Fig 2: Blimp1-deficient Tregs display impaired CNS Treg features and are insufficient to regulate CNS CD11b+ cells. a Expression of indicated proteins in FoxP3+ Tregs from the spleen (SP), spinal cord (SC), and brain (BR) of mice at d20 post-EAE induction, as in Fig. 1a. b MFI of each protein expressed in SC Tregs in a. c Splenic YFP+(FoxP3+) Tregs and CNS CD45+CD11b+ cells were sorted from MOG-immune mice and were co-cultured in triplicates with or without IL-33 for 40 h. Histogram overlays of Arg-1 and iNOS in CD11b+ cells. d Frequencies of Arg-1+ or iNOS+ CD11b+ cells in c. e Histogram overlays of GM-CSF in Tregs in C. Bottom, frequencies of GM-CSF+ Tregs. WT, FoxP3YFP-Cre; KO, Prdm1fl/flFoxP3YFP-Cre. Data represent one of two experiments (b n = 5/group). *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 (b, unpaired two-tailed Student’s t test; d, e, one-way ANOVA post hoc Tukey’s multiple comparisons test). Bars, mean ± SEM

Fig 3: GC function is required for cross-strain protective humoral responses.a Top panel indicates the design of the BALF transfer experiment for protective antibody in Bcl6-deficient mice. Mice intranasally administered with BALF or serum (60 µg/ml anti-Narita IgG) from WT (white), Bcl6?T (black) or Bcl6?B (red) mice infected with A/Narita/1/2009 were challenged intranasally with a lethal dose of A/Narita/1/2009 (left) (BALF WT n = 3, BALF Bcl6?T n = 4, Bcl6?B n = 4, Serum WT n = 3, Bcl6?T n = 5, Bcl6?B n = 6) or A/PR/8/1934 (right) (WT n = 8, Bcl6?T n = 10, Bcl6?B n = 4, Serum WT n = 4, Bcl6?T n = 5, Bcl6?B n = 6). The data indicate body weight change. Missing symbols mean that the mice were euthanized, as described in Fig. 1b. b A/H1N1pdm09-HA- and A/PR/8/1934HA-specific IgG1 and IgG2 titers in BALF from WT (white bars), Bcl6?T (black bars), and Bcl6?B (red bars) mice (n = 7 each) at 14 days post-infection with A/Narita/1/2009 were measured by ELISA. c The data indicate single or dual-binding B cells in the B220+IgM-IgD-Dump- population from MLN of WT (white bars, n = 4), Bcl6?T (black bars, n = 5), and Bcl6?B (red bars, n = 7) mice. UI represents unimmunized. Bars represent mean; *,#p < 0.05, **,##p < 0.01 by unpaired, two-tailed t-test (WT group with other groups respectively). All error bars represent SEM.

Fig 4: Selection of broadly reactive IgG is regulated differently from strain-specific IgG.a B220+IgM-IgD-Dump- B cells were obtained from spleen of the mice immunized with inactivated A/Narita/1/2009 (Vaccine, white bars; n = 3) and from MLN in the A/Narita/1/2009 infected mice (Infection, black bars; n = 3). Binding to A/Narita/1/2009-HA and/or A/PR/8/1934-HA was determined by flow cytometry analysis. b The data indicate binding to HA in flow cytometry analysis (left) and the number of single or dual-binding B cells (right) in spleen (white bars) or MLN (black bars) from the mice infected with A/Narita/1/2009 (n = 5). c Narita-HA binding B cells (2000 cells each) were isolated from vaccinated mice (spleen, white bar; n = 3) and infected mice (MLN, black bar; N = 3). The V-D-J combinations are shown by connected color bands between V and D (left), V and J (middle), and D and J (right). d Vh gene usage of A/Narita/1/2009-HA binding B cells in spleen from three vaccinated mice (Vaccine n = 3) or MLN (Infection n = 2) from two infected mice with A/Narita/1/2009. Total RNA isolated from sorted A/Narita/1/2009-HA binding B cells served as the template for cDNA synthesis followed by PCR to amplify IgG variable region genes. Amplified IgG gene libraries were sequenced by MiSeq. The IgG sequences were analyzed with the ImMunoGeneTics (IMGT) HighV-QUEST. e Seven hundred twenty HA probe+ B cells were single sorted and cultured on NB21.2D9 feeder cells with IL-4 in individual wells. After 20 days of culture, 303 IgG-producing cells were obtained. The binding of antibodies produced by these cells to A/Narita/1/2009-HA, A/PR/8/1934-HA, A/H1N1pdm09 derived LAH (blue), or rHA2 (red) was measured by ELISA. Finally, reliable binding and VH genetic information were obtained for 101 clones. Black circles indicate clones showing no binding to LAH or rHA2. Y-axis represents the ratio of the OD450 value in the sample vs the OD450 value obtained with IgG [anti-Narita-HA (1 µg/ml), anti-PR8-HA (1 µg/ml), anti-LAH (1 µg/ml), and rHA-2 (3 µg/ml)]. f Amino acid sequences encoded by Vh genes in single or dual-HA binding B cell clones in the region from CDR1 to CDR3 are shown with germline amino acid sequence (unhighlighted). Red boxes indicated CDR regions. The positions where amino acid substitutions occur are indicated by the red font. g Number of amino-acid mutations in the region extending from CDR1 to CDR3 of single or dual-binding B cell clones (single, blue dots; n = 18, dual, red dots; n = 31) as determined by referring to the germline sequences of the corresponding Vh in the IMGT database. *p < 0.05, **p < 0.01 by unpaired, two-tailed t-test (Vaccine group with Infection group in a and c, Spleen group with MLN group in b, Single group with Dual group in g). All error bars represent SEM.

Fig 5: Cathepsin B-deficient mice have abnormal follicle architecture.a Analysis of lymph node presence and morphology from WT and Ctsb-/- lymph nodes. b Percentage of B cells, CD4+ and CD8+ T cells in WT and Ctsb-deficient LNs determined using flow cytometry, with significance assessed using a Student’s t test. c Staining of WT and Ctsb-/- LNs with anti-B220 (B cells), anti-Podoplanin (Stroma), anti-CD4 (T cells) and anti-CD21/35 (follicular dendritic cells). d Staining of WT and Ctsb-/- LNs for CD19 (B cells) and Meca-79 (PNAd+ HEVs). e Entry of CFSE transferred WT B cells into the LN parenchyma of either WT or Ctsb-/- recipient mice was assessed by confocal microscopy. f Ratio of LN entry of KO:WT B cells into either WT or Ctsb-/- recipients. To determine the relative efficiency of WT vs Ctsb-/- B cells to enter into WT or Ctsb-/- recipients, equal numbers of CSFE (ThermoFisher)-labeled KO cells and CMTMR (ThermoFisher)-labeled WT cells were transferred into corresponding recipient mice. The ratio of transferred B (B220+) cells KO:WT was calculated by taking into account the relative efficiency of CFSE and CMTMR labeled survival post transfer by calculating the ratio of WT CSFE:WT CMTMR transferred cells. g Quantification of migrated CSFE-positive B cells by flow cytometry. h Analysis of Cxcl13 and Cxcr5 mRNA expression from total LN from WT and Ctsb-deficient mice using RT-qPCR. For panels (f-h), significance was assessed using a Student’s t test with p values provided for each comparison. Source data are provided as a Source Data file.