Figure 2. LAG3⁺ Treg regulate B-cell functions through Fas.

(a–d) Treatment of MRL/lpr mice with adoptive transfer of T-cell subsets. Ten-week-old MRL/lpr mice were injected i.v. with LAG3⁺ Treg (n=9), CD4⁺CD25⁻LAG3⁻ T cells (LAG3⁻ T; n=9), CD4⁺CD25⁺ Treg (CD25⁺ Treg; n=9) or CD4⁺CD25⁻CD45RB^high T cells (naive T; n=8) from MRL/+ mice (1 × 10⁵ cells each). The mice of LAG3⁺ Treg x3 group (n=8) were injected with LAG3⁺ Treg (1 × 10⁵ cells) at 10 weeks of age followed by a twice weekly injection of the same amount of LAG3⁺ Treg. The control group received PBS (n=13). (a) Proteinuria progression. *P<0.05 versus control group (Mann–Whitney U-test). (b) Quantification of serum anti-ds DNA antibodies. *P<0.05 (Bonferroni post-test). (c) Haematoxylin and eosine (H&E) staining (upper panels) and IgG immunofluorescent staining (lower panels) of kidney sections. Scale bars, 50 μm. (d) Glomerular scores. *P<0.05 (Mann–Whitney U-test). (e) LAG3⁺ Treg-mediated suppression of in vitro NP-specific antibody responses. B cells and Th cells purified from NP-OVA/alum-pre-immunized B6 mice and OT-II mice, respectively, were incubated with or without LAG3⁺ Treg from non-immunized OT-II mice in the presence or absence of anti-FasL blocking antibody, and supernatants were analysed for anti-NP-BSA antibodies using ELISA. See also Supplementary Fig. 1e (n=6 per group). *P<0.05 (Bonferroni post-test). (f) NP-specific antibody responses of Rag1KO mice injected with B6 B cells and OT-II Th cells with or without LAG3⁺ Treg from WT, B6/lpr or B6/gld mice, as outlined in Fig. 1e (n=6 per group). Anti-FasL blocking antibody (200 μg per mouse) was injected i.v. weekly. *P<0.05 (Bonferroni post-test). (g–i) Flow cytometry plots (g) and quantification of splenic CD4⁺CD25⁻CXCR5⁺PD-1⁺ T_FH (h) and B220⁺GL-7⁺Fas⁺ GCB (i) from the same mice as in f. Statistical significances in h,i were analysed by Bonferroni post-test (*P<0.05). The experiments in e,f were repeated three times. The means±s.d. are indicated.