All procedures involving animals were approved by the United Kingdom (UK) Home Office under the Animals (Scientific Procedures) Act 1986.
Study 1 was initiated in 2005 to test the hypothesis that scrapie can be transmitted to sheep via milk. As there was no scientific evidence at the time that scrapie was transmissible via milk, the number of sheep required to test the hypothesis could potentially be very large. It was considered feasible to do a study with 59 scrapie-affected sheep conducted over four years, which would be sufficient to confirm transmission if 5% of sheep (with 95% confidence interval) excreted the scrapie agent and transmitted disease to a milk-fed lamb. The interim results of Study 1  led to a modification of the project design, which was implemented in subsequent years, to address issues not known at the start of the study, such as the effect of potential lateral transmission, high somatic cell count and clinical status on the interpretation of the results (Study 2, started 2007), and to provide further information on infectivity of colostrum or milk (Study 3, started 2008).
Study 1. Feeding milk from scrapie infected sheep to lambs with mixing of lambs
The methods have been described in detail previously . Briefly, colostrum and milk (unless otherwise specified, colostrum and milk is subsequently referred to as milk) was collected from 12 VRQ/VRQ ewes naturally infected with scrapie either at clinical stage (eight) or pre-clinical stage (four sheep). The milk was frozen at below −20°C and fed subsequently to a total of 18 Cheviot VRQ/VRQ lambs from a flock free from classical scrapie : milk from individual ewes was fed to newborn lambs in the same order it was collected (e.g. milk from day 1 was fed shortly after birth, followed by milk from day 2 etc.), without pooling of milk, i.e. milk from individual ewes was only fed to one particular set of two lambs or single lambs if the volume was insufficient to feed two lambs; milk was not pooled. Milk recipients of individual ewe’s milk were housed separately until all scrapie milk was consumed, after which they were mixed and fed milk replacer (Lamlac, Volac International Ltd., Royston, UK).
Five ‘lateral transmission control’ lambs (weaned lambs from the same scrapie-free flock) were mixed with the milk recipients aged 70–72 days (three lambs) and 112 days (two lambs) respectively to assess whether horizontal transmission between sheep occurs after mixing of milk recipients.
Ten VRQ/VRQ Cheviot lambs were housed in the same accommodation but separate pen as ‘building controls’ to control for environmental contamination. These lambs were kept with their dam until weaning age.
At approximately 9 months of age, the scrapie status was evaluated based on a biopsy of the rectal mucosa under local anaesthesia (mixture of Lidocaine 2.5% and Prilocaine 2.5%, EMLA cream 5%, AstraZeneca, London, UK) in live animals (15 scrapie milk recipients, all five lateral transmission controls, nine building controls) to check for the presence of PrPd in RAMALT. For sheep that died of intercurrent diseases (three scrapie milk recipients, one building control) selected LRS tissue (distal ileum, mesenteric lymph node, spleen) and the brain were collected post mortem and examined for the presence of PrPd. Postmortem tests comprised the immunohistochemical examination of formalin-fixed and wax-embedded tissue sections of brain (obex) and LRS tissue, including RAMALT, using rat monoclonal antibody R145 as described previously . A section of the obex was also routinely stained with haematoxylin-eosin (H&E) and examined for the presence of vacuolar changes. A fresh sample of the caudal medulla was additionally examined for the presence of PrPres by Western immunoblot (Bio-Rad TeSeE Western blot, Bio-Rad Laboratories, Hemel Hempstead, UK) as described elsewhere .
All scrapie milk recipients and lateral transmission controls were culled upon development of definite clinical signs (clinical end-point). The clinical examination followed the same short examination protocol as used for scrapie-affected goats . Building controls were culled after the last scrapie-affected sheep was culled. The clinical end-point was reached when animals displayed abnormalities in sensation (frequent pruritic behaviour and a positive scratch test or alopecia) or movement (ataxia or tremor).
Study 2. Feeding milk from scrapie infected sheep to lambs without mixing of lambs
The milk from seven 18 month-old VRQ/VRQ ewes (Poll Dorset × Friesland) born in the VLA flock with a high incidence of scrapie , as mentioned in Study 1 above, was collected by hand and frozen at −80°C to be used in the following year. All ewes tested negative for MVV antibodies by Agar Gel Immunodiffusion Precipitin Test (AGIDT) using the Maeditect test kit (AHVLA Weybridge, Addlestone, UK) . An aliquot of each weekly milk sample collected after the colostral period was sent for SCC determination (National Milk Records plc, Chippenham, UK). A routine bacteriological examination was carried out (by AHVLA Winchester, Hampshire, UK) if the SCC exceeded 106 cells per ml. The scrapie status of each ewe was confirmed by immunohistochemical examination of a RAMALT biopsy collected prior to milking, at approximately one year of age, as described above. Assessment of clinical signs associated with scrapie in the lactating ewes was made by experienced animal husbandry staff familiar with the animals, who used a three point classification system to evaluate the clinical scrapie status. Mild signs included mild behaviour change, some pruritus with only minor fleece damage, minor muscle fasciculations; unequivocal signs included obvious behaviour change, pruritus with fleece damage but without skin lesions, tremor and some loss of weight or condition, and strong signs were ataxia, pruritus with minor skin damage and poor body condition.
Ewes were generally culled after development of at least unequivocal signs of scrapie, which was confirmed by a neurological examination prior to cull using the same methodology as described above. Disease was confirmed by histopathological (H&E) and immunohistochemical (antibody R145) examination of formalin-fixed and wax-embedded samples of the obex as described above as well as examination of a fresh sample of the caudal medulla by Western immunoblot (VLA Hybrid technique ).
The milk from the entire lactation was fed to a pair of VRQ/VRQ lambs born from ewes of the scrapie-free flock. The scrapie-free status of the ewes was confirmed by postmortem examination of the brain using immunohistochemistry (method as above) and ELISA (Bio-Rad TeSeE, Bio-Rad Laboratories) according to the manufacturer’s instructions . In addition, the inguinal (mammary) lymph nodes and the mammary gland were examined immunohistochemically for the presence of PrPd in lymphoid follicles.
The feeding protocol was as described for Study 1 (see above) but this time all pairs of lambs were kept in separate pens with separate equipment and entries to avoid lateral transmission. Prior to housing of sheep all pens were decontaminated with sodium hypochlorite (20% solution with 20,000 ppm available chlorine), which is effective against prions .
Milk replacer (Lamlac) was fed after all scrapie milk was consumed, followed by a diet of straw and concentrates at weaning age. Mixing of lambs only took place after PrPd was detected in a RAMALT biopsy in at least one lamb of each pair. Biopsies were taken at 4.5 months of age and repeated two weeks later if the tissue was inadequate for a diagnosis. Five VRQ/VRQ Cheviot lambs were housed in a separate pen of the same accommodation (building controls) and kept on a similar diet as the scrapie milk recipients (colostrum from the scrapie-free dam, milk replacer, then straw/concentrates).
The procedures for culling of animals and the protocol for examination of tissue were identical to Study 1 (see above).
Study 3. Feeding colostrum and milk from scrapie infected sheep to separate lambs
Milk from five 18–21 month-old VRQ/VRQ ewes from the VLA scrapie flock (one Poll Dorset, four Poll Dorset × Friesland) was collected. All ewes had no detectable antibodies against MVV by AGIDT. Identical to the previous study, milk was frozen and stored for a year and weekly samples were submitted for SCC. A RAMALT biopsy taken at 10 months of age in one and at 16 months of age in four ewes was examined immunohistochemically for the presence of PrPd to confirm scrapie status. Assessment of the clinical status and disease confirmation by postmortem tests was made as described above, including immunohistochemical examination of inguinal lymph nodes and mammary gland.
The milk from each ewe was separated into colostrum (= lactation up to day 4) and milk (lactation from day 5). Recipient lambs were born from ewes derived – as before – from the classical scrapie-free flock; 15 of the dams were used for another project and no follow-up on their scrapie-free status was possible; the other four, dams of five lambs, were scrapie-negative on brain examination by ELISA (Bio-Rad TeSeE, Bio-Rad Laboratories). Scrapie colostrum and milk were fed to a pair of lambs so that one pair received scrapie colostrum only and one pair received scrapie milk only. Scrapie colostrum was fed within 10 minutes after birth whereas scrapie milk was fed within 23 hours after birth but not earlier than 9 hours after birth since these lambs first received colostrum from their scrapie-free dams. The subsequent procedure was identical to previous studies: feeding of scrapie colostrum and milk in the same order it was collected, without pooling of milk, feeding milk replacer after all colostrum or milk was consumed and strict separation of paired lambs until scrapie diagnosis was made from a RAMALT biopsy taken from 4 months of age. Building controls comprised four VRQ/VRQ Cheviot lambs housed in a separate pen of the same accommodation and kept on a similar diet as the scrapie milk recipients (colostrum from the scrapie-free dam, milk replacer, then straw/concentrates). Sheep were again housed in medium security accommodation that also contained sheep and building controls from the previous two studies. As before, pens were decontaminated with sodium hypochlorite prior to movement of lambs. The procedures for culling of animals and the protocol for examination of tissue were identical to the first two studies (see above).
Application of PMCA to detect PrPscin milk
Milk samples were tested from six VRQ/VRQ scrapie-affected sheep that also provided milk for the transmission study to lambs (ewes from Study 3, see Table 2 and ewe 139/6 from Study 1, which fed one lamb culled at 105 days of age that was negative for scrapie by postmortem tests , see also “Additional file 1: summary”, which shows this dam and its milk recipient), with corresponding control samples from five healthy VRQ/VRQ sheep from the scrapie-free flock. Other milk samples were derived from one ARQ/ARQ sheep at 34 months post oral challenge with 5 g of BSE brainstem homogenate, which came from a BSE research flock , and two scrapie field cases (case 134/11: ARQ/ARQ, case 127/11: ARH/ARQ), with corresponding control samples from 11 healthy ARQ/ARQ sheep from the scrapie-free flock. With the exception of the samples that came from the two scrapie field cases that resembled colostrum, all tested samples were milk and not colostrum, i.e. they were collected on days after day 4 of lactation. Samples were collected at up to six different time points (depending on the length of lactation and availability of samples) and tested individually; in the ARX/ARQ TSE cases milk from each udder half was tested separately. Tables 3 and 4 list the animals with information on status of infection and lactation. Both scrapie field cases tested negative for antibodies against Maedi-Visna virus by AGIDT.
Inguinal lymph nodes and mammary gland of the sheep with BSE and the two scrapie field cases were examined by immunohistochemistry as described above.
Milk samples from ewes that also provided milk fed to lambs were collected weekly (volume 50 ml) and separated into whey, cream and cell fractions for further testing following a method described elsewhere  before being stored at −80°C. As whole milk was tested for comparison with the in vivo data, these fractions were thawed and recombined prior to further processing. All other milk samples were frozen at −80°C without being fractionated and thawed prior to testing. The methods used for the serial PMCA followed those described previously . Briefly, whole or recombined milk samples were mixed for 1 minute with 9.5% (v/v) EDTA (V4231 Promega Corp, Madison, WI, USA), 0.28% (w/v) Sodium Deoxycholate (D6750 Sigma-Aldrich, St Louis, USA) and 0.28% (v/v) Igepal CA-630 (I3021 Sigma-Aldrich) and clarified by centrifugation at 16000g for 10 minutes at 10°C. Supernatants below the fat layer were diluted 1:10 in PMCA substrate and supplemented with polyadenylic acid (PA) (P9403 Sigma-Aldrich) at 100 μg/ml to enhance the efficiency of amplification . Substrates were prepared from ovine brain tissue from a VRQ/VRQ sheep (PG1521/10) and an ARQ/ARQ sheep (PG0220/11) and were used for testing VRQ/VRQ-derived and ARQ/ARX-derived milk samples respectively; both brains were from sheep from the scrapie-free flock, which tested negative for TSE by ELISA (Bio-Rad TeSeE, Bio-Rad Laboratories). Diluted samples were subjected to four rounds (VRQ/VRQ samples) or 11 rounds (ARQ/ARX samples) of serial PMCA, diluting samples 1/3 in fresh substrate (same substrate with PA) between each round. Each round comprised 48 consecutive cycles of sonication (40 seconds at 250 W) and incubation (29 minutes 30 seconds). PMCA products were stored frozen at −20°C until analysed. Products were analysed by enzyme immunoassay (EIA) (IDEXX HerdChek BSE-Scrapie Antigen Test Kit, IDEXX Laboratories, Westbrook, USA) using a modified protocol. Briefly, PMCA products were diluted 1:5 in kit homogenisation buffer then 4:5 with kit plate diluent. The sample (100 μl) was applied to the capture plate for 180 minutes at room temperature (RT) (~20°C). Excess reagents were washed away (kit wash 1) and bound sample incubated with conditioning buffer for 10 minutes at RT. Following a further wash (kit wash 2) wells were incubated with the small ruminant anti-PrP horseradish peroxidase conjugate for 90 minutes at RT then followed by another wash (kit wash 2). Visualisation of bound PrPSc was achieved using 3,3′,5,5′-tetramethylbenzidine (TMB) substrate and measured at 450 nm using a reference filter at 620 nm (Perkin Elmer Envision 2104 multi-label reader). Amplification was determined by comparison with pre-amplification reference samples . In each experiment negative control samples were included to monitor both de novo synthesis and putative contamination. The substrate only controls used in each experiment comprised of substrate spiked with extract from a control milk sample (derived from ewes N290 and K193 from the scrapie-free flock for the VRQ/VRQ sample and ARQ/ARX sample test runs respectively). Eight replicates were included in each experiment. Each sample was analysed four times. Following PMCA all samples with an absorbance of greater than 2 were counted as positive. We use PrPsc to describe scrapie-associated prion protein from the PMCA experiments since they are not subjected to PK digestion prior to detection.