activity to infectivity is uncertain, we also estimated the infectivity of each tissue using the CPCA. Accordingly, we infected transgenically modified RK13 rabbit kidney epithelial cells expressing deer or elk PrP^C with WTD or elk prions of matching genotype. We calculated titres as previously described; briefly, the titre is the log of the number of spots (which is estimated to be the number of infected cells) per gram of tissue [31, 34, 35]. RK13 cells that do not express PrP^C served as negative controls. The titres of CWD prions from RPLN and brain of elk and red deer were similar (elk RPLN: 8.5; elk brain: 7.9; red deer RPLN: 6.7; red deer brain: 7.5), but the titres were lower in the RPLN compared to the CNS for experimentally infected WTD 816 and 817 (816 RPLN: 6.3; 816 brain: 7.0; 817 RPLN: 5.1; 817 brain: 6.9) (Fig. 1). Interestingly, preparations from naturally infected WTD brain failed to infect susceptible RK13 cells (Fig. 1b–c), although we detected proteinase K (PK)-resistant PrP^Sc by Western blot (data not shown) and amyloid seeding activity (Fig. 1a) in RT-QuIC. In contrast, samples from experimentally infected WTD (#816, 817) produced robust infections of deer PrP^C expressing cells. Therefore, prions were present in the naturally infected WTD brain, and the cells expressing deer PrP^C were competent for infection. Previous studies have suggested that CPCA may be sensitive to CWD prion strain properties, since other infectious CWD isolates do not efficiently infect cells in the CPCA [31]. The same CWD isolates (from naturally and experimentally infected WTD) had similar seeding activity in RT-QuIC; it seems that their variability was only detectable in CPCA. Due to the sensitivity of CPCA to CWD strain properties, we hypothesize that naturally infected deer brain contained a different strain than the other WTD samples.

Fig. 1.

Prion loads in RPLN and CNS from CWD-infected cervids are comparable when assessed by RT-QuIC and CPCA. (a) RPLN and brain samples from red deer, WTD or elk were tested in RT-QuIC to determine the seeding activity of each sample (represented by the RT-QuIC rate, which is the inverse of the lag phase). Filled circles, brain; open circles, RPLN. Error bars, SEM at each dilution. *, significant differences between RPLN and brain samples (P<0.05, two-sided Mann–Whitney test). Dotted grey lines, mean spontaneous conversion rate for a negative brain sample; solid grey lines, mean spontaneous conversion rate for a negative RPLN sample. (b) CPCA of RPLN or brain from red deer, WTD or elk using RK13 cells expressing deer PrP^C (WTD samples), or elk PrP^C (elk and red deer samples). Infected cells appear as immunopositive spots on elispot plates following treatment with proteinase K. (c) Titres indicate the log spots g⁻¹ tissue calculated in the linear range of curves, with a limit of detection of 10³ (indicated by the grey region).

To assess the infectivity of prions from the lymph nodes and CNS of the various CWD-infected cervid subspecies, we intracerebrally inoculated Tg(DeerPrP) 1536 mice (as previously described, [32]). All of the inoculated animals in all six study groups displayed a constellation of neurological signs consistent with CWD, with mean incubation times ranging from 270 to 370 days (Fig. 2a). Mice inoculated with brain extracts from naturally infected WTD had shorter mean incubation times than mice inoculated with RPLN from the same animal (P<0.05, Mantel–Cox) (Fig. 2a). While CWD prions from the CNS of naturally infected elk produced disease with a shorter mean incubation time than prions from RPLN (P<0.005, Mantel–Cox) (Fig. 2a), Tg mice inoculated with red deer brain extracts had a longer mean incubation time than Tg mice inoculated with red deer RPLN (P<0.005, Mantel–Cox) (Fig. 2a). Immunoblots of brain extracts showing the presence of PrP^Sc in the brains of diseased Tg(DeerPrP) 1536 mice confirmed that mice had succumbed to prion infection in all cases (Fig. 2b). We analysed coronal hippocampal sections with immunohistochemistry using formic acid treatment, antigen retrieval and antibody D18 for standard neuropathological assessment of CWD strains in Tg(Deer) 1536 mice [36]. Mice inoculated with brain or RPLN from red deer had similar PrP^Sc deposition patterns: florid plaques were present primarily in the inoculated hemisphere and the hippocampi were spared, which is reminiscent of type II CWD [37] (Fig. 3a, b). Mice inoculated with WTD RPLN had a similar deposition pattern to red deer-inoculated mice, with florid plaques in the inoculated (right) hemisphere, which is characteristic of type II CWD (Fig. 3e). However, mice inoculated with WTD brain had florid plaques in both hemispheres and hippocampal PrP^Sc accumulation, which is reminiscent of type I CWD (Fig. 3d).

Fig. 2.

Prions from the CNS and lymphoreticular system of CWD-infected deer and elk are transmissible to CWD-susceptible transgenic mice. (a) Tg(DeerPrP) 1536 mice were intracerebrally inoculated with 1 % homogenates of brain and RPLN tissues from CWD-infected cervids. Kaplan–Meier survival curves are shown for each infected cohort. Mean incubation times (±SEM), where n indicates the number of infected and diseased animals, are also shown. Asterisks indicate statistical significance between the mean incubation periods of matched samples (*, P<0.05; **, P<0.005, Mantel–Cox). (b). All of the diseased Tg(DeerPrP) 1536 mouse brains were analysed by immunoblotting to confirm prion disease. Samples from two representative mice in each group are shown. The inoculum tissues are listed above the blots, and the inoculum species and PK treatments are listed below the blots.

Fig. 3.

The PrP^Sc deposition patterns in mice varied with inoculum source. We prepared coronal sections through the hippocampus of Tg(DeerPrP) 1536 mice inoculated with: (a) red deer brain, (b) red deer RPLN, (c) nothing (uninoculated), (d) white-tailed deer brain, (e) white-tailed deer RPLN, (g) elk brain, (h) elk RPLN, 320 days p.i., (i) elk RPLN, 400 days p.i. We used prion knockout mice (PrnP^−/−) as a negative control for IHC (f). We stained paraffin-embedded sections with anti-PrP antibody 18 days after formic acid treatment and antigen retrieval. Scale bar (h), 500 µm. All mice were inoculated in the right hemisphere.

Mice inoculated with elk RPLN had an apparently bimodal distribution of incubation times (Fig. 2a). Mice with shorter incubation periods (320 days p.i.) had PrP^Sc in both hemispheres and the hippocampus (type I) (Fig. 3h), as did all mice inoculated with elk brain (Fig. 3g). Interestingly, mice inoculated with elk RPLN with longer incubation periods (400 days p.i.) displayed asymmetric deposition patterns (type II) (Fig. 3j). The variability in PrP^Sc deposition patterns between animals inoculated with brain versus lymphoid tissue and within the group inoculated with elk RPLN suggests the presence of multiple CWD strains in a single individual.

Understanding the biological properties of prions that are resident in non-CNS tissues of animals affected by CWD is important for several reasons. First, it offers clues about the rapid and contagious spread of CWD [38, 39]. While the pathogenesis of both CWD and sheep scrapie are characterized by prion replication in the lymphoid system [6, 7, 40], BSE is restricted almost exclusively to the nervous system, with a few reports of PrP^Sc detection in the distal ileum and the tonsil [25–27]. However, when cattle were intracerebrally inoculated with scrapie or CWD, there was essentially no PrPSc detected outside the CNS [41, 42]. In contrast, when sheep were inoculated with BSE, prions were detectable in the lymphoid system as well as the CNS [43, 44]. Taken together, these results suggest a significant host influence on prion lymphotropism, with deer and sheep tending to permit lymphotropism, while cattle do not. Additionally, the lymphoid system and CNS have different tolerances for cross-species prion transmission [45]. When Tg mice were inoculated with prion strains to which they are typically resistant, PrP^Sc propagation and prion infectivity were detectable in their spleens, despite an absence of prion replication in the brain [45]. CWD strains seem to have variable lymphotropic behaviour, which supports the hypothesis that the CNS and lymphoid system provide different environments for prion replication [37]. We propose that lymphotropism is an essential component of horizontally transmissible prion diseases, which are characterized by substantial titres of infectious prions outside the brain, particularly in the lymphoid system. Full characterization of the properties of peripheral CWD prions versus CNS prions and an assessment of their zoonotic potential remains to be determined. Our data suggest that (1) peripheral tissues accumulate high titres of infectious CWD prions and (2) different CWD strains may be harboured by peripheral tissues versus central nervous tissues.