Oxygen level is a critical regulator of human B cell differentiation and IgG class switch recombination

Abstract

The generation of high-affinity antibodies requires an efficient germinal center (GC) response. As differentiating B cells cycle between GC dark and light zones they encounter different oxygen pressures (pO₂). However, it is essentially unknown if and how variations in pO₂ affect B cell differentiation, in particular for humans. Using optimized in vitro cultures together with in-depth assessment of B cell phenotype and signaling pathways, we show that oxygen is a critical regulator of human naive B cell differentiation and class switch recombination. Normoxia promotes differentiation into functional antibody secreting cells, while a population of CD27⁺⁺ B cells was uniquely generated under hypoxia. Moreover, time-dependent transitions between hypoxic and normoxic pO₂ during culture - reminiscent of in vivo GC cyclic re-entry - steer different human B cell differentiation trajectories and IgG class switch recombination. Taken together, we identified multiple mechanisms trough which oxygen pressure governs human B cell differentiation.

Keywords: B cells; antibody-secreting cell; class switch recombination; differentiation; germinal center; hypoxia.

Figure 1

Differential pO₂ controls human B cell differentiation into CD27⁺ and antibody secreting cell compartments. (A) Schematic overview of B cell in vitro culture system. 250 resting human naive B cells (CD19⁺CD27^-IgD⁺) were stimulated using a human CD40L-expressing feeder cell layer (subtype ‘High’ (12),); recombinant human IL-4 (25ng/ml) and IL-21 (50ng/ml); and cultured at 5% pCO₂ and 21, 3, or 1% pO₂ for a maximum of 11 days. (B) Representative biaxial CD27/CD38 FACS plots after 5, 7, and 11 days of culture at 21, 3, or 1% pO₂. (C) Quantification of the percentages of CD27 and CD38 subpopulations within total CD19⁺ B cells over time (n = 9). (D) CD138 expression within the CD27⁺CD38⁺ antibody secreting cell population (n = 6). (E) Cumulative secretion of IgM and IgG measured in culture supernatants after 11 days (n = 9). (F) gMFI of AID expression over time (n = 3). (G) Percentage of IgG⁺ cells within CD19⁺ B cells, combined surface and intracellular staining (n = 8). Bars represent means of biological replicates each composed of 2 technical replicates in (D, F) 2, (E, G) 3 or (C) 4 independent experiments. Statistical differences were determined using (C, F) mixed-effects analysis using Tukey’s test for multiple comparisons (D, E, G) repeated measures one-way ANOVA using Tukey’s test for multiple comparisons. *p < 0.05, ***p < 0.001, ****p < 0.0001.

Figure 2

Proliferative human B cells adopt glycolysis and mitochondrial-associated ROS levels are increased in hypoxic cultures (A) gMFI of HIF1α expression over time (n = 6) and histogram overlay at day 3. (B) Glucose uptake (2-NBG; (2-(N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino)-2-deoxyglucose) by B cells cultured for 7 days (n = 9). Lactate production measured in supernatants at day 11 in µM per 10⁴ cultured B cells (n = 6). (C) gMFI of MitoTracker GREEN, MitoTracker RED, and MitoSOX after 7 days of culture (n = 8) indicative for mitochondrial mass, potential and ROS production, respectively. (D) Fatty acid (FA) uptake by B cells cultured for 7 days (n = 5). Bars represent means of biological replicates each composed of two technical replicates of (E) 1, (A, C) 2 or (B, D) 3 independent experiments. Statistical differences were determined using a repeated measures one-way ANOVA using Tukey’s test for multiple comparisons. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 3

pO₂ steers canonical signaling pathways that direct naive B cell differentiation pathways (A-D) gMFI of (A) BCL-6 (n = 6) (B) IRF4 (n = 6) (C) PNA+ (n = 6) (D) CD95 (n = 6) (E) CD86 (n = 6) (F) NFκB active subunit p65 (n = 9) (G) c-Myc (n = 9) (H-I) gMFI of (H) tSTAT6 and pSTAT6 and (I) tSTAT3 and pSTAT3 over time in culture (n = 6) and (left panels) representative histogram overlays of pSTAT and tSTAT expression on day 5 of culture (J) gMFI of PAX5, BLIMP1 and %XBP-1s+ in BLIMP1hi cells (n=6) on day 7 (A-J) Bars represent means of biological replicates each composed of two technical replicates of 2 (H-J) or (A-G) 3 independent experiments (A-I) Differences in gMFI were determined using mixed-effects analysis using Tukey’s test for multiple comparisons. (J) Differences in gMFI were determined using repeated measures one-way ANOVA using Tukey’s tests for multiple comparisons. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 4

Hypoxic pO₂ drives generation of a unique CD27⁺⁺ B cell population, with enhanced antibody secreting cell differentiation capacity and Ig production upon restimulation (A) Quantification of percentage of CD27^-CD38^-, CD27^+/-CD38^-, CD27⁺⁺CD38^- and CD27⁺⁺CD38⁺⁺ cells on day 7 of culture (n = 5) (B) Representative biaxial flow cytometry plot of CD27 and CD38 expression, with heatmap depicting MFI expression of PAX5, BLIMP1 and XBP-1s on day 7. (C) gMFI of PAX5, BLIMP1 and XBP-1s in respective populations of CD27^-CD38^-, CD27^+/-CD38^-, CD27⁺⁺CD38^- and CD27⁺⁺CD38⁺⁺ cells on day 7 (n = 6) (D) Percentage of IgG+ B cells on day 7 of of culture at 1% pO₂ within CD27⁺⁺ CD38^-, CD27^+/- CD38^- or CD27^- CD38^- populations. (n = 5) (E) Representative example of FACS sort gating strategy on day 7 to isolate CD27⁺⁺ CD38^-, CD27^+/- CD38^- or CD27^- CD38^- B cells with schematic experimental time line. (F) Quantification of %CD27⁺CD38⁺ cells by flow cytometry and IgG and IgM secretion by ELISA, 4 or 7 days after sort (day 11 or 14 of culture) (n = 5). (A-F) Bars represent means of biological replicates each composed of two technical replicates of (A, C, D) 2 or (F) 1 independent experiment (A) Differences in %gated cells were determined using mixed-effects analysis using Tukey’s test for multiple comparisons. (C, D, F) Differences in gMFI and % gated cells were determined using repeated measures one-way ANOVA using Tukey’s test for multiple comparisons. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 5

Time-dependent pO₂ transitions alter naive B cell differentiation and IgG class switch recombination. (A) Representative biaxial CD27/CD38 FACS plots of 21, 3, 1% pO2 cultures and 1 – 3% pO₂ transition cultures at day 3, 5 or 7 shown for day 7 and 11 of culture. (B) Quantification of CD27⁺CD38⁺ and CD27⁺CD38^- B cells over time in culture (n = 11) (C) gMFI of BLIMP1 on day 7 (n = 3) (D) Frequency of IgM⁺ and IgG⁺ B cells on day 11 (n = 5). (E) CD138 expression within the CD27⁺CD38⁺ population (n = 5). (F) IgM and IgG levels were measured by ELISA in culture supernantant after 11 days (n = 11). Bars represent means of biological replicates each composed of two technical replicates of (B, F) 3, (D) 2 or (C, E) 1 independent experiment. (B) Statistical differences were determined using mixed-effects analysis using Tukey’s test for multiple comparisons (C-F) Differences in gMFI, % gated cells and Ig production were determined using repeated measures one-way ANOVA using Tukey’s test for multiple comparisons. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 6

pO₂ as a critical driver of B cell differentiation in vitro. Schematic representation of the effect of differential pO₂ on B cell differentiation and underlying signaling pathways.