J. Comp. Path. 2014, Vol. 150, 57e70

Available online at www.sciencedirect.com

ScienceDirect www.elsevier.com/locate/jcpa

EXPERIMENTALLY INDUCED DISEASE

Influence of Polymorphisms in the Prion Protein Gene on the Pathogenesis and Neuropathological Phenotype of Sheep Scrapie after Oral Infection L. Gonz alez*, J. L. Pitarch†, S. Martin*, L. Thurston‡, H. Simmons‡, C. Acın† and M. Jeffrey* * Animal Health and Veterinary Laboratories Agency, Penicuik, Midlothian EH26 0PZ, UK, † Facultad de Veterinaria, Universidad de Zaragoza, 50013 Zaragoza, Spain and ‡ Animal Health and Veterinary Laboratories Agency, Addlestone, Surrey KT15 3NB, UK

Summary The prion protein gene (Prnp) plays a crucial role in the susceptibility of sheep to scrapie in terms of attack rate and/or incubation period. However, the influence of Prnp on the pathogenesis of the disease, specifically the involvement of tissues of the lymphoreticular system (LRS), pathways of neuroinvasion and neuropathological phenotypes, remains controversial. This study reports the onset and progression of disease-associated prion protein (PrPd) accumulation in the LRS and nervous tissues of sheep of six different Prnp genotypes infected by oral administration of the same mixed scrapie brain homogenate. Sheep homozygous for glutamine (Q) at codon 171 of PrP, with either valine (V) or alanine (A) at codon 136 (i.e. VRQ/VRQ, VRQ/ARQ and ARQ/ARQ), showed early and consistent PrPd accumulation in LRS tissues of the pharynx and gut. In contrast, LRS involvement was minimal, inconsistent and occurred late in the incubation period in sheep heterozygous for arginine (R) at codon 171 (i.e. VRQ/ARR and ARQ/ARR). Despite this difference, all five groups were susceptible to infection and developed clinical disease, albeit with significantly different incubation periods (shortest in VRQ/VRQ and longest in ARQ/ARR sheep). The remaining group of ARR/ARR homozygous sheep did not show evidence of infection at the end of the experiment or at previous predetermined time points. As for LRS tissues, the sites of initial PrPd accumulation in the brain were determined immunohistochemically. These were the same in all susceptible sheep (except for ARR/ARR sheep), regardless of their Prnp genotype which, together with an early and consistent accumulation of PrPd in circumventricular organs and a late or inconsistent involvement of the enteric and autonomic nervous system and of the spinal cord, suggests neuroinvasion occurring via the blood. The neuropathological phenotype (PrPd profile in the central nervous system) of clinically affected sheep was similar in the three V136 carrier groups, but showed some differences in the two A136 homozygous groups, suggesting a codon 136-driven selection of different strains from the mixture contained in the inoculum. ARQ/ARR sheep showed an irregular distribution of brain PrPd, contrasting with the more widespread distribution of the other four groups. The results indicate that (1) ARQ/ARR sheep are more susceptible to oral scrapie infection than would be predicted from incidence figures in natural disease, (2) amplification and accumulation of PrPd in LRS tissues is host genotype dependent, but does not necessarily have a marked effect on the outcome of the infection and (3) the neuropathological phenotype of scrapie is related to the host genotype, but possibly in combination with the infecting source. Crown Copyright Ó 2014 Published by Elsevier Ltd. All rights reserved. Keywords: prion protein; scrapie; sheep; transmissible spongiform encephalopathy

Correspondence to: L. Gonzalez (e-mail: [email protected]). 0021-9975/$ - see front matter http://dx.doi.org/10.1016/j.jcpa.2013.10.001

Crown Copyright Ó 2014 Published by Elsevier Ltd. All rights reserved.

L. Gonz alez et al.

58

Introduction Scrapie is a progressive and invariably fatal transmissible spongiform encephalopathy (TSE) originally described in sheep more than two centuries ago (reviewed by Schneider et al., 2008). It occurs as a natural infection in many countries and can also be transmitted by a variety of routes (refer to preceding article in this issue) including orally, a route that most closely mimics natural transmission. The relationship between sheep host genetics and susceptibility to scrapie is well established and it was speculated that scrapie might be a genetic condition (Hunter et al., 1997). However, it is now recognized that scrapie is an infectious and contagious disease, almost unanimously believed to be caused by a proteinaceous infectious particle or prion (Prusiner, 1982), the susceptibility to which is strongly modulated by polymorphisms in the prion protein gene (Prnp). The ancestral Prnp gene in sheep encodes for alanine (A), arginine (R) and glutamine (Q) at codons 136, 154 and 171 of the open reading frame, respectively, and mutations at these codons are mainly responsible for increased resistance or susceptibility to sheep scrapie, measured in terms of attack rate or incubation period (reviewed by Goldmann, 2008). Thus, valine (V) at codon 136 is associated with increased susceptibility, while R at codon 171 is associated with increased resistance, making VRQ/VRQ sheep the most susceptible and ARR/ ARR sheep the most resistant to many classical scrapie sources. The relationship between scrapie and Prnp genotype constitutes the basis for the implementation of a scrapie control programme in sheep based on genetic selection (Dawson et al., 2008). More recently, polymorphisms at other codons, in particular substitution of methionine (M) by threonine (T) or of leucine (L) by phenylalanine (F) at codons 112 and 141, respectively, have been shown to provide increased resistance to scrapie (Laegreid et al., 2008; Gonzalez et al., 2012b). The influence of the sheep Prnp genotype on the pathogenesis of scrapie has not been studied as systematically as its impact on susceptibility. It is generally accepted that in most sheep of susceptible Prnp genotypes, and as judged by the detection of disease-associated prion protein (PrPd), the infectious agent initially replicates in the tissues of the lymphoreticular system (LRS) prior to entering the brain. However, some sheep, particularly those of VRQ/ ARR genotype, can show accumulation of PrPd in the brain without apparent LRS involvement (van Keulen et al., 1996, 2002; Andreoletti et al., 2000; Jeffrey et al., 2002; Ersdal et al., 2005). In natural sheep scrapie, VRQ/VRQ sheep have been found to

show the earliest and most consistent LRS involvement followed by VRQ/ARQ and ARQ/ ARQ sheep (reviewed by Jeffrey and Gonz alez, 2007). Sheep that are homozygous or heterozygous for the AHQ and ARH alleles have also shown involvement of LRS tissues (Gonzalez et al., 2006), but the rarity of these alleles in the sheep population makes numbers of scrapie cases studied too low and restricted to clinical disease to allow further interpretation. In summary, there is no firm evidence at present to confirm whether the LRS plays a decisive role in the pathogenesis of sheep scrapie by facilitating or speeding up the process of neuroinvasion. Differences in Prnp genetics in sheep with susceptible genotypes have also been implicated as drivers or selectors of the disease phenotype, particularly the morphology and distribution of PrPd aggregates in the brain, as judged by immunohistochemistry (IHC) (Ligios et al., 2004; Spiropoulos et al., 2007). Other studies, however, suggested that variability in immunohistochemically-determined phenotypes of PrPd accumulation in the brain is mostly due to infecting strain diversity (Gonzalez et al., 2002, 2005, 2010b; Jeffrey et al., 2003, 2006). A recently published experiment on cross-Prnp genotype scrapie transmissions strongly suggests that it is the interaction between the infecting agent or strain and the host genotype that dictates the disease phenotype (Gonzalez et al., 2012b). The aim of the present study was to determine the effect of the Prnp genotype on the pathogenesis of sheep scrapie, including the role of the LRS and the possible pathways of neuroinvasion, and on the disease phenotype, including its incubation period and neuropathological features.

Materials and Methods Experimental Design

One hundred and twelve Cheviot lambs of five different Prnp genotypes were dosed orally with a brain pool inoculum. Details of the inoculum and of the inoculation procedure are described in the preceding article of this issue and in Ryder et al. (2009), respectively. The experimental groups (Prnp genotypes indicate amino acid positions at codons 136, 154 and 171 of PrP) are described in Table 1. VRQ/ARQ sheep included twenty-six 10- to 15day-old lambs dosed with 1 g of inoculum. Twentyone of these animals were killed in the absence of clinical signs of scrapie in groups of three at 15, 29, 43, 85, 112, 177 and 203 days post inoculation (dpi) and five sheep were killed once they developed definite neurological signs compatible with scrapie.

Influence of Prnp Genotype on Scrapie

59

Table 1 Onset and progression of PrPd accumulation in tissues of sheep of different genotypes orally dosed with the same pooled scrapie inoculum.

AR, attack rates calculated considering only sheep killed after the first positive. Results expressed as positive sheep/sheep examined at the different time points (days post-infection, dpi) in the different tissue compartments. In bold, earliest indication of involvement of each compartment. Note that results at 15 dpi for the first four genotype groups are omitted since all sheep were negative at that time point; for the same reason results from ARQ/ARR at 373, 736 and 905 dpi and the whole ARR/ARR group are also excluded. Note also that, for the first four groups, columns are aligned to provide a better visual comparison of days post-infection; this alignment does not apply to the ARQ/ARR group. LRS, lymphoreticular system; ENS, enteric nervous system; SPC, spinal cord; ANS, autonomic nervous system (*results are displayed only for sheep with nodose or cranial mesenteric ganglia available for examination). For details of specific tissues or tissue samples included in each group refer to materials and methods in the text of the preceding article in this issue or to legend of Fig. 1. For details of individual tissue results refer to Supplementary Table 1. †, the positive tissues belonged to the same sheep; ‡, the positive tissues belonged to two different sheep in each time point. Different letters at the right of each group indicate that the survival times of the sheep left to develop clinical disease (last column) are significantly different in the Mann-Whitney test (P < 0.01).

60

L. Gonz alez et al.

ARQ/ARQ sheep included twenty-six 10- to 14day-old lambs dosed with 1 g of inoculum. Twentyone of these animals were killed in groups of three at 15, 29, 43, 85, 163, 220 and 340 dpi and the remaining five were killed once they developed clinical disease. VRQ/ARR sheep comprised eighteen 13- to 16day-old lambs dosed with 1 g of inoculum; of these, four groups of three animals were killed at 15, 29, 43 and 163 dpi and six sheep were killed after showing definite clinical signs. ARQ/ARR sheep included twenty-six 10- to 15day-old lambs dosed with 5 g of inoculum (1 g for five consecutive days) and killed in groups of three at 373, 736, 905, 1,086, 1,290, 1,444 and 1,837 dpi. Some sheep had post-mortem evidence of infection at the four later time points, so a group of five sheep were left to develop clinical disease and killed at that time. ARR/ARR sheep comprised twenty-six 11- to 22day-old lambs dosed with 5 g of inoculum (1 g for five consecutive days) and killed in groups of three at 414, 780, 1,129, 1,500, 1,885 and 2,584 dpi. The remaining eight animals were killed at the end of the experiment (2,685e2,760 dpi). For a more complete comparison between different Prnp genotypes, this report also includes a further group of sheep from the preceding article in this issue. These are the 27 VRQ/VRQ Cheviot sheep dosed orally with 1 g of the same pooled inoculum and killed or left to develop clinical disease, as shown in Table 1. In this report, VRQ/ARR and ARQ/ARR will be sometimes referred to as R171 carriers, as opposed to Q171 homozygotes (VRQ/VRQ, VRQ/ARQ and ARQ/ARQ) and VRQ/VRQ, VRQ/ARQ and VRQ/ARR as V136 carriers as opposed to A136 homozygotes (ARQ/ARQ and ARQ/ARR).

Tissue Sampling and Analysis

All procedures concerning tissue sampling at postmortem, tissue grouping (LRS1 and LRS2, brain, spinal cord, enteric nervous system [ENS] and autonomic nervous system [ANS]), immunohistochemical procedure for PrPd detection, assessment of PrPd distribution in the neuroparenchyma and circumventricular organs (CVOs) of sheep culled at preclinical stages of infection and immunophenotyping of PrPd in the CNS of clinically affected sheep are identical to those reported in the preceding article of this issue with two exceptions: (1) the systematic examination of the third eyelid, which was included in the LRS2 group (see details in Supplementary Table 1) and (2) the inclusion of the reticular formation at the level of the obex as an additional structure in the assess-

ment of magnitude and morphological types of PrPd in clinical scrapie cases. In addition, in order to determine whether or not R171 carriers accumulated PrPd of the ARR polymorphism, serial sections of selected tissues from such animals were subjected to IHC with R145 or 2A11 monoclonal antibodies. 2A11 is a mouse monoclonal antibody binding amino acid residues 163e171 of ovine PrP (Brun et al., 2004), which does not recognize R at codon 171 and, therefore, does not produce positive immunolabelling in bovine spongiform encephalopathy (BSE)-infected ARR/ARR sheep tissues, while other antibodies such as R145 do (Bilheude et al., 2007). Prnp Genotyping

At the start of the experiment, amino acids at codons 136, 154 and 171 of ovine PrP were determined by sequencing with an ABI Prism 377 DNA sequencer according to the manufacturer’s instructions (PE Applied Biosystems, Warrington, UK). In 2013, in order to ascertain the existence of polymorphisms at other codons of ARQ allele-bearing sheep, archival blood samples from all 78 such sheep (26 each of ARQ/ARQ, VRQ/ARQ and ARQ/ARR genotype) were taken for PCR amplification and sequencing of the whole open reading frame of the Prnp gene on a 3130 Genetic Analyzer with the BigDyeÒ terminator v3.1 cycle sequencing kit as per the manufacturer’s protocol (PE Applied Biosystems).

Results Attack Rates and Onset and Progression of PrPd Accumulation

None of the 26 ARR/ARR sheep showed PrPd accumulation in any of the tissues examined (a zero attack rate) and are therefore excluded from the tables and figures in this report. For the other genotype groups, if only animals examined later than the first indication of infection in each group are considered in the calculations, attack rates were 64% for ARQ/ARR sheep and 100% for the remaining four genotypes (Table 1). There was a significant difference in survival times of sheep left to develop clinical disease between the different genotype groups (ManneWhitney test, P