TCF1 and LEF1 promote B-1a cell homeostasis and regulatory function

Mice

Tcf7^flox/floxLef1^flox/flox mice were generated by H.-H.X. (Center for Discovery and Innovation)^20,61. Mb1^Cre mice, Rag1^−/− and CD45.1 mice were maintained on a C57BL/6 background and housed in specific pathogen-free conditions at the Australian National University (ANU) Bioscience Facility. Tcf7^flox/floxLef1^flox/flox mice were backcrossed at least six times onto the C57BL/6 background, then were subsequently crossed to Mb1^Cre mice to generate conditional Tcf1;Lef1-knockout mice. Mice were used from 8 to 12 weeks, and they were age matched when comparing the effect of two genotypes, except for ageing assessment (4–21 weeks). The Gramd2a-T2A-iCre-P2A-mApple strain was generated by the Genetic Modification Service at the Francis Crick Institute. Following adeno-associated virus (AAV) induction of a packaged rAAV donor molecule CRIPSR–Cas9 reagents were electroporated into fertilized single-cell C57BL/6 zygotes. The donor vector contained an in-frame T2A iCre P2A mApple cassette flanked by 5′ and 3′ regions of homology (400 bp, respectively) designed to insert the donor template in-frame immediately upstream of the Gramd2a stop codon in exon 12, such that expression of iCre and mApple was controlled by the endogenous Gramd2a machinery. The AAV donor vector was synthesized and packaged into AAV serotype 1 by VectorBuilder. The guide sequence used was 5′-ATGTGCAGGTAACAGTCACT-3′ and this was synthesized as a synthetic guide RNA by IDT. Mouse procedures were approved by the ANU’s Animal Experimentation Ethics Committee or UK Home Office under project licence (PP2867252). Ets1^flox/flox mice crossed more than 12 generations onto a C57BL/6 background were obtained from B. Kee at the University of Chicago. CD19–Cre mice on a C57BL/6 background were obtained from Jackson Laboratories and crossed to Ets1^flox/flox mice to generate knockout CD19–Cre Ets1^flox/flox (Ets1^Δ)⁶² and control CD19–Cre (Ets1^WT) mice. Mice were used at both 9 weeks and 20 weeks of age to compare phenotypes. Mouse procedures involving Ets1^flox/flox mice were approved by the University at Buffalo Institutional Animal Care and Use Committee.

Human samples

All patients and healthy controls provided written informed consent. Patients with confirmed bacterial pleural infection were invited to donate samples (Extended Data Table 1); all patients had pleural neutrophil counts of more than 10 million per litre, a neutrophil:lymphocyte ratio of more than 4 and were on antibiotic treatment. All clinical specimens (pleural fluid and blood) were collected for Oxford Radcliffe Pleural Biobank (ethical approval reference: 19/SC/0173, South Central-Oxford C Research Ethics Committee) and the study was approved by the Oxford Radcliffe Biobank Tissue Access Committee (reference number: 22/A093). All donors provided written informed consent.

Patients and healthy donors were also recruited following favourable study approval on 4 March 2024 by the London-Brent Research Ethics Committee (REC reference 24/PR/0218, IRAS 330505) and as part of the CAPTURE (NCT03226886) study, a prospective longitudinal cohort study investigating COVID vaccine responses in patients with cancer (Extended Data Table 2). CAPTURE is sponsored by the Royal Marsden Hospital and was approved as a substudy of TRACERx Renal (NCT03226886) by the National Research Ethics Service Committee London, Fulham on 1 May 2020 (REC 11/LO/1996). Baseline samples before COVID vaccination from patients with untreated CLL were selected. Collected as part of the EXACT study, a prospective longitudinal cohort study investigating response to immune checkpoint inhibitor therapy in patients with cancer. EXACT is sponsored by the Royal Marsden Hospital and was approved by the National Research Ethics Service Committee, West Midlands-Black Country on 22 November 2021 (REC/WM/0251). A baseline sample from one patient with untreated CLL and before immune checkpoint inhibitor therapy was selected.

Patients and healthy donor peripheral blood samples (Extended Data Table 3) were obtained upon written informed consent, according to protocols NCT03394053 (www.clinicaltrials.gov) approved by the Institutional Review Boards of Children’s National Hospital and of the National Institutes of Health and 93-I-0119, approved by the Institutional Review Board of the National Institutes of Health.

Adenoid tissue samples were collected from children (2–6 years of age, two girls and four boys) undergoing elective adenoidectomy surgery for either otitis media with effusion (OME) or sleep disordered breathing (SDB) at the Great North Children’s Hospital. The parents of all donors provided written consent for tissue donation for research and this project was approved by Newcastle University Biobank (project code: NRTB-1).

Ethical approval to obtain blood from healthy individuals was provided by the London-Brent Regional Ethics Committee (REC: 21/LO/0682).

Human sample processing

Pleural fluid and blood specimens were transferred and processed in the laboratory within hours after collection. Pleural fluid specimens were centrifuged at 800g for 10 min. The acellular supernatant was removed and stored in −80 °C. Then, the cellular pellet was resuspended in red blood cell lysis buffer (5–10 ml; J62990.AK, Thermo Scientific) and incubated for 5 min. The sample was centrifuged at 500g for 5 min and the supernatant was discarded. If there were red blood cells, the process was repeated, otherwise the cells were washed in 5 ml of PBS and centrifuged at 500g for 5 min. The supernatant was discarded, and the cells were resuspended in 5 ml of RPMI enriched with 10% FBS, passed through a 70-µm filter and counted. Whole blood was collected in EDTA tubes (VWR) and stored at 4 °C until processing or processed freshly. All samples were processed within 24 h. Before processing, tubes were brought to room temperature. Peripheral blood mononuclear cells and plasma were isolated by density-gradient centrifugation using centrifugation tubes (SepMate, StemCell) per the manufacturer’s instructions. Peripheral blood mononuclear cells either proceeded to be stained with antibodies directly or resuspended in recovery cell culture freezing medium (Fisher Scientific) containing 10% DMSO, placed overnight in CoolCell freezing containers (Corning) at −80 °C and then transferred to liquid nitrogen for long-term storage. Adenoid tissue samples were collected in PBS and immediately mechanically dissociated into a single-cell suspension using scalpel blades and a 100-µm filter. Cells were then cryopreserved in freezing media (10% DMSO and 90% FBS) in liquid nitrogen before flow cytometry experiments.

Flow cytometry

Single-cell suspension from mouse peritoneal cavity, spleens and bone marrow were treated with TruStain FcX rat anti-mouse CD16/32 antibodies (101320, BioLegend). The cells were then stained with primary antibodies used for mouse samples along with LIVE/DEAD Fixable Aqua Dead Cell strain (Invitrogen) or eBioscience Fixable eFluor780 viability dye (Invitrogen). The primary antibodies included: B220-BUV737 (RA3-6B2; 612838, BD Horizon), CD19-BV605 (6D5; 115540, BioLegend), CD23-BV421 (B3B4; 101621, BioLegend), CD5-APC (53-7.3; 17-0051-82, eBioscience), IgM-PE-Cy7 (II/41; 25-5790-82, eBioscience), CD19-BUV395 (1D3; 563557, BD Horizon), CD3-BV650 (17A2; 100229, BioLegend), CD21-BV605 (7G6; 747763, BD), IgD-PerCP/Cyanine5.5 (11-26c.2a; 405710, BioLegend), CD93-BV480 (AA4.1; 746239, BD), CD93-PE/Cyanine7 (AA4.1; 136506, BioLegend), CD24-Pacific Blue (M1/69; 101820, BioLegend), CD43-BV605 (S7; 563205, BD), CD3-APC-Cy7 (17A2; 100222, BioLegend), CD4-APC-Cy7 (GK1.5; 100414, BioLegend), CD11b-APC-Cy7 (M1/70; 101226, BioLegend), CD11b-AF700 (M1/70; 101222, BioLegend), TER119-APC-Cy7 (TER-119; 560509, BD), Gr1-APC-Cy7 (RB6-8C5; 108424, BioLegend), Sca-1-BV421 (Ly-6A/E; 108128, BioLegend), Kit-APC (2B8; 105812, BioLegend), CD127-PE-Cy7 (A7R34; 135014, BioLegend), CD16/32-PerCP/Cyanine5.5 (93; 101324, BioLegend), CD135-PE (A2F10; 135306, BioLegend), CD45.1-BV711 (A20; 110739, BioLegend), CD45.2-BUV737 (104; 612779, BD), CD274-BV711 (B7-H1; 124319, BioLegend), CD86-PE-Cy7 (GL-1; 105014, BioLegend), FCRL5-AF488 (FAB6757G, biotechne), CD223(LAG-3)-PE (C9B7W; 125208, BioLegend), CD138-BV605 (281-2; 142515, BioLegend), CD138-PE (281-2; 142504, BioLegend), CD43-BV711 (S7; 740668, BD), CD43-BV605 (S7; 747726, BD), CD1d-PerCP/Cy5.5 (1B1; 123513, BioLegend), Ig light chain κ-AF700 (RMK-45; 409508, BioLegend), Ig light chain λ-FITC (R26-46; 553434, BD), IgG3-biotin (RMG3-1; 406803, BioLegend), streptavidin-BV605 (405229, BioLegend), BLNK phosphorylated at Tyr84 (J117-1278; 558442, BD), ERK1/2 phosphorylated at Thr202/Tyr204 (20A; 561991, BD), phosphorylated PLCγ2 (K86-1161; 560134, BD), SYK phosphorylated at Tyr525/526 (C87C1; 6485S, Cell Signaling Technology), BTK phosphorylated at Tyr551 (M4G3LN; 12-9015-42, eBioscience), Ki-67-PerCP-eFluorTM710 (SolA15; 46-5698-82, eBioscience), MYC-AF647 (Y69; ab190560, abcam), TCF1/TCF7-AF488 (C63D9; 6444S, Cell Signaling Technology) and LEF1-AF488 (C12A5; 8490S, Cell Signaling Technology). DOPC/CHOL/fluorescein-DHPE (54:45:1) and rhodamine-DH PE liposomes (DOPC/CHOL/rhodamine-DHPE(54:45:1)) were used (F60103F2-R, FormumMax).

For cytokine intracellular staining, the cells were stimulated with phorbol 12-myristate 13-acetate (PMA), ionomycin, brefeldin A and LPS for the terminal 5 h of culture. Cells were harvested and stained for surface markers, including ebioscience Fixable eFluor780 viability dye (Invitrogen) to exclude dead cells before cells were fixed. Intracellular staining was performed with the Cytofix/Cytoperm kit (BD) with IL-10-PE (JES5-16E3; 505008, BioLegend) and IL-10-BV421 (JES5-16E3; 563276, BD) as recommended. For transcription factors, the eBioscience FOXP3/transcription factor staining buffer set (Invitrogen) was used per the manufacturer’s instructions. For staining phosphorylated BCR signalling components, peritoneal cells were isolated using ice-cold FACS buffer (2% FBS in PBS) and stained on ice with CD19, B220 and CD5 antibodies, followed by immediately being fixed with pre-warmed 1.6% paraformaldehyde for 10 min at 37 °C. Cells were then permeabilized with PERM III buffer (558050, BD) for 30 min on ice and subsequently stained with phosphorylated BCR signalling components for 40 min at room temperature.

Human samples were treated with human TruStain FcX CD16, CD32 and CD64 antibodies (BioLegend) and then stained with the following anti-human antibodies: CD3-BV785 (HIT3a; 740961, BD Bioscience), CD19-BUV737 (SJ25C1; 612756, BD Bioscience), CD19-PerCP/Cyanine5.5 (SJ25C1; 363016, BioLegend), CD27-PECy7 (M-T271; 560609, BD Bioscience), CD38-BV605 (HIT2; 569699, BD Bioscience), CD43-BV510 (1G10; 563377, BD Horizon), CD43-FITC (1G10; 555475, BD), CD38-BV510 (HIT2; 303540, BioLegend), CD5-APC (UCHT2; 300612, BioLegend), CD5-PE/Cyanine7 (UCHT2; 300622, BioLegend), IgA-PerCP-Vio700 (IS11-8E10; 130-113-478, MACS), CD24-BV711 (ML5; 563401, BD), IgM-BUV395 (G20-127; 563903, BD), TCF1/TCF7 (C63D9; 2203, Cell Signaling Technology), LEF1 (EPR2029Y; ab137872, Abcam), Alexa Fluor-488 donkey anti-rabbit IgG (minimal x-reactivity Poly4064; 406416, BioLegend) and Alexa Fluor-647 donkey anti-rabbit IgG (minimal x-reactivity Poly4064; 406414, BioLegend). Flow cytometers (LSRFortessa X-20, FACSAria II and LSR II; BD) and software (CellQuest and FACSDiva; BD) were used for the acquisition of flow cytometric data, and FlowJo software (Tree Star) was used for analysis.

Generation of bone marrow and fetal liver chimeras

To generate fetal liver chimera, 1 × 10⁶ fetal liver cells from Tcf7^+/floxLef1^+/floxCre^Mb1 or Tcf7^flox/floxLef1^flox/floxCre^Mb1 mice (at embryonic day 14.5) were transferred intravenously into sublethally irradiated (500 rad) Rag1^−/− recipients. To generate bone marrow chimeras, 2 × 10⁶ bone marrow- derived haematopoietic stem cells from 8-week-old mice with aforementioned genotypes were transferred intravenously into sublethally irradiated Rag1^−/− recipients. Mice were given Bactrim in their drinking water for 48 h before injection and for 6 weeks after injection and housed in sterile cages. After 6 weeks of reconstitution, mice were taken down for phenotyping by flow cytometry.

Adoptive cell-transfer experiments

Equal numbers (1 × 10⁵ cells) of flow-cytometry-sorted peritoneal CD19⁺B220⁻CD5⁺CD23⁻ B-1a cells from Tcf7^+/+Lef1^+/+ (CD45.2⁺) or Tcf7^flox/floxLef1^flox/floxCre^Mb1 (CD45.2⁺) and wild type (CD45.1⁺) were adoptively transferred into Rag1^−/− recipients intraperitoneally. The frequency of donor B cells among total peritoneal and splenic B cells was analysed at 2 months after transfer.

Peritoneal cell transfer and in vivo BrdU incorporation assay

Peritoneal cells were harvested from Tcf7^+/+Lef1^+/+ (CD45.2⁺) or Tcf7^flox/floxLef1^flox/floxCre^Mb1 (CD45.2⁺) donors by injecting 5 ml of serum-free, DMEM medium without l-glutamine (Gibco) into the peritoneal cavity and withdrew as much fluid as possible, followed by centrifuging to harvest the cells; 3 × 10⁶ cells were then injected intraperitoneally into the CD45.1 recipients. The recipients were fed with water containing 0.8 mg ml⁻¹ BrdU for 12 days. Mice were then killed; the frequency of BrdU-positive cells among indicated donor cells in the peritoneal cavity was analysed using a BrdU staining kit (8811-6600-42, Thermo Fisher) and analysed by flow cytometry.

ELISA

Flat-bottom 96-well ELISA plates (3855, Thermo Fisher) were coated with goat anti-mouse κ-UNLB (1050-01, Southern Biotech) overnight. The plates were subsequently washed and blocked using 1% bovine serum albumin in 1× PBS for 1.5 h at 37 °C. Serial dilution of the mouse serum was added to the wells and incubated overnight at 4 °C. Targeted antibodies were detected using AP-conjugated goat anti-mouse IgM (1020-04, Southern Biotech) and AP-conjugated goat anti-mouse IgG3 (1100-04, Southern Biotech). Plates were developed using 1 mg ml⁻¹ phosphatase substrate tablets (S0942, Sigma-Aldrich), and the absorbances at 405 nm and 605 nm were measured using an Infinite 200 PRO plate reader (Tecan) equipped with i-control (v1.9) software. Serum was added in serial dilution before the addition of secondary antibody: alkaline phosphatase (AP)-conjugated goat anti-mouse IgM (1020-04, Southern Biotech) and AP-conjugated goat anti-mouse IgG3 (1100-04, Southern Biotech).

Flow cytometric analysis of intracellular IL-10 synthesis

In brief, isolated leukocytes or purified cells were resuspended (1 × 10⁶ cells per millilitre) with LPS (10 μg ml⁻¹), PMA, ionomycin and brefeldin A (1:500; BioLegend) for 5 h. For IL-10 detection, cells were stained with surface markers followed by fixation, permeabilization with the Cytofix/Cytoperm kit (BD Bioscience) and staining with IL-10-PE (JES5-16E3, BioLegend) or IL10-BV421 (JES5-16E3, BD) according to the manufacturer’s instructions.

RNA-seq library preparation and data analysis

Total RNA was purified from sorted peritoneal B-1a cells (CD19⁺CD3⁻B220⁻CD5⁺CD23⁻7AAD⁻) using the PicoPure RNA isolation kit. Library construction and sequencing were performed in the Biomolecular Resource Facility, the John Curtin School of Medical Research, ANU. The single end reads of 76-bp sequencing were generated on a HiSeq2000 machine with a depth of more than 30 million reads per sample. The raw reads were aligned to the mm10 (GRCm38) genome assembly using hisat2 (ref. ⁶³) and the mapped reads were assigned with FeatureCounts (v2.4)⁶⁴ based on the genome-build GRCm38.p4 annotation and NCBI Refseq gene mode by removing ribosomal genes and non-coding RNA, respectively. Differential expression analyses were performed with voom-limma⁶⁵, after removal of lowly expressed genes and normalized using the trimmed mean of M-values method^66,67. Significantly differentially expressed genes were identified by applying a Benjamini–Hochberg adjusted P value threshold of 0.05 (ref. ⁶⁸). Gene set enrichment or pathway analysis were performed using clusterProfiler⁶⁹ and Camera⁷⁰ against the Gene Ontology database, KEGG database and HALLMARK C2 and C7 gene sets in the MSigDB (v7.5).

scRNA-seq

Peritoneal B cells (CD19⁺CD3⁻7AAD⁻) from three 8-week-old mice, B-1Ps (Lin⁻CD93⁺IgM⁻CD19⁺B220^low/−) from embryonic day 18.5 fetal livers (three fetal livers were pooled together as one replicate) or bone marrow from three 8-week-old mice, and B-1 cells (CD19⁺B220^low/−) from the peritoneal cavity and spleen from three 8-week-old mice were sorted by the FACS Aria II cell sorting system (BD Immunocytometry Systems). Cells (n = 10,000–20,000) per sample were run on the 10X Chromium platform (10X Genomics). Library preparation and sequencing were performed by the Australian Cancer Research Foundation Biomolecular Resource Facility, the John Curtin School of Medical Research, ANU or Genomics STP, The Francis Crick Institute according to the manufacturer’s instructions for the Chromium Next GEM Single Cell 5′ Kit v2 or v3. Two libraries were generated and the mRNA transcript expression and BCR repertoire were measured. The samples were sequenced using the NovaSeq 6000 (Illumina) system or NovaSeq S2. The FASTQ files were aligned to the mm10 mouse reference genome using the 10X Genomics CellRanger pipeline (v6.0.1 or v7.0.1). Data were processed and analysed using the Seurat package (v5.0.3) in R (v4.4.1). Raw gene expression matrices were filtered to exclude low-quality cells based on the thresholds for mitochondrial gene content (set at 5%) and the number of detected genes per cell, using the PercentageFeatureSet() and subset() functions.

For multiplexed samples labelled with cell hashing antibodies, demultiplexing was performed using the MULTIseqDemux() function in Seurat. Hashtag oligonucleotide counts were normalized using centred log-ratio (CLR) transformation via the NormalizeData() function with normalization.method = ‘CLR’, and cells were classified as singlets, doublets or negatives based on the hashtag oligonucleotide signal. Only singlets were retained for downstream analysis.

Normalization and integration were performed using the scTransform() workflow, followed by the identification of highly variable features. Dimensionality reduction was performed using principal component analysis via the RunPCA() function, and the top 20–30 principal components were selected to construct a shared nearest neighbour graph using FindNeighbours(). Clustering was conducted using the FindClusters() function. Clusters were visualized using UMAP via the RunUMAP() function. Differential expressed genes for each cluster were identified using the FindAllMarkers() function with test.use set as either ‘wilcox’ or ‘MAST’. Where applicable, cell types were annotated based on canonical marker expression or informed by publicly available datasets.

Cell trajectory analysis

Cell trajectory analysis was performed using Monocle2 (v2.32.0) in R (v4.4.1). The annotated Seurat object was first transformed into a CellDataSet object using the as.CellDataSet() function from the Seurat package. To filter relevant genes, those expressed in at least 10 cells and with a mean expression value of 0.1 or greater were selected. Next, highly variable genes across cell subsets were identified using the differentialGeneTest() function with the formula fullModelFormulaStr = ‘~cell_subset’. On the basis of their q-value, the top 2,000 genes were selected, and setOrderingFilter() was applied to prioritize them for ordering in pseudo-time analysis. Dimensionality reduction was performed using reduceDimension() with the parameter reduction_method = ‘DDRTree’. Cells were subsequently ordered along the trajectory using the orderCells() function and visualized with plot_cell_trajectory(). All other parameters were set to their default values.

Single-cell BCR sequencing analysis

BCR repertoire analysis was conducted with the Immcantation framework, following the guidelines provided by its developers (http://immcantation.org/). This was implemented in a Python environment (v3.11.10). We began with 10X Genomics BCR sequencing data processed through the CellRanger pipeline (v7.0.1) to obtain annotated clonotype and contig sequences. For V(D)J gene assignment, the AssignGenes.py command from the Change-O toolkit was used, referencing the mouse IMGT and IgBLAST (v1.22.0) databases. MakeDb.py was used to standardize the data to AIRR format. Cells expressing more than one heavy chain or only light chains were removed. Clonotypes were assigned using the DefineClones.py tool based on heavy chain sequences, using the default Hamming distance substitution model. Subsequently, light_cluster.py was applied to refine clonotype grouping by incorporating light chain information. Finally, the resulting datasets were merged with the annotated Seurat file and key metrics, including V(D)J gene usage, CDR and FWR sequences, lengths and mutational content were extracted for downstream analysis and visualization.

For diversity analysis, we used the iNEXT package (v3.0.1) to perform coverage-based rarefaction and extrapolation analyses. We computed the three most common Hill numbers: species richness (q = 0), Shannon diversity (q = 1) and Simpson diversity (q = 2). The resulting sampling curves, which illustrate diversity estimates with respect to sample coverage, were visualized using ggiNEXT() function. In addition, we constructed circle packing plots using the packcircles package (v0.3.6) to visualize the immune repertoire composition of individual samples. These analyses were conducted in R (v4.4.1).

scRNA-seq analysis on human prenatal B-1 cluster

Processed and annotated scRNA-seq data of fetal immune cells, specifically the haematopoietic stem and progenitor cells and the B cell lineages, in Suo et al.²³ were downloaded from the developmental cell atlas portal (https://developmentcellatlas.cellgeni.sanger.ac.uk/fetal-immune/lymphoid/). The mean expression of Tcf7 and Lef1 in the cell types of the above-mentioned lineages are represented as dot plots (Fig. 3h and Extended Data Fig. 6f). The B-1 cells were classified into four clusters based on the expression of LEF1 and TCF7. The total B-1 cells expressing LEF1 exclusively is 463, TCF7 exclusively is 325, both is 73 and neither LEF1 nor TCF7 is 4,789 (Extended Data Fig. 6e). All the analyses were performed using scanpy (v1.9.8).

ChIP–seq analysis

FASTQ files were downloaded from the NCBI’s Gene Expression Omnibus database for TCF1 and LEF1 chromatin immunoprecipitation followed by sequencing (ChIP–seq) data (SRP142342)⁴². FASTQ files were aligned to Enembl’s mouse GRCm38 genome using BWA (v0.7.15). The resulting BAM files were sorted, duplicates marked and indexed using Picard (v2.1.1). Peaks were called using MAC2 (v2.1.1) that were enriched in TCF1 or LEF1 relative to input using default parameters. Peaks were annotated using HOMER (v4.8). BAM files were normalized to 10 million reads, and IGVTools (v2.3.75) was used to generate coverage files.

ATAC-seq analysis

FASTQ files were downloaded from the NCBI’s Gene Expression Omnibus database for ATAC-seq samples in B-1a cells (GSM2461745)¹⁷. FASTQ files were aligned to Ensembl’s mouse GRCm38 genome using BWA (v0.7.15). The resulting BAM files were sorted, duplicates marked and indexed using Picard (v2.1.1). IGVTools (v2.3.75) was used to generate coverage file for visualization.

In vitro stimulation

Cells were cultured in RPMI 1640 medium supplemented with 10% FBS, 2 mM l-glutamine, 100 U penicillin–streptomycin, 0.1 mM non-essential amino acids, 100 mM HEPES and 55 μM 2-mercaptoethanol at 37 °C in 5% CO₂. B-1 cells were magnetically purified from the peritoneal cavity using the Pan B cell isolation kit II (130-104-443, Miltenyi Biotec) with anti-mouse CD45R (B220) antibody (130-110-707, Miltenyi Biotec) and labelled with CellTrace Violet (C34557, Thermo Fisher) followed by stimulating with or without 5 µg ml⁻¹ LPS (O111:B4; L4391, Sigma), 0.4 µg ml⁻¹ IL-10 (210-10-10UG, Thermo Fisher), 10 µg ml⁻¹ InVivoMAb anti-mouse IL-10R (CD210; BE0050, BioXCell) and its isotype control, InVivoMAb rat IgG1 isotype control, anti-horseradish peroxidase (HRP; BE0088, BioXCell) for 72 h.

Cytometric bead array

B-1 cells were magnetically purified from the peritoneal cavity using the Pan B cell isolation kit II (130-104-443, Miltenyi Biotec) with anti-mouse CD45R (B220) antibody (130-110-707, Miltenyi Biotec) and cultured in RPMI 1640 medium supplemented with 10% FBS, 2 mM l-glutamine, 100 U penicillin–streptomycin, 0.1 mM non-essential amino acids, 100 mM HEPES, 55 μM 2-mercaptoethanol and 5 µg ml⁻¹ LPS (O111:B4; L4391, Sigma) for 3 days at 37 °C in 5% CO₂. Supernatant was collected from cell culture, and IL-10 production was measured by the cytometric bead array mouse IL-10 Flex set (BD).

EAE

EAE was induced by immunization with MOG_35–55 peptide, emulsified in Complete Freund’s adjuvant (Sigma-Aldrich) and pertussis toxin (Sigma-Aldrich). On day 3.5, 2 × 10⁵ B-1 cells that had been previously activated with R848 (0.1 μg ml⁻¹) for 48 h were transferred intravenously. Mice were examined daily and scored using the following scoring system: 0 for no disease, 1 for loss of tail tonicity, 2 for hindleg weakness, 3 for complete hindleg paralysis, 3.5 for complete hindleg paralysis with partial hind body paralysis, 4 for full hindleg and foreleg paralysis, and 5 for moribund or dead animals⁷¹.

Immunohistochemistry

Sections (3 µm) were staining on the Leica Bond Rx automated staining platform using sequential application of antibodies with Opal TSA fluorophore detection followed by heat stripping of the antibody complex between stainings using ER1 or ER2 antigen retrieval solution (Leica) for 20 min at 95 °C. The following antibodies, secondaries and opal TSA reagents were used: mApple (1:100; STJ140269, St John’s Laboratory), CD45R/B220 (1:750; 553086, BD Biosciences), CD19 (1:200; ab245235, Abcam), Immpress HRP Horse anti-goat IgG polymer (MP-7405-50, Vector), streptomycin-HRP (1:500; P0397, Dako), NovoLink Max Polymer (RE7260-CE, Leica) and Opal 520/570/690 (FP1487001KT/FP1488001KT/FP1497001KT, Akoya Biosciences). Slides were counterstained with DAPI and imaged on the PhenoImager HT (Akoya Biosciences).

Statistical analyses

Statistical methods used in RNA-seq, scRNA-seq and single-cell BCR-seq are described above. Comparison between groups was performed using parametric t-test, two-tailed Welch’s t-test, Mann–Whitney t-test, one-way ANOVA with Tukey multiple-comparison test, two-way ANOVA and two-tailed Pearson correlation analysis from GraphPad Prism10 (GraphPad Software). Sample sizes are provided in the figures, and statistically significant differences are indicated as exact P values.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.