Congenital T-cell activation impairs transitional-to-follicular B-cell maturation in humans

Abstract

Patients with cytotoxic T-lymphocyte-associated protein 4 (CTLA4) deficiency exhibit profound humoral immune dysfunction, yet the basis for the B-cell defect is not known. We observed a marked reduction in transitional-to-follicular (FO) B-cell development in patients with CTLA4 deficiency, correlating with decreased CTLA4 function in regulatory T cells, increased CD40L levels in effector CD4+ T cells, and increased mammalian target of rapamycin complex 1 (mTORC1) signaling in transitional B cells (TrBs). Treatment of TrBs with CD40L was sufficient to induce mTORC1 signaling and inhibit FO B-cell maturation in vitro. Frequent cell-to-cell contacts between CD40L+ T cells and immunoglobulin D-positive CD27- B cells were observed in patient lymph nodes. FO B-cell maturation in patients with CTLA4 deficiency was partially rescued after CTLA4 replacement therapy in vivo. We conclude that functional regulatory T cells and the containment of excessive T-cell activation may be required for human TrBs to mature and attain metabolic quiescence at the FO B-cell stage.

Graphical abstract

Figure 1.

Increased frequency of circulating TrBs and decreased frequency of circulating FO B cells in patients with CTLA4 deficiency. Flow cytometric analysis of peripheral blood B-cell populations in patients with CTLA4 deficiency (n = 11) and healthy controls (n = 16). (A) Representative plots from a healthy control (left) and a patient with CTLA4 deficiency (right, homozygous CTLA4 3′ UTR mutation carrier), shown at 5% of events with gating strategy outlined (full description of B-cell gating strategy in supplemental Figure 1A). (B) Quantification of major B-cell subsets in the peripheral blood, including naïve, marginal zone (MZ), SWM, and DN populations as subset from total CD19⁺ B cells by the markers IgD and CD27. (C) Quantification of naïve (IgD⁺CD27^–) B-cell subsets in the peripheral blood, including T1/2, T3a, and T3b TrBs, resting mature FO B cells, aN B cells, and MZP B cells as further subsets by the indicated markers. For the CD45RB by CD21 and count by CD73 B-cell subset plots, final IgD⁺CD27^– B-cell frequency was derived as a percentage of the indicated parent gate. Symbols represent unique individuals; bars represent means (± standard deviation [SD]) of all data. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001 or as listed by 2-tailed Mann-Whitney U test. MZP, MZ precursor.

Figure 2.

Reduced frequency of circulating FO B cells correlates with degree of congenital Treg dysfunction in CTLA4 deficiency. CHO-CD80-GFP cells were cocultured with activated T cells from the peripheral blood of patients with CTLA4 deficiency (n = 10) and healthy controls (n = 7). Gated CD4⁺FOXP3⁺ Tregs that had acquired GFP from CHO expressing CD80-GFP cells were scored positive for intact CTLA4-mediated transendocytosis. (A) Representative contour plots from a healthy control (left) and a patient with CTLA4 deficiency (right, heterozygous CTLA4 c.410C>T, p.P137L mutation carrier), shown at 2% of events with gating strategy outlined. Quantification of (B) total cellular CTLA4 levels by intracellular staining or (C) CTLA4 function by CD80-GFP transendocytosis as percent GFP⁺CTLA4⁺ in stimulated peripheral blood CD4⁺FOXP3⁺ Tregs from patients with CTLA4 deficiency compared with healthy controls (left) and compared by CTLA4 mutation type (right). Symbols represent unique individuals and bars represent means (± SD) of all data. ∗∗P < .01 by 2-tailed Mann-Whitney U test. (D-E) Correlations between FO B-cell frequency in the peripheral blood and severity of the underlying CTLA4 mutation type. Y-axes are FO B-cell frequency as percent IgD⁺CD27^– B cells in the peripheral blood (data from Figure 1C, here shown as means ± standard error of the mean for all analyzed patients with the indicated CTLA4 mutation type or healthy control state, respectively). (D) X-axis is total cellular CTLA4 levels as MFI by intracellular staining in stimulated CD4⁺FOXP3⁺ Tregs (data from panel B, here consolidated as geometric mean per CTLA4 mutation type or healthy control state, respectively). (E) X-axis is CTLA4 function as CD80-GFP transendocytosis (percent GFP⁺CTLA4⁺) in stimulated CD4⁺FOXP3⁺ Tregs (data from panel C, here consolidated as geometric mean per CTLA4 mutation type or healthy control state, respectively). Simple linear regression with correlation coefficient (R²) and P value shown. Homo, homozygous; het, heterozygous.

Figure 3.

Increased CD40L levels in effector CD4⁺ T cells in CTLA4 deficiency correlate with a decrease in the frequency of circulating FO B cells. (A) Extracellular flow cytometric analysis of peripheral blood T-cell populations in patients with CTLA4 deficiency (n = 11) and healthy controls (n = 10). Representative contour plots from a healthy control (left) and 2 patients with CTLA4 deficiency (right, heterozygous CTLA4 c.410C>T, p.P137L mutation carriers) shown at 5% of events (full description of T-cell gating strategy in supplemental Figure 1B). (B) Quantification of the major CD4⁺ T-cell subsets defined by the extracellular markers CXCR5, CCR7, and CD45RA, including TFH, naïve, CM, TEMRA, and EM populations. Symbols represent unique individuals; bars represent means (± SD) of all data. ∗P < .05; ∗∗P < .01 by 2-tailed Mann-Whitney U test. (C) Correlation between FO B-cell frequency as percent IgD⁺CD27^– B cells in the peripheral blood (data from Figure 1C; x-axis) and naïve CD4⁺ T-cell frequency as percent CD4⁺CXCR5^– T cells in the peripheral blood (data from Panel B; y-axis). Simple linear regression with correlation coefficient (R²), P value, and 95% confidence interval (CI) shown. (D) Intracellular flow cytometric analysis of peripheral blood T-cell populations in patients with CTLA4 deficiency (n = 8) and healthy controls (n = 10). Representative overlaid histograms from a healthy control (blue) and a patient with CTLA4 deficiency (red, heterozygous CTLA4 c.173G>C p.C58S mutation carrier) shown. (E) Quantification of total cellular CD40L levels (MFI) by intracellular staining in the major CD4⁺ T-cell subsets described above in patients with CTLA4 deficiency and healthy controls. Symbols represent unique individuals; bars represent means (± SD) of all data. ∗P < .05; ∗∗P < .01 or as listed by 2-tailed Mann-Whitney U test. (F) Correlation between FO B-cell frequency as percent IgD⁺CD27^– B cells in the peripheral blood (data from Figure 1C; x-axis) and total cellular CD40L levels as MFI in TEMRA CD4⁺ T cells (data from panel E; y-axis). Simple linear regression with correlation coefficient (R²), P value, and 95% CI shown. CM, central memory; EM, effector memory; TFH, T follicular helper.

Figure 4.

Increased mTORC1 signaling due to abundance of transitional and ABC-like aN B-cell clusters in patients with CTLA4 deficiency. (A) Volcano plot of shared differential gene expression analysis between bulk RNA-seq samples of flow-sorted resting FO and TrB populations (T1/2, T3; n = 3 healthy controls; significant differentially expressed genes defined by adjusted P value ≤ .01). (B) Adjusted P value of ingenuity pathway analysis pathway overlaps with significantly differentially expressed genes in TrB vs resting FO B cells. (C) Uniform manifold approximation and projection (UMAP) projection of total naïve (IgD⁺CD27^–) B cells profiled by scRNA-seq from patients with CTLA4 deficiency (n = 4) and healthy controls (n = 3), annotated by naïve B-cell subtype. (D) Dot plot showing the marker gene expression of FO, TrB, and aN B-cell types by cell type annotation; color represents average normalized gene expression and size represents the percent cells with non-zero gene expression. (E) Heat map of scaled gene expression of top 30 differentially expressed genes between annotated cell types. (F) UMAP projection of total naïve (IgD⁺CD27^–) B cells profiled by scRNA-seq, annotated by control or patients with CTLA4 deficiency. (G) Fraction representation of each annotated cell type within each donor scRNA-seq sample. (H) Average log₂-fold change of significant differentially expressed genes (adjusted P value ≤ .05) between distinguishing CD38⁺ TrBs vs CCR7⁺ FO B cells plotted on the x-axis and average log₂-fold change of significant differentially expressed genes between ITGAX⁺ ABC-like aN B cells vs CCR7⁺ FO B cells plotted on the y-axis, showing correlation in differential gene expression distinguishing both TrBs and ABC-like aN B cells from FO B cells. Differentially expressed genes that overlap with significant differentially expressed genes from the bulk RNA-seq analysis in panel A are outlined. (I) Single-cell module scores defined by TrB and ABC-like aN cells shared differentially expressed genes in both TrB and ABC-like aN cells plotted on the x-axis (TrB coexpression score), vs module scores defined by shared FO differentially expressed genes in FO cells on the y-axis (FO B score). Single cells from healthy controls are plotted separately from single cells from patients with CTLA4 deficiency to demonstrate consistent trends in cell type population scores. Pearson correlation rho values (−0.313 for healthy controls, −0.316 for patients with CTLA4 deficiency) were both significant (P < 2E−16). (J) Violin plots of module scores over msigDB hallmark oxidative phosphorylation and mTORC1 signaling pathways (see “Methods”) in healthy controls and patients with CTLA4 deficiency. Violin plots of single-cell module scores shown and evaluated for significance separately for healthy controls and patients with CTLA4 deficiency; ∗∗P < .001; ∗∗∗P < .0001. DEG, differentialy expressed gene; ERK, extracellular signal-regulated kinase; IL4, interleukin 4; msigDB, molecular signatures database; NAD, nicotinamide adenine dinucleotide.

Figure 4.

Increased mTORC1 signaling due to abundance of transitional and ABC-like aN B-cell clusters in patients with CTLA4 deficiency. (A) Volcano plot of shared differential gene expression analysis between bulk RNA-seq samples of flow-sorted resting FO and TrB populations (T1/2, T3; n = 3 healthy controls; significant differentially expressed genes defined by adjusted P value ≤ .01). (B) Adjusted P value of ingenuity pathway analysis pathway overlaps with significantly differentially expressed genes in TrB vs resting FO B cells. (C) Uniform manifold approximation and projection (UMAP) projection of total naïve (IgD⁺CD27^–) B cells profiled by scRNA-seq from patients with CTLA4 deficiency (n = 4) and healthy controls (n = 3), annotated by naïve B-cell subtype. (D) Dot plot showing the marker gene expression of FO, TrB, and aN B-cell types by cell type annotation; color represents average normalized gene expression and size represents the percent cells with non-zero gene expression. (E) Heat map of scaled gene expression of top 30 differentially expressed genes between annotated cell types. (F) UMAP projection of total naïve (IgD⁺CD27^–) B cells profiled by scRNA-seq, annotated by control or patients with CTLA4 deficiency. (G) Fraction representation of each annotated cell type within each donor scRNA-seq sample. (H) Average log₂-fold change of significant differentially expressed genes (adjusted P value ≤ .05) between distinguishing CD38⁺ TrBs vs CCR7⁺ FO B cells plotted on the x-axis and average log₂-fold change of significant differentially expressed genes between ITGAX⁺ ABC-like aN B cells vs CCR7⁺ FO B cells plotted on the y-axis, showing correlation in differential gene expression distinguishing both TrBs and ABC-like aN B cells from FO B cells. Differentially expressed genes that overlap with significant differentially expressed genes from the bulk RNA-seq analysis in panel A are outlined. (I) Single-cell module scores defined by TrB and ABC-like aN cells shared differentially expressed genes in both TrB and ABC-like aN cells plotted on the x-axis (TrB coexpression score), vs module scores defined by shared FO differentially expressed genes in FO cells on the y-axis (FO B score). Single cells from healthy controls are plotted separately from single cells from patients with CTLA4 deficiency to demonstrate consistent trends in cell type population scores. Pearson correlation rho values (−0.313 for healthy controls, −0.316 for patients with CTLA4 deficiency) were both significant (P < 2E−16). (J) Violin plots of module scores over msigDB hallmark oxidative phosphorylation and mTORC1 signaling pathways (see “Methods”) in healthy controls and patients with CTLA4 deficiency. Violin plots of single-cell module scores shown and evaluated for significance separately for healthy controls and patients with CTLA4 deficiency; ∗∗P < .001; ∗∗∗P < .0001. DEG, differentialy expressed gene; ERK, extracellular signal-regulated kinase; IL4, interleukin 4; msigDB, molecular signatures database; NAD, nicotinamide adenine dinucleotide.

Figure 5.

CD40L induces mTORC signaling and arrests FO B-cell maturation in vitro. (A) Levels of (p)-S6 by intracellular flow cytometry in transitional (T) and FO B-cell subsets, sorted and cultured for 24 hours in vitro with CD40L treatment (+CD40L) compared with without (untx) as indicated. Histograms are representative of 3 independent experiments. Quantified data, from 3 independent experiments, are means (± SD) of all data. ∗∗P < .01 by unpaired Student t test. (B) Differentiation by extracellular flow cytometry of T3a and T3b B cells, sorted and cultured for 24 hours in vitro with, compared with, without CD40L treatment or AICAR treatment to inhibit mTORC signaling as a control. Histograms are representative of 3 independent experiments, here showing differences in T3a B-cell levels of MitoTracker (MTG) and CD73 at 24 hours in culture by treatment condition as indicated. Differentiation to the FO B-cell stage was scored by the acquisition of the CD73⁺MTG⁻ phenotype. Quantified data, from 2 independent experiments, are means (± SD) of all data. ∗∗P < .01; ∗∗∗∗P < .0001 by unpaired Student t test. AICAR, 5-aminoimidazole-4-carboxamide ribonucleotide; untx, untreated.

Figure 6.

Treatment with abatacept (CTLA4 immunoglobulin) restores FO B-cell maturation coincident with decreased CD40L levels on effector T cells in CTLA4 deficiency. (A-B) Flow cytometric analysis of peripheral blood B-cell populations in patients with CTLA4 deficiency before (n = 11) and after abatacept (n = 8) therapy (6 patients captured at paired intervals before/after abatacept therapy) and healthy controls (n = 16). Representative contour plots and histograms from a healthy control (left) and the same patient with CTLA4 deficiency (middle and right) before/after abatacept therapy (homozygous CTLA4 3′ UTR mutation carrier in panel A, and heterozygous CTLA4 c.410C>T, p.P137L mutation carrier in panel B) shown at 10% and 5% of events, respectively. Quantitation of B-cell subset frequency as percent IgD⁺CD27⁻ B cells in the peripheral blood. Symbols represent unique individuals; bars represent means (± SD) of all data; dotted lines represent paired pretreatment to posttreatment time points. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001, or as listed by analysis of variance (ANOVA) for multiple comparisons or by Wilcoxon matched-pairs signed-rank test, as indicated. (C) scRNA-seq of total IgD⁺CD27⁻ B cells (from heterozygous CTLA4 c.410C>T, p.P137L mutation carrier) before and after abatacept therapy. UMAP projection of cells is shown by cell type clustering and state of abatacept treatment. Fraction representation of each clustered scRNA-seq cell type. (D) Flow cytometric analysis of peripheral blood T-cell populations in patients with CTLA4 deficiency before (n = 8) and after abatacept (n = 6) therapy (4 patients captured at paired intervals before/after abatacept therapy) and healthy controls (n = 10). Representative overlaid histograms from the same patient with CTLA4 deficiency before/after abatacept therapy (heterozygous CTLA4 c.173G>C p.C58S mutation carrier). Quantitation of total cellular CD40L by intracellular staining in CD4⁺ T-cell subsets. Symbols represent unique individuals; bars represent means (± SD) of all data; dotted lines represent paired before to after treated unique patients. ∗P < .05 or as listed by ANOVA for multiple comparisons or by Wilcoxon matched-pairs signed-rank test, as indicated. (E) Correlation between FO B-cell frequency as percent IgD⁺CD27⁻ B cells in the peripheral blood (data from panel A; x-axis) and total cellular CD40L levels as MFI in TEMRA CD4⁺ T cells (data from panel D; y-axis). Data are from untreated patients with CTLA4 deficiency, treated patients with CTLA4 deficiency, and healthy controls, as indicated. Simple linear regression with correlation coefficient (R²), P value, and 95% CI shown. ns, not significant.