To test this hypothesis, immunized mice were treated with an agonistic anti-GITR

mAb to disrupt the suppressive activity of Treg cells.48–50 Splenic GCs persist for at least 4 weeks so the GC response was monitored at days 8, 12, 18 and 24 post-challenge. Preliminary experiments tested the effects of continuous anti-GITR mAb injections on the GC response. When injected twice weekly for up to 4 weeks, however, anti-GITR mAb administration resulted in a high death rate in immunized mice, preventing an appropriate kinetic analysis (data not shown). Similar to previous studies18,22,23,26 a three-injection PLX4032 nmr protocol was therefore used whereby 250 μg of either anti-GITR mAb or control rat IgG (rIgG) was injected on days −2, +1 and +5. Mice were immunized with SRBC on day 0 and splenic GCs were analysed during the ensuing 4 weeks. Naive mice kept in specific pathogen free conditions do not have detectable GC B cells in their spleens, as previously described1,5 and C59 wnt datasheet shown in Fig. 1(a). Upon challenge with SRBC, a robust GC response is induced and easily detected as a B220+ PNAhi population (refs. 1,5 and Fig. 1a). Using a multi-colour approach, the IgM+ (non-switched)

B cells and switched GC B cells can be further delineated (Fig. 1a). When comparing the GC response from immunized mice injected with anti-GITR mAb or rIgG, it is clear that Treg-cell disruption resulted in a higher frequency and total number of splenic B220+ PNAhi GC B cells at all time-points out examined

(Fig. 1b). As expected, the ratio of IgM+ to switched GC B cells remained steady over the course of the response in control rIgG-treated mice, even as the reaction waned (Fig. 1c). However, immunized mice treated with anti-GITR mAb exhibited a higher frequency and total number of IgM− switched GC B cells at day 8, an imbalance which increased over time (Fig. 1c). When comparing the distribution of IgG isotypes expressed on switched GC B cells in anti-GITR mAb and rIgG treated mice, a significant increase in the percentage of IgG1+ GC B cells was observed at day 8 in the Treg-cell-disrupted group (data not shown). At all other time-points, IgG isotype expression within the switched GC pool did not differ between the two groups. Taken together, disruption of Treg cells led not only to a larger GC response, but to an inability to control the proportion of IgM+ to switched GC B cells. Given the marked changes observed in splenic GC B cells after Treg-cell disruption, the non-GC (B220+ PNAlo/neg) B-cell population was also monitored. As shown in Supplementary material, Fig. S1(A), a significant difference in the total number of non-GC B cells was observed after anti-GITR mAb treatment only at day 12 post-challenge. To assess which non-GC B-cell sub-sets were affected at day 12, a detailed analysis of follicular, pre-marginal zone, marginal zone, transitional 1 (T1), T2 and B1 B cell percentages was performed (see Supplementary material, Fig. S1B,C).