P < 0·05 was regarded as the significant level of probability throughout. Two trials designated Experiments 5 and 6 were conducted (Table 1), so numbered as they were part of a larger series of trials sharing the same design. Both experiments contained a group of sheep which had received a trickle immunising Romidepsin solubility dmso infection of 2000 T. circumcincta infective larvae three times per week for 8 weeks,

and a group of control sheep which had not received the trickle infection. All were dosed with fenbendazole one week prior to challenge with a single dose of 50 000 infective larvae, with surgery to cannulate the gastric lymph duct being carried out on 10 sheep in each experiment during the intervening week. Sheep were killed on days 5, 10 or 21 post-challenge. It was known from prior work using this experimental model that in previously infected sheep the cellular and humoral immune responses in lymph all occurred by day 9 after challenge. Therefore, lymph collection from the previously infected lambs was stopped GS-1101 mouse after 10 days. Large cells or lymphoblasts were determined as those with a diameter of >9 μm when measured by Coulter Counter, with small lymphocytes represented as those with a diameter of between 3 and 9 μm. During FACS analysis, small cells were those appearing within region R1 on a control sample Fsc vs. Ssc plot (Figure 1),

blast cells were designated as the gated lymphocytes which fell within region R2 and total lymphocytes within R3 (=R1 + R2). Downstream

FACS analyses of stained cells were gated to contain only those cells present in R3. Surface staining of lymphocytes from gastric lymph, and flow Tyrosine-protein kinase BLK cytometry, were carried out as detailed previously (6). Monoclonal antibodies that recognise border disease virus as isotype controls (clones VPM21 (isotype IgG1, 1/500 dilution) and VPM22 (isotype IgG2, 1/500) (25)), ovine CD4 (clone 17D, IgG1, 1/1000 (26)), CD8 (clone 7C2, IgG2a, 1/1000 (27)), γδ T cell receptor (clone 86D, IgG1, 1/1000 (28)), CD25 (an activated T cell marker, clone ILA111, IgG2a, 1/2000 (29)), CD21 (a pan B cell marker, clone CC21, IgG1, 1/10 (30)) and IgA (MCA628, Serotec, Oxford, UK, IgG1, 1/1000) were used. The percentage of total cells positive for the isotype control antibodies was observed to be below 0·15% for 99·3% of all samples. Detection and quantification of antibody in the gastric lymph was carried out as detailed previously (10). Briefly, total IgA was measured using a sandwich ELISA, with purified sIgA as a standard. Antigen specific IgA was measured for both somatic L4 antigen, and L4 excretory/secretory (ES) products, with a positive reference sample included on each plate. Previously infected lambs had significantly (P < 0·05) fewer parasites than controls on day 10 after challenge in both experiments (Figure 2a). However, on day 5 a significant difference (P < 0·05) was only observed within Experiment 5.