Stimulation with LPS + IL-4 promotes B cell proliferation and class switch recombination

(A-D) Cells from the lymph nodes of C57BL6/J mice were stained with Cell Trace Violet (CTV) and stimulated with IL-4 (10ng/ml) +/- LPS (20μg/ml) for 72 hours and analysed by flow cytometry. Data for naive B cells (stained on day 0) is also shown. Live CD19+ B cells were identified using the gating strategy described in Figure S7A. (A) Representative flow cytometry plots comparing IgG1 expression and CTV staining after 72 hours of IL-4 or LPS + IL-4 stimulation. (B) Representative plots for FSC and SSC at 72 hours of IL-4 or LPS + IL-4 stimulation with geometric means for the (C) forward scatter and (D) side scatter. Data shows the results of three biological replicates. Data was analysed by one-way ANOVA followed by multiple comparison testing via Sidak’s analysis, for comparisons to naïve B cells, where ****p<0.0001. Full ANOVA results are given in Supplementary Table S4. (E-H) Cells from the lymph nodes of C57BL6/J mice, stimulated with LPS (20μg/ml) and IL-4 (10ng/ml) for 24 hours, and CD19+ B cells were isolated by FACS. Alternatively, naïve CD19+ B cells were sorted directly from ex vivo lymph node cells. Cells were lysed and analysed by proteomics as described in the methods. Samples from four mice for LPS + IL-4 and three for naive were generated. (E) Total protein content (pg/cell) was estimated from the proteomic data. Statistical power was determined using an unpaired two-tailed t-test, where **p<0.01. (F) Volcano plot depicting changes in estimated protein copy number in naïve vs LPS + IL-4 stimulated B cells. (G) Volcano plot showing the estimated cellular protein concentration (μM) in naïve vs LPS + IL-4 stimulated B cells. Horizontal dashed lines indicate q < 0.05. Vertical dashed lines indicate log2 fold change of one standard deviation away from the median. (H) Enrichment analysis of the upregulated proteins in (G) against GO-term and KEGG databases.

Proteins involved in cell cycle progression are upregulated in LPS + IL-4 activated B cells

(A-B) B cells were purified from the spleens of C57BL6/J mice and either fixed on isolation (naive) or stimulated with LPS (20μg/ml) and IL-4 (10ng/ml) for 24 hours before fixation. Cells were then stained with DAPI and CD19. The cell cycle stage were analysed using the gating strategy shown in Figure S7B. (A) Representative histogramsshowing the proportion of B cells in different phases of the cell cycle. (B) Quantification shows three technical replicates from cells isolated from one mouse and is representative of three independent experiments. Statistical power was determined using two-way ANOVA followed by multiple comparison testing via Sidak’s analysis, where p<0.01 is indicated by **, p<0.001 by *** and p<0.0001 by **** for comparisons between the naive and LPS + IL-4 conditions. (C-F) Graphs depicting changes in the estimated cellular concentration (nM) of proteins implicated in entry into the cell cycle, determined from the proteomic dataset described in Figure 1. (C) Cyclin D, (D) CDK4, (E) CDK6, (F) p27Kip1. (G-H) B cells were purified from the spleens of C57BL6/J mice and stimulated with LPS (20μg/ml) and IL-4 (10ng/ml) for 24 hours before fixing and staining for phospho-retinoblastoma (p-Rb). Gating strategy for p-Rb staining is shown in Figure S7B. (G) Representative histogram comparing p-Rb staining in naïve and LPS + IL-4 stimulated B cells. (H) Quantification shows three technical replicates from cells isolated from one mouse and is representative of three independent experiments. (I) Heat map showing the expression of proteins encoded by E2F target genes, derived from the proteomic data. (J-M) Graphs depicting changes in the estimated cellular concentration (nM) of proteins implicated in cell cycle progression, determined from the proteomic dataset described in Figure 1. (J) Cyclin A, (K) Cyclin B, (L) CDK2, (M) CDK1. p values were determined using an unpaired two-tailed Student’s t-test for (H) or represent adjusted p values from the FDR calculations applied to the proteomic dataset (D, E, F, L), where p<0.01 is indicated by **, p<0.001 by *** and p<0.0001 by ****.

LPS + IL-4 stimulation promotes protein synthesis

(A, B) Graphs show changes in cellular concentration (nM) of the sum of proteins that make up the large (A) and small (B) ribosomal subunits with heat maps showing the expression of the individual proteins making up the large (C) and small (D) subunits. (E-G) The proteomic dataset was mined for proteins involved in the biogenesis of the large and small ribosomal subunits based on the GO terms: GO:0000027, GO:0042273, GO:0000028, GO:0042274. (E) shows the sum of the proteins involved in the biogenesis of the large subunit and (F) shows the small subunit, with individual proteins represented on the volcano plot (G) Horizontal dashed lines on (G) indicate q < 0.05 while vertical dashed lines indicate log2 fold change more than one standard deviation away from the median. (H-I) Splenocytes from C57BL6/J mice were stimulated with LPS (20μg/ml) and IL-4 (10ng/ml) for 24 hours before fixing and staining for the uptake of puromycin analog O-propargyl-puromycin (OPP) to measure protein synthesis. Gating strategy for OPP staining is shown in Figure S7C. (H) Representative histogram comparing OPP uptake between naïve and LPS + IL-4 stimulated B cells. (I) Quantification shows three technical replicates from cells isolated from one mouse. Statistical power was determined using an unpaired two-tailed Student’s t-test, where p<0.001 is indicated by *** and p<0.0001 by ****.

Amino acid transporter SLC7A5 is required for key B cell functions

(A) Heat map showing the expression of genes encoding for proteins involved in plasma membrane amino acid transport determined from the proteomic dataset described in Figure 1. (B, C) Graphs depicting changes in cellular concentration (nM) of (B) SLC7A5 and (C) SLC3A2 derived from the proteomic data. p(adj)<0.0001, based on the proteomic FDR calculations, is indicated by ****. (D-E) Lymph node cells from C57BL6/J mice were stimulated with LPS (20μg/ml) and IL-4 (10ng/ml) for 24 hours before staining for CD98. (D) Representative FACS plot comparing CD98 expression. (E) Quantification shows three technical replicates from cells isolated from one mouse and is representative of two independent experiments. Statistical power was determined using one-way ANOVA followed by multiple comparison testing via Dunnett’s analysis, for comparison to naïve B cells, where ***p<0.001 and ****p<0.0001. (F-I) B cells were purified from the spleens of wild type (WT) and Slc7a5 fl/fl/Vav-iCre+/- (SLC7A5 KO) mice and stained with CTV before stimulation with LPS (20μg/ml)and IL-4 (10ng/ml) for 72 hours. (F) Percentage of live B cells (7AAD-ve) in WT and SLC7A5 KO mice. (G) Histogram representing CTV staining of WT and SLC7A5 KO B cells. (H) Live B cell number of WT and SLC7A5 KO mice. (I) Percentage of B cells that are IgG1+ve in WT and SLC7A5 KO mice. Data shows the results of four biological replicates per genotype. p<0.0001 is indicated by **** (unpaired two-tailed Student’s t-test). (J-K) B cells were purified from the spleens of WT and SLC7A5 KO mice and stimulated with LPS (20μg/ml) +/- IL-4 (10ng/ml) for 24 hours before fixing and staining for the uptake of kynurenine to measure amino acid uptake. Gating strategy for kynurenine uptake in Figure S7D. (J) Quantification of kynurenine MFI between B cells from WT and SLC7A5 KO mice with (+) or without (-) LPS + IL-4, kynurenine or aminobicyclo-(2,2,1)-heptane-2-carboxylic acid (BCH). (K) Quantification of kynurenine MFI between B cells from WT and SLC7A5 KO mice with (+) or without (-) LPS, kynurenine or BCH. Data shows the results of four biological replicates per genotype. Statistical power was determined for using two-way ANOVA followed by multiple comparison testing via Sidak’s analysis, where **** indicates p<0.0001 for comparisons between genotypes.

LPS + IL-4 stimulation upregulates cholesterol metabolism in B cells

(A-E) Analysis of the genes involved in cholesterol metabolism in the proteomic dataset. (A) Heat map of all the enzymes involved in cholesterol biosynthesis. Cellular concentration (nM) of (B) HMGCR, (C) SQLE, (D) LDLR, and (E) SREBP2. A p(adj) of <0.01 is indicated by **. (F) Splenocytes from C57BL6/J mice were plated in cholesterol-free media and stimulated with LPS (20μg/ml) and IL-4 (10ng/ml). The cells were fixed at the stated time points and stained with filipin. The gating strategy for filipin staining in Figure S8A. Data shows three technical replicates from cells isolated from one mouse. Statistical power was determined by one-way ANOVA followed by multiple comparison testing via Dunnett’s analysis, where * indicates p<0.05 and **** indicates p<0.0001 for comparisons to naïve B cells. (G-H) Splenocytes from C57BL6/J mice were plated in cholesterol-free media and pre-treated with DMSO as a vehicle control or varying concentrations of Fluvastatin or NB-598 for 45 minutes before stimulation with LPS (20μg/ml) and IL-4 (10ng/ml). The cells were fixed after 24 hours and stained with filipin before acquisition. (G) Filipin staining of Fluvastatin titration. (H) Filipin staining of NB-598 titration. Data shows three technical replicates from cells isolated from one mouse and is representative of two independent experiments, each with one biological replicate. Statistical power was determined using a one-way ANOVA followed by multiple comparison testing via Dunnett’s analysis, where ***p<0.001 and ****p<0.0001 for comparison to LPS + IL-4 stimulated B cells. (I) Spleenocytes from C57BL6/J mice were plated in normal or cholesterol-free (CF) media and pre-treated with Fluvastatin (10μM) or NB-598 (10μM) before stimulation with LPS + IL-4. The cells were fixed after 24 hours and stained with filipin before acquisition. Data shows the results of three biological replicates. Statistical power was determined using two-way ANOVA followed by multiple comparison testing via Tukey’s analysis, where **** indicates p<0.0001 for comparisons to the LPS + IL-4 condition.

Blocking rate-limiting enzymes in the cholesterol biosynthesis pathway reduces B cell growth, survival and proliferation

(A-D) B cells were purified from the spleens of C57BL6/J mice and cultured in normal or cholesterol-free (CF) media. The cells were stained with CTV, then pre-treated with DMSO as a vehicle control or Fluvastatin (10µM) or NB-598 (10µM), where indicated, for 45 minutes prior to stimulation with LPS (20μg/ml) and IL-4 (10μg/ml) for 48 hours.Gating strategy for CTV staining is shown in Figure S8B. (A) shows representative CTV staining, (B) live B cell number, (C) percentage of live B cells (7AAD-ve) (D) forward scatter. Data shows the results of three biological replicates. (E-H) Same as (A-D) but with pretreatment using FGTI-2734 (10µM). (E) shows representative CTV staining, (F) live B cell number, (G) percentage of live B cells (7AAD-ve) after FGTI-2734 treatment and (H) forward scatter of B cells. Data shows the results of three biological replicates. Statistical power was determined using two-way ANOVA followed by multiple comparison testing via Tukey’s analysis, p<0.01 is indicated by **, <0.001 by *** and <0.0001 by **** for comparisons to the LPS + IL-4 condition. ns indicates p>0.05. For all panels, cells in the absence of LPS + IL-4 were naïve B cells analysed on the day of isolation.

Mevalonate supplementation rescues the effect of Fluvastatin treatment in B cells

(A-B) Splenocytes from C57BL6/J mice were plated in normal or cholesterol-free (CF) media and pre-treated with HEPES as a vehicle control or mevalonate (2mM) for 1 hour prior to treatment with DMSO or Fluvastatin (10µM), where indicated, for 45 minutes before stimulation with LPS (20μg/ml) and IL-4 (10ng/ml). The cells were fixed after 24 hours or 48 hours of LPS + IL-4 stimulation and stained with filipin. (A) Filipin staining comparing B cells +/- Fluvastatin or mevalonate after 24 or 48 hours of LPS + IL-4 stimulation in normal media. (B) Filipin staining comparing B cells +/- Fluvastatin or mevalonate after 24 or 48 hours of LPS + IL-4 stimulation in cholesterol-freemedia. Data shows the results of three biological replicates. (C-F) B cells were purified from the spleens of C57BL6/J mice and cultured in normal or cholesterol-free media. The cells were stained with CTV, then pre-treated with HEPES as a vehicle control or mevalonate (2mM) for 1 hour. The cells were treated with DMSO or Fluvastatin (10µM), where indicated, for 45 minutes prior to stimulation with LPS (20μg/ml) and IL-4 (10μg/ml) for 48 hours. (C) Shows percentage of live B cells (7AAD-ve), (D) forward scatter of B cells, (E) representative CTV staining, (F) live B cell number. Data shows the results of three biological replicates. Where shown, statistical power was determined using two-way ANOVA followed by multiple comparison testing via Sidak’s analysis. For comparison to Fluvastatin treated B cells, p<0.05 is indicated by *, <0.01 by **, <0.001 by ***, <0.0001 by **** and >0.05 by ns. For all panels, cells in the absence of LPS + IL-4 were naïve B cells analysed on the day of isolation.

GGPP supplementation rescues the effect of Fluvastatin treatment in B cells

(A-H) B cells were purified from the spleens of C57BL6/J mice and cultured in normal (A-D) or cholesterol free (CF, E-H) media. The cells were stained with CTV, pre-treated with methanol:ammonium hydroxide solution (CH3OH:NH4OH) as a vehicle control or geranylgeranyl pyrophosphate (GGPP) (10μM) for 1 hour before treatment with Fluvastatin (10μM), where indicated. The cells were stimulated with LPS (20μg/ml) and IL-4 (10μg/ml) and cultured for 48 hours. (A) shows representative CTVstaining for B cells cultured in normal media, (B) live B cell number, (C) percentage of live B cells and (D) forward scatter of B cells. (E) shows representative CTV staining for B cells cultured in CF media, (F) live B cell number, (G) percentage of live B cells and (H) forward scatter of B cells. Data shows the results of three biological replicates. Statistical power was determined using one-way ANOVA followed by multiple comparison testing via Dunnett’s analysis. For comparison to Fluvastatin treated B cells, p<0.001 is indicated by *** and p<0.0001 by ****. ns indicated p<0.05. For (B) and (F) this data was log-transformed then statistically analysed due to unequal variance. (I-J) Splenocytes were plated in normal or cholesterol-free media and pre-treated with methanol:ammonium hydroxide solution (CH3OH:NH4OH) as a vehicle control or geranylgeranyl pyrophosphate (GGPP) (10μM), where indicated for 1 hour before treatment with DMSO or Fluvastatin (10μM). The cells were then stimulated with LPS (20μg/ml) and IL-4 (10ng/ml). The cells were fixed after 24 hours and stained with filipin before acquisition. (I) Filipin staining comparing B cells +/- Fluvastatin or GGPP after 24 or 48 hours of LPS + IL-4 stimulation in normal media. (J) Filipin staining comparing B cells +/- Fluvastatin or GGPP after 24 or 48 hours of LPS + IL-4 stimulation in cholesterol-free media.Data shows the results of three biological replicates. Statistical power was determined using two-way ANOVA followed by multiple comparison testing via Sidak’s analysis. For comparison to Fluvastatin treated B cells p<0.01 is indicated by **, <0.001 by *** and p<0.0001 by ****. ns indicated p<0.05. For all panels, cells in the absence of LPS + IL-4 were naïve B cells analysed on the day of isolation.

Cholesterol is required for the growth and proliferation of B cells regardless of stimuli

(A) Splenocytes from C57BL6/J mice were plated in normal or cholesterol-free (CF) media, before stimulation with LPS (20μg/ml), IL-4 (10μg/ml) or a combination of LPS and IL-4. The cells were fixed after 24 hours and stained with filipin. Data shows the results of three biological replicates. Statistical power was determined using two-way ANOVA followed by multiple comparison testing via Tukey’s analysis. For comparisons to naïve B cells, p<0.0001 is indicated by ****. (B) Splenocytes from C57BL6/J mice were plated in normal media before stimulation with LPS (20μg/ml), Resiquimod (1μg/ml), ODN 1826 (1μg/ml), Anti-IgM (10μg/ml), or CD40L (500ng/ml). The cells were fixed after 24 hours and stained with filipin. (B) Filipin staining comparing cholesterol content between different stimuli. Data shows the results of three biological replicates. Statistical power was determined using one-way ANOVA followed by multiple comparison testing via Dunnett’s analysis. For comparisons to naïve B cells, p<0.0001 is indicated by ****. (C-E) B cells were purified from the spleens of C57BL6/J mice and cultured in normal media. The cells were stained with CTV, then pre-treated with DMSO as a vehicle control or Fluvastatin (10µM) for 45 minutes before stimulation with LPS (20μg/ml), Resiquimod (1μg/ml), ODN 1826 (1μg/ml), Anti-IgM (10μg/ml), or CD40L (500ng/ml) as indicated for 72 hours. (C) shows representative histogram for CTV staining, (D) the percentage of live B cells (7AAD-ve), (E) live B cell number. Data shows the results of three biological replicates. Statistical power was determined using two-way ANOVA followed by multiple comparison testing via Tukey’s analysis, where p<0.0001 is indicated. For (E) this data was log-transformed and then statistically analysed due to unequal variance. Statistical power was determined using two-way ANOVA followed by multiple comparison testing via Sidak’s analysis, where ****p<0.0001.

MACS antibodies.

Flow cytometry antibodies