Effect of co-infection with a small intestine-restricted helminth pathogen on oral prion disease pathogenesis in mice

The early replication of some orally-acquired prion strains upon stromal-derived follicular dendritic cells (FDC) within the small intestinal Peyer’s patches is essential to establish host infection, and for the disease to efficiently spread to the brain. Factors that influence the early accumulation of prions in Peyer’s patches can directly influence disease pathogenesis. The host’s immune response to a gastrointestinal helminth infection can alter susceptibility to co-infection with certain pathogenic bacteria and viruses. Here we used the natural mouse small intestine-restricted helminth pathogen Heligmosomoides polygyrus to test the hypothesis that pathology specifically within the small intestine caused by a helminth co-infection would influence oral prion disease pathogenesis. When mice were co-infected with prions on d 8 after H. polygyrus infection the early accumulation of prions within Peyer’s patches was reduced and survival times significantly extended. Natural prion susceptible hosts such as sheep, deer and cattle are regularly exposed to gastrointestinal helminth parasites. Our data suggest that co-infections with small intestine-restricted helminth pathogens may be important factors that influence oral prion disease pathogenesis.

Pathogen induced alterations to GALT also have the potential to influence prion uptake and disease susceptibility. For example, orally-acquired prions are initially transported across the gut lumen into Peyer's patches by a specialized population of phagocytic epithelial cells known as M cells [19][20][21] . Oral prion disease is blocked in mice lacking these cells 21,22 , and exacerbated in mice with an increased M cell-density 22 . Oral infection with certain pathogenic bacteria or exposure to inflammatory stimuli such as cholera toxin can increase M-cell density in the intestine [23][24][25] , and with the damage and/or inflammation caused by pathogen infection 26 this could potentially exacerbate oral prion disease pathogenesis by increasing the efficiency of the initial uptake of the prions from the gut lumen. Mononuclear phagocytes (MNP) are a diverse population of macrophages and classical dendritic cells (DC) [27][28][29] that also play important and contrasting roles during prion disease depending on the subset 30 . Whereas CD11c + classical DC aid the propagation of orally acquired prions towards FDC in Peyer's patches in order to establish infection 31,32 , the uptake of prions by macrophages can lead to their destruction 33,34 . Therefore, pathogen-induced alterations to the abundance, trafficking or activity of MNP could similarly influence disease pathogenesis by enhancing the propagation or clearance of orally-acquired prions.
Studies in mice have shown that the host's immune response to a gastrointestinal helminth infection alters susceptibility to co-infection with a variety of other pathogenic microorganisms 35 including Salmonella enterica serovar Typhimurium 36 , Citrobacter rodentium 37 and norovirus 38 . Natural prion susceptible hosts such as sheep, deer and cattle are regularly exposed to gastrointestinal helminth parasites, but little is known of the effects that co-infection with these pathogens may have on oral prion disease pathogenesis. Since neuroinvasion of orally-acquired prions occurs directly from the GALT in the small intestine 12 , in this study the natural mouse small intestine-restricted helminth pathogen Heligmosomoides polygyrus was selected. This parasite is an excellent model for the study of gastrointestinal helminth infections in livestock and humans 39 , being phylogenetically similar to the ruminant parasites Haemonchus contortus and Teladorsagia circumcincta, and human hookworms Ancylostoma duodenale and Necator americanus 40 . We used H. polygyrus to test the hypothesis that the pathology caused by a pathogen co-infection specifically within the small intestine would significantly influence oral prion disease pathogenesis. These data are essential for the identification of important factors can that influence the risk of oral prion disease transmission and to help design effective intervention and control strategies.

Results
Oral H. polygyrus infection causes pathology in the small intestine. Groups of four female C57BL/6J mice were orally infected with 200 H. polygyrus L3 larvae by gavage and faecal egg burdens measured at intervals afterwards to monitor the magnitude of the parasite infection. As anticipated, maximum egg production was observed by 18 days post-infection (dpi) with H. polygyrus and had declined by 32 dpi (Fig. 1A). Histopathological analysis of inflammatory cell infiltrate and changes to the intestinal architecture (using the evaluation scheme described in ref. 41 ) confirmed the presence of detectable pathology within the small intestine by 8 dpi with H. polygyrus, and was significantly greater in the duodenum by 14 dpi (Fig. 1B,C). Goblet cell density in epithelium of the duodenum was also significantly increased from 8 dpi with H. polygyrus (Fig. 1D).
Numerous macroscopic granulomas were also evident in the duodenal wall from 8 dpi with H. polygyrus infection ( Fig. 2A,B). Consistent with previous reports 42 our immunohistochemical (IHC) analysis showed that these structures typically comprised a central core which predominantly contained CD11b + MNP, surrounded by other populations of CD68 + and CD11c + MNP (Fig. 2C). Independent studies have suggested that an inflammatory PrP C -expressing stromal cell population within soft tissue granulomas can support prion replication 43 . IHC analysis did not reveal significant cell-associated PrP C expression within the granulomas induced in the gut wall during H. polygyrus infection. However, they were typically surrounded by a border of PrP C -expressing stromal-like cells (Fig. 2D), as well as cells expressing the neuronal synaptic vesicle marker synaptophysin 1, indicative of enteric nerves (Fig. 2E), suggesting a potential novel site for prion accumulation or neuroinvasion in the gut wall.
Effect of H. polygyrus co-infection on oral prion disease susceptibility. Next, groups of C57BL/6J mice (n = 6-8) were orally-infected with H. polygyrus and 0, 1, 8 or 14 d later the mice were subsequently orally co-infected with a limiting dose of ME7 scrapie prions. The days on which the mice were co-infected with prions were selected based on the known pathogenesis of the H. polygyrus infection in the small intestine 44 : 1 dpi, when the infective L3 larvae penetrate the lining of the duodenum; 8 dpi, during the period when the adult worms re-emerge back into the gut lumen; 14 dpi, when adult worms are tethered around the villi in the duodenum. A separate group of mice were orally-infected with prions alone as a control.
Oral exposure of C57BL/6 mice to a limiting dose of ME7 scrapie prions has an incomplete clinical disease incidence 22,32 , allowing the potential effects of H. polygyrus co-infection on survival times and disease susceptibility to be determined. As expected, four of eight of the mice exposed to prions alone developed clinical disease with a mean survival time of 394 ± 14 d (median = 401 d; Fig. 3A, Table 1). The remaining mice in this group did not develop clinical prion disease up to at least 504 d after exposure. The mice co-infected with prions on d 0, 1 and 14 after H. polygyrus infection also succumbed to clinical prion disease with similar survival times and disease incidences to those infected with prions alone (Fig. 3A, Table 1). However, when the mice were co-infected with prions on d 8 after H. polygyrus infection their mean survival time was significantly extended by approximately 39 d (mean = 440 d, median = 436 d, n = 3/6; Fig. 3A, Table 1).
All the brains from the mice in each infection group that developed clinical signs of prion disease displayed the characteristic spongiform pathology (neuronal vacuolation), microgliosis, astrogliosis and accumulations of prion disease-specific PrP (referred to as PrP d ) that are associated with a terminal, clinical, infection with ME7 scrapie prions (Fig. 3B). The distribution and severity of the spongiform pathology was also similar in the brains of the mice that developed clinical signs of prion disease (Fig. 3C). None of these histopathological signs of prion disease were detected in any of the brains from the clinically-negative mice (data not shown). Thus although   Effect of H. polygyrus co-infection on oral prion disease. Groups of 6-8 C57BL/6J mice were orallyexposed to a limiting dose of ME7 scrapie prions on either d 0, 1, 8 or 14 after oral infection with H. polygyrus L3 larvae. A separate group of mice were exposed to prions alone as a control. (A) Mean prion disease survival times (±SEM) for the orally-infected mice from each group. When the mice were co-infected with prions on d 8 post-infection with H. polygyrus infection, the mean survival times were significantly extended by approximately 39 days when compared to mice infected with prions alone. n = 3-6 clinically positive mice/ group *P < 0.05; **P < 0.01, One-way ANOVA. (B) High levels of spongiform pathology (H&E, upper row), heavy accumulations of PrP d (brown, second row), active microglia expressing Iba-1 (brown, third row) and reactive astrocytes expressing GFAP (brown, bottom row) were detected in the brains of all the orally-exposed mice with clinical prion disease at the time of cull. Sections were counterstained with haematoxylin to detect cell nuclei (blue). Images show the thalamus. (C) The severity and distribution of the spongiform pathology (vacuolation) within each clinically-affected brain was scored on a scale of 1-5 in nine grey matter regions: G1, dorsal medulla; G2, cerebellar cortex; G3, superior colliculus; G4, hypothalamus; G5, thalamus; G6, hippocampus; G7, septum; G8, retrosplenial and adjacent motor cortex; G9, cingulate and adjacent motor cortex; Each point represents the mean vacuolation score ± SEM, n = 3-6 clinically positive mice/group, not significantly different to Prions alone, Two-way ANOVA. Although granulomas were present in the small intestines of many of the co-infected mice, no histopathological evidence of PrP Sc accumulation was detected within them (Fig. 4E).
Our data suggested that specific pathological effects on the small intestine at 8 dpi with H. polygyrus had impeded the initial accumulation of prions upon FDC in Peyer's patches. Further experiments were undertaken to attempt to identify the pathological factors that mediated this effect.

H. polygyrus infection does not affect the density and function of M cells.
Orally-acquired prions are initially transferred across the intestinal epithelium into Peyer's patches by M cells, and this is essential to establish infection 21 . Furthermore, changes to M cell-density in the specialized follicle-associated epithelium (FAE) that covers the Peyer's patches can directly influence oral prion disease susceptibility 22 . Whole-mount IHC showed that the density of M cells expressing the mature M-cell marker glycoprotein 2 (GP2; refs. 45,46 ) in FAE covering the Peyer's patches was unchanged after H. polygyrus infection (Fig. 5A,B). The in vivo ability of M cells to acquire and transcytose particulate antigens can be assessed by measuring the uptake of fluorescently-labelled latex microbeads from the gut lumen into Peyer's patches 22,45,47,48 . Here, C57BL/6J mice (n = 4) were orally infected with H. polygyrus and at 5 dpi orally-gavaged with fluorescent microbeads. Uninfected mice were orally-administered fluorescent microbeads as a control. Seventy two hours later (d 8 post-H. polygyrus infection) the number of microbeads that had been transcytosed into the subepithelial dome (SED) region of their Peyer's patches was quantified on tissue sections by fluorescence microscopy. Similar levels of fluorescent microbeads were detected in the SED regions of the Peyer's patches from mice from each group (Fig. 5C). These data show that the density of M cells in the FAE, and their ability to transcytose particulate antigens from the gut lumen into Peyer's patches, was not impaired following H. polygyrus infection.

Effects of H. polygyrus infection on FDC in Peyer's patches and the MLN. FDC initially retain
complement-opsonized prions on their surfaces in a complement receptor-dependent (CR) manner, and CR1 (CD35) is considered to play a prominent role 49 . IHC and morphometric analysis showed that the FDC were similar in size in the Peyer's patches from H. polygyrus-infected and uninfected control mice (area of CD35 + immunostaining; Fig. 6A,B). The expression of PrP C by FDC is essential to sustain prion replication upon their surface 14,50 . Our analysis also showed that the magnitude of the PrP C -expression co-localized upon FDC was similar in Peyer's patches from each group (Fig. 6A,C). The overall area of the Peyer's patches was also unchanged during H. polygyrus-infection (Fig. 6D). In contrast, in the MLN a significant increase in the size of the FDC and the expression of PrP C was observed on d 14 post H. polygyrus infection when compared to uninfected control mice ( Fig. 6E-G). These data clearly show that the delayed accumulation of PrP Sc in the Peyer's patches and MLN of mice infected with prions on d 8 post H. polygyrus infection was not simply due to reduced FDC size or PrP C expression.

Effects of H. polygyrus infection on the MNP in the lamina propria. Prions are considered to utilize
CD11c + MNP to mediate their delivery to FDC 31,32 , whereas macrophages appear to phagocytose and destroy prions 34 . IHC and morphometric analysis suggested that the % area of CD11c + (Fig. 7A,B) and CD68 + (Fig. 7C,D) immunostaining in the lamina propria was unchanged in the small intestines of H. polygyrus infected mice when compared to uninfected controls. However, a significant increase in the % area of CD11b + immunostaining (  Prion disease durations for individual mice (from time of injection with prions to cull at clinical end-point).
The notation "NX > 504" means mice that were free of the clinical and histopathological signs of prion disease up to at least this time after oral exposure. d Incidence = no. animals displaying clinical signs of prion disease/no. animals tested. e Incidence = no. animals with histopathological signs of prion disease in the brain (spongiform pathology)/no. animals tested. f Statistical differences between survival times were compared by one-way ANOVA with Tukey's post hoc test. *P < 0.05, when compared to mice orally-exposed to prions alone. NS, not significant. ( www.nature.com/scientificreports www.nature.com/scientificreports/  29 . We have shown that prions are also subsequently acquired by CD11c + MNP after oral exposure 14,31,51 and propagated by them towards the FDC in the B cell follicles of Peyer's patches 32 . IHC and morphometric analysis revealed that the abundance of CD11c + MNP was specifically and significantly increased in the Peyer's patches on d 8 after H. polygyrus infection (Fig. 8A,B). In contrast, the number of CD68 + MNP (indicative of tissue macrophages) in the Peyer's patches was unchanged during H. polygyrus infection (Fig. 8A,C). Stratification of the Peyer's patches into three distinct anatomical regions (SED, B cell mantle and B cell follicle; Fig. 8D) revealed that the distribution of the CD11c + MNP was also altered after H. polygyrus infection. A specific and significant increase in the number of CD11c + MNP within the B cell mantle region of the Peyer's patches on d 8 after H. polygyrus infection was observed (Fig. 8E). The distribution of the CD68 + MNP, in contrast, was unchanged (Fig. 8F) consistent with the cell abundance data (Fig. 8C). This analysis suggested that the significantly disturbed distribution of the MNP within the Peyer's patches on d 8 after H. polygyrus infection may have impeded the ability of these cells to efficiently propagate the prions from the SED region down towards the FDC within the B cell follicles.

DISCUSSION
Factors that influence the early accumulation of prions in the small intestinal Peyer's patches can directly influence oral prion disease pathogenesis. Gastrointestinal pathogens can cause significant disturbances to the gut mucosa and the GALT. Therefore, in this study we determined whether pathology specifically in the small intestine, such as that caused by co-infection with the natural mouse helminth parasite H. polygyrus would influence www.nature.com/scientificreports www.nature.com/scientificreports/ www.nature.com/scientificreports www.nature.com/scientificreports/ oral prion disease pathogenesis. Mice were co-infected with prions at distinct intervals after H. polygyrus infection based on the parasite's known pathogenesis in the host's duodenum 39,44 to determine whether the pathology in the duodenum during these distinct phases of the helminth infection affected oral prion disease pathogenesis. For example, we reasoned that damage to the epithelium might increase gut permeability, and by doing so raise the uptake of prions across it and exacerbate prion disease pathogenesis. Conversely, it was plausible that increased MNP abundance may enhance the sequestration of prions.
Our data revealed that the early accumulation of prions within the small intestinal Peyer's patches was reduced and survival times significantly extended when the mice were co-infected with prions on d 8 after H. polygyrus infection. This effect was transient and specific to this period of the H. polygyrus infection, as no effects were observed when the mice were co-infected with prions at other intervals in relation to the helminth infection. This suggests that specific effects of the H. polygyrus co-infection on the Peyer's patches had impeded the efficient spread of orally acquired prions from the gut lumen to the brain. We have previously shown that co-infection of mice with the large intestine-restricted helminth pathogen Trichuris muris does not influence oral prion disease pathogenesis 12 . This is most likely because the large intestinal GALT are not important early sites of prion accumulation and neuroinvasion 12 . In contrast, our data in the current study suggest that co-infections with small intestine dwelling helminth pathogens may be important factors that can significantly influence oral prion disease pathogenesis.
Little, if anything, was known of the effects of H. polygyrus infection on the key cell populations within the Peyer's patches that play an important role in oral prion disease pathogenesis. M cells are considered to be the critical gate-keepers of oral prion infections 21,22 . However, the density or function of these cells was not affected by H. polygyrus infection. The FDC within the B cell follicles are the essential sites of prion replication in the GALT 12,52,53 , but the status of these cells was similarly unaffected in the Peyer's patches of H. polygyrus infected mice. This clearly demonstrated that the effects of helminth infection on oral prion disease pathogenesis were independent of direct effects on M cells or FDC in Peyer's patches.
Infections with gastrointestinal helminth parasites such as H. polygyrus characteristically induce robust T helper (Th) 2 immune responses associated with expression of cytokines such as interleukin (IL)-4 and IL-13, and potent regulatory T-cell responses 54 . However, data show that deficiency in T cells [55][56][57][58] or the cytokines 59 IL-4 or IL-13 do not influence peripheral prion disease pathogenesis or susceptibility. This suggests that the effects of helminth co-infection on oral prion disease pathogenesis were similarly independent of effects on T cells.
Following their initial transcytosis across the gut epithelium by M cells the prions are subsequently acquired by CD11c + MNP such as the classical DC in the underlying SED regions of the Peyer's patches 14,31,51 . The propagation of prions by these MNP towards the FDC in the Peyer's patches is important to establish host infection and efficient neuroinvasion 32 . Our analysis revealed that the abundance of CD11c + MNP in the B cell mantle region of the Peyer's patches was specifically and significantly increased on d 8 after H. polygyrus infection. This implied that the disturbed distribution of the MNP within the Peyer's patches on d 8 after H. polygyrus infection may have impeded the ability of these cells to efficiently propagate the prions from the SED region down towards the FDC within the B cell follicles. Other MNP populations such as macrophages appear to play a host protective role during prion disease by aiding the clearance of prions 30,33,34 . The "alternative" activation of macrophages is a characteristic feature of the helminth-induced Th2 response 60 . These alternatively-activated macrophages typically contribute to the regulation of inflammatory responses, and promote wound healing and parasite expulsion during helminth infections 39 . Whether effects of H. polygyrus infection on the activation status of the MNP in the Peyer's patches may have also enhanced their ability to phagocytose and destroy the prions remains to be determined.
The factor/s during H. polygyrus infection that were responsible for the altered positioning and trafficking of the CD11c + MNP within the Peyer's patches are not known. Day 8 after H. polygyrus infection was selected as it is around this time that the adult worms emerge from the submucosa into the lumen of the duodenum 39,44 . During this process, significant amounts of parasite-derived antigens and debris from damaged host cells are liberated into the gut lumen. Thus, it is plausible that the presence of this material in the gut lumen stimulates the recruitment of CD11c + MNP to the Peyer's patches.
Orally-acquired prions accumulate first in the small intestinal Peyer's patches and subsequently spread via the lymphatics and bloodstream to most other secondary lymphoid tissues, including the MLN and the spleen. However, our previous data show that the subsequent accumulation of the prions within these tissues is not essential for efficient neuroinvasion from the intestine 10,12,22,32 . In mice co-infected with prions on d 8 after H. polygyrus infection the accumulation of prions within the MLN was also reduced despite the presence of larger FDC networks. This suggests that any prions that are initially transported to the MLN immediately after oral exposure either reach this tissue in insufficient abundance to establish infection on FDC, or alternatively, are not propagated towards the B cell follicles. As a consequence, the prions are most likely phagocytosed and destroyed by the macrophages within the MLN 33,34 .
Studies of sheep with natural scrapie have revealed that orally acquired prions are similarly first detected in the GALT in the upper gastrointestinal tract before spreading to the draining lymph nodes and onwards to other lymphoid tissues. The large intestinal GALT such as the caecal patches are not early sites of prion accumulation [61][62][63][64] , consistent with data from experimental oral transmissions to mice 12 . Natural prion disease susceptible hosts such as sheep, deer and cattle are regularly exposed to helminths in field situations. Thus, co-infection with immunolabelling in MLN from mice from each group revealed a significant increase on d 14 post H. polygyrus infection when compared to uninfected controls (*P < 0.05; **P < 0.01; ***P < 0.007; one-way ANOVA). Horizontal bars, median. Data derived from 1-7 fields of view (images) per mouse, n = 3-4 mice/group. www.nature.com/scientificreports www.nature.com/scientificreports/ small intestine-restricted helminth pathogens around the time of oral exposure may similarly influence oral prion disease pathogenesis in naturally affected host species. An independent study has previously attempted to determine the influence that helminth infections may have on oral prion disease pathogenesis 65 . The gastrointestinal www.nature.com/scientificreports www.nature.com/scientificreports/ nematode parasite T. circumcincta establishes infection in the mucosa and lumen of the abomasum of sheep. When lambs with high genetic susceptibility to natural sheep scrapie were experimentally co-infected with T. circumcincta at monthly intervals from 6 to 11 months old, prion disease survival times were shortened 65 . However, very young lambs are highly susceptible to prion infection 66 and are considered to acquire natural sheep scrapie from infected ewes via the oral route around the time of birth 67 . Treatments that can delay the accumulation of prions in the GALT do not influence disease pathogenesis once neuroinvasion has occurred and the prions have established infection in the peripheral nervous system 5 . Since the lambs in the above study 65 were co-infected with T. circumcincta long after prion neuroinvasion from the intestine is likely to have occurred 61,62 , the reduced survival times are unlikely to be a consequence of effects of the parasite infection on the uptake or accumulation of the prions within the GALT. However, it is plausible that the parasite-mediated damage to the abomasum stimulated the systemic release of inflammatory mediators that exacerbated the development of neuropathology within the CNS 68 .
In the current study mice were given a single infection with H. polygyrus. Although experimental helminth infections in mice are commonly studied after administration of a single dose, natural host species in field situations are likely to be regularly exposed to these pathogens over long periods. The effects that chronic helminth exposure may have on MNP abundance and distribution within the Peyer's patches are not known. It is plausible that repeated helminth exposure may lead to the prolonged recruitment of MNP into the B cell mantel region of the Peyer's patches further enhancing the sequestration of orally-acquired prions.
Soft tissue granulomas are a prominent feature within chronically-inflamed tissues in humans and animals. One study has shown that an inflammatory PrP C -expressing stromal cell population in experimentally-induced soft tissue granulomas can support prion replication in infected mice 43 . Similar evidence of ectopic PrP Sc accumulation has been observed in the newly formed lymphoid follicles of chronic lymphofollicular inflammatory sites in scrapie-affected sheep 69 . Granulomas can form around persistent foreign bodies, and are induced in the intestines of H. polygyrus infected mice in response to the invasion of the L3 larvae into the submucosa. However, the granulomas induced in the intestines of H. polygyrus infected mice predominantly contained MNP, lacked organized lymphoid follicles and PrP C -expressing stromal cells within them. Although the granulomas appeared to be surrounded by a border of PrP C -expressing stromal-like cells they were unable to support PrP Sc accumulation in mice co-infected with prions.
Minor oral (lingual) lesions can enhance susceptibility to CWD prions in cervid PrP-expressing transgenic mice by facilitating direct neural invasion 26 . However, our own studies using two distinct gastrointestinal helminth pathogens show that the damage these parasites cause to the mucosa in the small (current study) and large intestines 12 has little, if any, effect. Instead our data suggest that oral prion disease pathogenesis is significantly altered when the density or the status of the cells utilized by prions to establish infection within small intestinal Peyer's patches is affected, including: M cells 15,21,22 ; MNP 31,34 (current study); and FDC 5,16 . Natural prion disease susceptible hosts such as sheep, deer and cattle are regularly exposed to helminths. Thus, co-infection with small intestine-restricted helminth pathogens around the time of oral prion exposure may be important factors that can influence prion disease pathogenesis in naturally affected host species.

Materials and Methods
Ethics Statement. Approvals for the mouse studies and regulatory licences were received from The Roslin Institute's and University of Edinburgh's ethics committees, and the UK Home Office. The animal experiments described in this study were undertaken under the authority of UK Home Office Project Licences (PPL60/4325 & PA75389E7) under the terms, conditions and regulations of the UK Home Office ' Animals (scientific procedures) Act 1986' . Where appropriate and required, necessary efforts were undertaken to reduce harm and suffering. Mice were humanely culled at the end of the experiments using a UK Home Office Schedule One method.
Mice. Six-eight weeks old female C57BL/6J (Charles River, Margate, UK) were used throughout this study. All mice were maintained under SPF conditions.
Oral infection with H. polygyrus. Mice were orally infected by gavage with 200 H. polygyrus third stage (L3) larvae prepared as described 70 . Egg counts were determined by weighing faecal pellets from individual mice prior to resuspension in 1 ml of saturated sodium chloride/50% w/v glucose solution. Eggs were then counted using a McMaster chamber and the eggs/g faeces determined. For analysis of pathology, blinded sections of duodenum and ileum (2 per mouse) were scored for inflammatory cell infiltrate (0-4) and changes to the intestinal architecture (0-4) using the evaluation scheme described for H. polygyrus pathology in 41 . Mean scores per mouse were combined to give a maximum score of 8.
Oral Prion Exposure and Disease Monitoring. Individual food pellets were first doused with 50 µl of scrapie brain homogenate derived from mice with terminal ME7 scrapie prion infection. Each pellet contained approximately 3.3 log 10 intracerebral ID 50 units. Each mouse was then individually housed in a bedding-and food-free cage with water provided ad libitum, and fed an individual prion doused food pellet 5,10,12,16,21,31 . The mice were returned to their original cages as soon as no signs of the food pellet were evident (indicating that it had been completely ingested). The mice were then coded and assessed by independent technicians at weekly intervals for the clinical signs of prion disease and scored as "unaffected", "possibly affected" and "definitely affected" using standard criteria established for mice clinically-affected with ME7 scrapie prions 5,10,12,16,21,31 . Mice were either culled at the standard clinical endpoint, or survival times recorded for those that remained free of the clinical signs of prion disease after exposure. The clinical status in each mouse was confirmed by the histopathological assessment of prion disease pathology in the brain. The magnitude and distribution of the spongiform pathology (vacuolation) in nine distinct grey-matter regions of the brain was compared as described 71 . www.nature.com/scientificreports www.nature.com/scientificreports/ IHC and Immunofluorescent Analyses. To detect M cells by whole-mount immunostaining 21 Peyer's patches were first fixed using BD Cytofix/Cytoperm (BD Biosciences, Oxford, UK), and then immunostained with rat anti-mouse GP2 mAb (MBL International, Woburn, MA). The Peyer's patches were then stained with Alexa Fluor 488-conjugated anti-rat IgG Ab (Invitrogen, Paisley, UK), rhodamine-conjugated Ulex europaeus agglutinin I (UEA-1; Vector Laboratories Inc., Burlingame, CA) and Alexa Fluor 647-conjugated phalloidin to detect F-actin (Invitrogen).
Goblet cells were detected using a Periodic Acid Schiff (PAS) Stain Kit (Abcam, Cambridge, UK) according to the manufacturer's instructions.
To detect disease-specific PrP (PrP d ) in intestines, MLN and brains were first fixed in periodate-lysine-paraformaldehyde fixative and then embedded in paraffin wax and 6 µm sections cut. The sections were first pre-treated with a combination of hydrated autoclaving (15 min, 121 °C, hydration) followed by 5 min immersion in formic acid (98%) to enhance PrP d detection, before immunostaining with 1B3 PrP-specific pAb. Astrocytes were detected by immunostaining with anti-glial fibrillary acidic protein (GFAP; DAKO, Ely, UK). Microglia were detect by pre-treatment with citrate buffer, followed by immunostaining with anti-ionized calcium-binding adaptor molecule 1 (Iba-1; Wako Chemicals GmbH, Neuss, Germany). To detect FDC, the sections were pre-treated with Target Retrieval Solution (DAKO) before immunostaining with rat anti-CD21/35 mAb (clone 7G6, BD Biosciences). Sections were counterstained with biotin-conjugated species-specific secondary antibodies (Stratech, Soham, UK) followed by HRP-conjugated to the avidin-biotin complex (ABC kit, Vector Laboratories, Peterborough, UK) with DAB (Sigma) as a substrate. Haematoxylin was used as a counterstain to detect cell nuclei. Paraffin-embedded tisssue immunoblot analysis 73 was used to detect proteinase K-resistant PrP Sc using with PrP-specific pAb 1B3. Image analysis. ImageJ software (http://rsb.info.nih.gov/ij/) was used for morphometric image analysis on coded sections as described 74 . First, pixel intensities across all the areas of each image (immunostained and non-stained regions) were measured using an ImageJ macro and used to obtain the background pixel intensity threshold values. These values were then applied to all subsequent analyses. The number of pixels in each colour channel (black, red, green, yellow etc.) in each image were then determined automatically. These data are presented as the proportion of positively immunostained pixels/total number of pixels (all colours) in the specific region of the image. We typically analysed 1-7 images from each mouse, and multiple mice were analysed from each group (n = 3-8 mice/group). The figure legends display the sample sizes for every parameter analysed by this method in this study. For analysis of CD11c + cell location, images were orientated with the FAE at facing upwards and the height of the patch and the Y-axis coordinate of each CD11c + cell recorded. The location of each CD11c + cell was calculated as a percentage of the total patch height and then allocated to three regions approximately representing the B cell follicle (0-33.33% of total patch height), B cell mantle (33.34-66.67%) and SED (66.68-100%).
Oral gavage with fluorescent microbeads. Mice were given a single oral gavage of 2 × 10 11 of Fluoresbrite Yellow Green labelled 200 nm microbeads (Polysciences, Eppelheim, Germany) in 200 µl PBS. Mice were culled 72 h later and Peyer's patches and small intestine segments were snap-frozen at the temperature of liquid nitrogen. Serial frozen sections (6 µm in thickness) were cut on a cryostat and counterstained with DAPI. The number of beads in the SED from 3 sections of two Peyer's Patches per mouse (n = 3-4 mice/group; total 4-19 SED/mouse studied) were counted. Images of SED were acquired using Nikon Eclipse E400 fluorescent microscope using Micro Manager (http://www.micro-manager.org). Tissue auto-fluorescence was subtracted from displayed images using ImageJ.
Statistical Analyses. All data are presented as mean ± SD. Unless indicated otherwise, differences between groups were analysed by One-or Two-way ANOVA where appropriate. Comparisons between groups were performed using a Tukey's (One-way ANOVA) or Sidak's (Two-way ANOVA) post-hoc test. In instances where there was evidence of non-normality (identified by the Kolmogorov-Smirnov, D' Agostino & Pearson omnibus, or Shapiro-Wilk normality tests), data were analysed by appropriate non-parametric tests. Values of P < 0.05 were accepted as significant.

Data Availability
All relevant data are present within the paper.