The present study was performed as a response to the recent outbreak of CWD among reindeer in Norway. This outbreak has raised concerns about the possible spread of CWD to both wildlife and to the high number of sheep that graze in the region of the affected reindeer population. This concern still exists despite the effective depopulation of the reindeer in Nordfjella zone 1 [16] since prions have been shown to maintain infectivity in the environment for a long period [38–40], and can therefore potentially be taken up by other species even after the removal of the affected population.
In general, the barrier for interspecies transmission of prions is high [41]. The potential interspecies transmission of CWD from reindeer to sheep will mainly depend on the amount of prions available in the environment [42], to which degree sheep ingest contaminated material (by licking on contaminated surfaces or ingesting contaminated food, soil, or water) and the ability of prions to infect and propagate in the new host [18, 43]. Little is known about the shedding of prions from Norwegian reindeer infected with CWD, but studies of several cervid species in North America have shown that prions are shed in saliva, urine [44], feces [45], and antler velvet [46]. The distribution of prions in peripheral tissues in the Norwegian reindeer cases [7] suggests that shedding of infected materials from these animals occurs in the same way as in cervids with North American CWD. Environmental contamination of prions will therefore depend on the number of infected animals and their shedding capacity, but other factors such as climate and soil characteristics [47–49] will inflict on the availability and persistence of prions in the environment.
In Nordfjella, there has been extensive use of mineral licks intended for sheep [16], but these have also been used by reindeer and other cervids as documented by us (Fig. 2) and others [16]. These mineral lick sites are regarded as high-risk spots for environmental prion contamination and potential cross-species transmission [18, 50, 51]. On the other hand, we know that at the time of stamping out, the prevalence of the disease was less than 1 %, and in approximately half of the infected animals, prions were only found in lymphoid tissues (personal communication S. Benestad), indicating that the individuals probably not had shed prions for a long time and possibly only in small amounts relative to an animal in a later stage [9].
GPS data show that the now eradicated reindeer herd utilized a large area, including the high-density sheep grazing area Fødalen, where about 3200 sheep and lambs grazed during the summer of 2018 (Fig. 1B). We examined 475 of them (approximately 15 %), and prions were not detected in GALT samples from these animals. The other sampling area was Lærdal, which has a lower sheep density (Fig. 1B) but higher reindeer usage than Fødalen (Fig. 1C). From there, we sampled 28 adult sheep that had used the same pasture area for several seasons. GPS data from six of these sheep confirmed considerable overlap of area use with a CWD positive reindeer (Fig. 1D), and the affected reindeer herd in general. Neither of the sheep had evidence of prions in their GALT. To increase the confidence in the negative immunohistochemical results, all RAMALT samples were additionally tested with a validated ELISA rapid test [52]. All samples were negative using this commercial kit, reported to have nearly 100 % specificity when retropharyngeal lymph nodes were analyzed in both cervids and sheep [53, 54], meaning that the likelihood of false negative test results are low.
Transmission of North American mule deer CWD prions via the intracerebral route has been successfully performed for genetically susceptible sheep, but the number of animals and different PrP genotypes were limited [55]. The second passage of the disease in sheep showed the ability of the prions to adapt to the new host and increase the attack rate [56]. In a conference paper, sheep were reported to be infected when challenged intracerebrally with elk CWD, but not after oral inoculation [57]. The difficulties to overcome the species barrier were also shown by intracerebral inoculations of transgenic ovinized mice with white-tailed deer CWD and cervidized mice, with classical scrapie, both of which showed very low attack rates [59]. However, several recent papers report differences between Norwegian and North American CWD isolates [11, 59, 60]. Norwegian reindeer CWD has a remarkable low attack rate in bank voles [10] but transmits efficiently into transgenic and gene-targeted cervidized mice, albeit with longer incubation times than North American moose CWD [59]. An in vitro protein misfolding cyclic amplification (PMCA) study suggests that Norwegian reindeer CWD has a lower species barrier than North American isolates towards sheep, cattle, hamsters, and mice (all transgenic mice substrates), but a higher species barrier towards humans (substrate from humanized mice) compared to North American isolates [60].
The majority of sheep investigated for prion uptake in this study were lambs born during March/April and sampled in September the same year. Generally, oral uptake of prions and further propagation of the disease is more efficient in young lambs compared to adult sheep [61, 62]. Thus, lambs are suitable indicators of oral uptake of environmental prions able to infect sheep, although older sheep that have grazed in the same area for several seasons could be better for demonstrating propagation and aggregation of prions in tissues. Sheep of all age groups investigated in the present study were included due to their history of close temporospatial contact with the population of infected reindeer and because tissues from these animals were accessible from the abattoir.
Choosing tissue type to screen a large number of animals for prion disease should be based on knowledge of the disease progress, including uptake and propagation of prions, in the species of interest. Scrapie in sheep and CWD in cervids have similar pathogenesis with an early accumulation of prions in peripheral tissues and later involvement of the CNS [23, 63]. It is therefore hypothesized that if sheep are able to take up and propagate CWD prions, the pattern of dissemination will be similar between these species. The early accumulation of prions in GALT, including the lymphoid follicle-rich RAMALT [30] and the easy accessibility of the tissue, makes it ideal for sampling and screening purposes [34, 43]. The technique is relatively easy to perform on live animals [30] and RAMALT is also well suited for postmortem sampling [30, 31] of a large number of animals in a short time, as was done on fresh slaughterhouse material in the present study. Other lymphoid tissues such as the medial retropharyngeal lymph node, the third eyelid, distal jejunal lymph node, and the IPP have also proven to be tissues with high diagnostic value [22, 28, 43]. These tissues, however, were less accessible and would have required a longer time to sample in this particular abattoir, which made them less suitable for mass sampling. The IPP in sheep is shown to undergo involution early in life [6], which makes it less reliable as sampling tissue as the animal gets older, but since some studies suggest that accumulation of prions occurs earlier in the IPP than in other lymphoid organs [43], we included IPP from 37 of the lambs in the study.
The histological analysis of over 500 slides confirmed that RAMALT is well suited for post-mortem screening of a large number of sheep. Some publications refer to 6 or more lymphoid follicles in a sample as “sufficient”, while samples with 5 or fewer follicles are considered as “insufficient” [33, 65]. When using these criteria in our work, 83.7 % of the samples are considered “sufficient”. However, even samples with less than 6 lymphoid follicles can contain detectable prions. In sheep with scrapie, the probability of false-negative results based on the number of lymphoid follicles in the sample has been calculated [66]. These calculations are intended to predict the probability of an individual scrapie-positive sheep within a herd to escape detection. Our work, on the contrary, is designed to detect any evidence of prions in sheep that spend time in a certain geographical region, and not necessary to locate an individual animal with prion disease. Thus, sample volume seems to be more important than individual lymphoid follicle count for our type of study. Since we have sampled approximately 15 % of all animals that grazed in Fødalen during the summer of 2018 and examined nearly 13000 lymphoid follicles from these animals, we are confident that the likelihood of false-negative results in our study is low.
The lymphoid aggregates in the rectum area are most abundant in the circumference starting from the mucocutaneous junction and approximately 1 cm proximally [30, 67]. Thus, sampling should be directed to that region. We demonstrated that using the mucocutaneous junction as a guideline is beneficial to achieve high-quality samples. We further showed that when the mucocutaneous junction was missing from the sample due to sampling from a more proximal location of the rectum (designated as SSE0), the mean number of lymphoid follicles was significantly lower than the mean number of lymphoid follicles counted in samples that included the mucocutaneous junction. However, when the sample was covered by more than 50 % SSE, the number of lymphoid follicles seemed to be declining. In our sample pool, only 6 samples fell into this category, a number that was too low to show statistical significance (Fig. 4A).
The samples from the IPP contained a higher mean number of lymphoid follicles than the RAMALT samples. In addition, the sample with the lowest number of follicles in the IPP group contained 9 follicles, while for the RAMALT group, 3.7 % of the samples contained no follicles at all (Fig. 4B, Table 1). This shows that the IPP is a reliable tissue for screening individual young lambs, and although the IPP undergoes involution and the number of lymphoid follicles gradually declines with age, some also remain in adulthood [68]. The number of RAMALT follicles decreases as well with age, but the process is much slower [32]. This favors RAMALT as a reliable tissue for screening different age groups.