Heme enhances B-cell proliferation and plasma cell formation through reduced p21 and Rb expression
Herbey O Padilla-Quirarte, Anna K Kania, Nicolas V Janto, Bagdeser Akdogan-Ozdilek, Sakeenah L Hicks, Carly J Roman, Mansi Gupta, Jeremy M Boss, Christopher D Scharer
TL;DR
Heme boosts B-cell growth and plasma cell formation by reducing p21 and Rb, which helps cells progress through the cell cycle.
Contribution
This study reveals a novel mechanism by which heme promotes plasma cell formation through p21 and Rb downregulation.
Findings
Heme treatment increases gene expression and chromatin accessibility in B cells and plasma cells.
Heme enhances B-cell proliferation and plasma cell formation by promoting G1 to S phase transition.
Reduced p21 and Rb levels mediate heme's effect on cell cycle progression.
Abstract
Antibodies are secreted by specialized antibody-secreting cells, also known as plasma cells (PCs), which differentiate from antigen-activated B cells. Antibodies are critical for protection against many types of infection and are correlates of vaccine efficacy. Iron metabolism is important for antibody responses, and heme (the major source of Fe2+) augments PC formation. However, the full spectrum of heme–molecular interactions and effects during B-cell differentiation are not fully understood. Here, we found that heme treatment of differentiating mouse B cells resulted in significant augmentation of the gene expression and chromatin accessibility landscape of both activated B cells and PC, with the largest effect occurring in genes regulating the G1 to S cell cycle transition. Consistent with this effect, naïve and memory B cells displayed enhanced proliferation and PC formation in the…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7- —National Institutes of Health10.13039/100000002
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsT-cell and B-cell Immunology · Blood groups and transfusion · Xenotransplantation and immune response
Introduction
Antibodies (Abs) are crucial for protection against multiple types of infections and are frequently used as correlates of vaccine efficacy. Abs are secreted by specialized antibody-secreting cells, also known as plasma cells (PCs), which differentiate from antigen-activated B cells. During the differentiation process, B cells go through multiple cell division–coupled epigenetic, transcriptomic, and metabolic changes that allow them to sustain their newly acquired function to secrete excessive quantities of Abs.1 As B cells are activated and begin to divide there is a remodeling of transcription factor (TF) expression with naive B cell–specific-TFs, such as PAX5 and BACH22^,^3 being repressed and IRF4 and BLIMP1, which drive PC differentiation,4^,^5 being induced. In addition, cellular metabolism is also dynamically altered to support the unique needs of each differentiation stage. For example, after activation B cells transition from a quiescent energy state to importing high levels of glucose to sustain glycolysis and oxidative phosphorylation metabolism.6^,^7 Cell division is also required for the formation of CD138^+^ PCs,8 with in vivo differentiation requiring up to 8 divisions before PCs.9 However, how these processes are regulated and interact to control B-cell differentiation is not fully understood.
Among the pathways that have been shown to be important in PC differentiation is the metabolism of iron.10–13 Iron is a transition metal and is an essential cofactor for fundamental biochemical enzymes, such as ribonucleotide reductase14 and the cytochrome proteins15 that participate in DNA synthesis and ATP generation as part of the mitochondrial electron transport chain, respectively. For B cells, the availability of iron is an important correlate for the production of Abs as individuals with diminished serum iron levels and transferrin saturation generated significantly fewer Abs following measles virus vaccination.13 Consistent with this finding, mice that were fed an iron-deficient diet developed overall weaker T-dependent and independent B-cell responses compared to animals receiving an iron-sufficient diet.13 Furthermore, human and murine memory B cells (MBCs) present a unique epigenetic and transcriptomic signature enriched for genes that are necessary for the transport and usage of iron,12 suggesting that increased iron utilization and metabolism is a fundamental process supporting B-cell responses.
In mammals, iron is primarily found in the form of heme, which is composed of a ferrous ion (Fe^2+^) enclosed in a protoporphyrin IX ring.16 Several studies have found that murine and human B-cell cultures exposed to heme displayed around 3-fold increased PC formation.10–12 Additionally, mice deficient for heme-oxygenase 1 (HO-1), which catabolizes heme into Fe^2+^, carbon monoxide, and biliverdin, showed higher basal levels of IgM at the expense of high-affinity class-switched Ig antibodies compared to HO-1–sufficient mice.17 Proliferating, activated B cells following influenza vaccination expressed high levels of the transferrin receptor CD7118 that promotes the uptake of ferric iron, further suggesting that iron consumption may be important during B-cell proliferation. These data indicate that heme can be used by B cells as an iron source during PC differentiation; however, the molecular mechanisms by which heme enhances PC differentiation are not fully understood.
As mentioned, a key function for heme includes its role in oxidative phosphorylation (OXPHOS). Recent findings indicate that heme modulates mitochondrial function by enhancing respiratory capacity and OXPHOS in ex vivo differentiating human MBCs.12 These observations suggest that heme plays a pivotal and diverse role in regulating metabolism during naïve and MBC differentiation. Additionally, the stability of the TF BACH2 is regulated directly by heme. BACH2 maintains the B-cell program in part by repressing the expression of Prdm1, encoding BLIMP1, the master transcriptional regulator of PC fate. BACH2 contains 5 Cys-Pro heme-binding motifs, which when bound by heme result in its degradation and subsequent derepression of Prdm1.11^,^19 However, given the multitude of iron-dependent cellular processes, there may be additional mechanisms for heme in promoting PC differentiation and function.
In this study, we found that heme-exposed B cells displayed a transcriptomic and epigenomic state during their transition to PCs that was accompanied by changes in genes regulating G1/S cell cycle checkpoint. In exploring the role of heme in controlling cell cycle progression during B-cell differentiation to PCs, we found that heme increased PC numbers in a time- and cell division–dependent manner. This occurred irrespective of whether the stimulation was T-independent or T-dependent and across a range of heme doses that spanned normal and hemolysis concentrations. The increase in PCs was facilitated by the downregulation of the negative cell cycle regulators p21 and Rb. Consequently, heme enhanced the proliferative capacity of both naïve and memory B cells by promoting the G1/S transition of the cell cycle and increased the proportion of S phase cells at later divisions. Thus, the presence of exogenous heme accelerates the process of B-cell differentiation through its role in regulating key components that control cell cycle progression.
Materials and methods
Mice
C57BL/6J mice were purchased from Jackson Laboratory (stock #000664) and bred within the Division Animal Resources facility at Emory University. Mice were housed within a 12-h dark/light continuous cycle and access to water and food was maintained ad libitum. For each experiment, at least 2 independent biological replicates with groups of 3 to 5 eight- to 12-week-old animals were used with a 50:50 ratio of male to female mice. To generate MBCs, mice were anesthetized using inhaled isoflurane prior to intranasal administration of 15,000 viral foci units of influenza A/Puerto Rico/8/34 in a 30-µL volume. Memory B cells were isolated 35 days postinfection. All animal protocols were approved by the Emory Institutional Animal Care and Use Committee.
B-cell isolation and ex vivo culture
Spleens were isolated and maintained on ice in B-cell media containing RPMI 1640 (Corning Cellgro; 50-020-PC), 10% heat-inactivated FBS (Sigma-Aldrich), 10 mM HEPES (HyClone; SH30237), 1% MEM nonessential amino acids (Sigma-Aldrich; ENBF3930-01), 10 μM sodium pyruvate (Sigma-Aldrich; RNBF6686), 1× penicillin-streptomycin-glutamine (10378016; Life Technologies), and 0.0035% 2-mercaptoethanol (Sigma-Aldrich). To obtain single cell cultures, individual spleens were mashed through a 100-µM cell strainer (Fisher: 22363549) and red blood cells were lysed with ACK lysis buffer (0.15 M NH_4_Cl, 10 mM KHCO_3_, 0.1 mM EDTA) on ice for 3 min. Naïve B cells or MBCs were purified from single cell cultures using CD43^–^ B cell (Miltenyi: 130090862) or MBC (Miltenyi: 130095838) isolation kits, respectively. When indicated, B cells were stained with CellTrace Yellow (Invitrogen: C34567) prior to stimulation. B cells were stimulated with either 20 µg/mL LPS (L2630; Sigma), 20 ng/mL IL-2 (575406; BioLegend), and 5 ng/mL IL-5 (581504; BioLegend) or 500 ng/mL CD40L (R&D Systems, 8230-CL), 10 ng/mL IL-4 (R&D Systems, 404-ML), and 10 ng/mL IL-5. Heme (Sigma: H9039) was dissolved in 1.4 M NH_4_OH (Sigma: 338818) and added in the indicated final concentration to cultures 24 h following the first stimulation. Unless otherwise indicated, heme was added to a final concentration of 60 µM, resulting in a final NH_4_OH concentration of 2 mM in the culture. For titration experiments, vehicle NH_4_OH concentrations were dosed equally to the matched heme treatment concentration to ensure the paired vehicle and heme cultures contained the same amount of NH_4_OH.
Quantitative RT-PCR
Total RNA was isolated from mouse B cells using the Quick-RNA Microprep Kit (Zymo Research, R1051) according to the manufacturer’s protocol. To eliminate residual genomic DNA, samples were treated with DNase at 37 °C for 30 min. Complementary DNA (cDNA) was synthesized using SuperScript II Reverse Transcriptase (Life Technologies, 18064-014). Quantitative RT-PCR was carried out on at least 3 independent RNA samples using gene-specific primers for Trp53 (Fwd: 5′-CCATGGCCCCTGTCATCTTT-3′, Rev: 5′-TGAGGGGAGGAGAGTACGTG-3′) and Cdkn1a (Fwd: 5′-GCAGAATAAAAGGTGCCACAGG-3′, Rev: 5′-CGAAGAGACAACGGCACACT-3′). Gene expression was normalized to Gapdh levels (Fwd: 5′-TGAAGTCGCAGGAGACAACC-3′, Rev: 5′-CCTGGAGAAACCTGCCAAGT-3′).
Western blot
After B-cell culture, cells were harvested and protein lysates were prepared by resuspending them in RIPA buffer (50 mM Tris, pH 8.0; 150 mM NaCl; 0.5% sodium deoxycholate; 0.1% SDS; 1% IGEPAL; 20% glycerol) and were incubated on ice for 20 min. Protein samples (30 µg per lane) were then separated on a 12% acrylamide gel by electrophoresis and transferred onto a nitrocellulose membrane, which was subsequently blocked for 1 h at room temperature with 0.5% nonfat dry milk in TBST buffer (200 mM Tris, pH 7.4; 1.5 M NaCl; 20 mM KCl; 0.5% Tween-20). After blocking, membranes were incubated overnight at 4 °C with rabbit polyclonal anti-p53 (1:5,000; Proteintech, 10442-1-AP) or mouse monoclonal anti-β-actin (1:10,000; clone C4, Santa Cruz Biotechnology, sc-47778) primary antibodies. The next day, membranes were washed and exposed to HRP-conjugated secondary antibodies (sheep anti-mouse IgG, Sigma-Aldrich: A6782: goat anti-rabbit IgG, Sigma: A0545) for 1 h at room temperature, the membrane exposed to Pierce ECL substrate (Thermo 32109) for 5 min and visualized with a ChemiDock MP Imaging System (Bio-Rad). For the analysis of p53 expression, ImageJ software was used to quantify the density of the bands. Western blot image was adjusted for brightness (+40%) and contrast (−40%) using Microsoft PowerPoint v16.
Flow cytometry and FACS sorting
Following stimulation, 1 to 2 × 10^6^ cells were washed with FACS buffer (1× PBS, 2 mM EDTA, 1% BSA), incubated with Fc Block (BD; 553141) for 10 min, and stained with antibody cocktail in a 100 µL total volume of FACS buffer for 30 min on ice. Specific fluorophore-conjugated antibodies used include 0.05 µg B220-PE-Cy7 (RA3-6B2, BioLegend: 103222), 0.0025 µg CD138-BV711 (281-2, BD Horizon: 563193), 0.05 µg CD138-APC (281-2, BioLegend: 142506) 0.05 µg CD11b-APC-Cy7 (M1/70, BioLegend: 101226), 0.05 µg F4/80-APC-Cy7 (BM8, BioLegend: 123118), 0.05 µg Thy1.2-APC-Cy7 (30-H12, BioLegend: 105328), 0.1 µg p21 (EPR18021, Abcam: 237265), 0.1 µg Rb (EPR17512, Abcam: 300158), 0.1 µg rabbit IgG isotype control-AF647 (EPR25A, Abcam: 199093). 0.25 µL Zombie NIR (BioLegend: 103905) was used to stain dead cells, and Phase-Flow kit for proliferating cells (BioLegend: 370704). Intracellular staining was performed using the Foxp3/Transcription Factor Staining Buffer Set (Invitrogen: 00552300). Flow cytometry data were acquired on an FACSymphony A3 (BD Biosciences) and cell sorting was performed on a FACSAriaII (BD Biosciences) using the Emory Integrated Flow Cytometry Core. All data were analyzed using FlowJo v10.
RNA sequencing
For each sample, 1,000 cells were sorted directly into RLT buffer (79216; Qiagen) containing 1% 2-mercaptoethanol. RNA was isolated using the Zymo Quick-RNA MicroPrep Kit (11-328M; Zymo Research). Synthesis of cDNA was performed using the SMART-Seq v4 Ultra Low Input RNA Kit (634894; Takara Bio) kit. Final libraries were generated using 200 pg of cDNA as input for the NexteraXT kit (Illumina, FC-131-1024) with 12 cycles of PCR amplification. Final RNA sequencing (RNA-seq) libraries were quantitated by QuBit (Life Technologies, Q33231), size distributions determined by bioanalyzer (Agilent 2100), pooled at equimolar ratios, and sequenced at Novogene on a NovaSeq6000 using a PE150 run. Raw fastq reads were mapped to the mouse mm10 genome using STAR v2.7.6a20 with the Gencode vM17 reference transcriptome. Duplicate reads were removed from downstream analysis using the PICARD markduplicates v2.23.8 function (http://broadinstitute.github.io/picard/). Reads mapping to exons for all unique ENTREZ genes were compiled and normalized using GenomicRanges v1.38.021 and all genes expressed at 3 or more reads per million in all samples of any one biological group were considered expressed. For gene set enrichment analysis (GSEA),22^,^23 all detected genes were ranked by multiplying the -log_10_ of the P value by the sign of the fold change.
Assay for Transposase-Accessible Chromatin Sequencing
For each sample, 10,000 cells were sorted into FACS buffer and transposition was performed as previously described.24 In brief, cells were resuspended in 12.5 μL 2× TD buffer, 2.5 μL Tn5, 2.5 μL 1% Tween-20, 2.5 μL 0.2% digitonin, and 5 μL H_2_O and incubated at 37 °C for 1 h. Cells were then lysed with the addition of 2 μL 10 mg/mL proteinase-K, 23 μL tagmentation clean-up buffer (326 mM NaCl, 109 mM EDTA, 0.63% SDS), and incubated at 40 °C for 30 min. Tagmented DNA was purified and size selected for small fragments using AMPureXP beads (Beckman Coulter, A63881) and PCR amplified (Roche, KK2602) with dual indexing primers (Illumina, FC-131-2004) to generate a sequencing library. Final libraries were again size selected using AMPureXP beads, quantitated by QuBit (Life Technologies, Q33231), size distributions determined by bioanalyzer (Agilent 2100), pooled at equimolar ratios, and sequenced on a NovaSeq6000 using a PE150 run. Raw fastq reads were mapped to the mm10 genome using Bowtie v2.4.2,25 enriched peaks were called using MACS2 v2.2.7.1,26 and the mapped data were normalized to reads per peak per million (rppm). Differential analysis was performed using DESeq2 v1.26.0.27 Homer v4.11.128 was used for annotating the peaks and the “findMotifsGenome.pl” function was used for motif enrichment analysis of differentially accessible peaks.
Statistical analysis
All statistical analyses for flow cytometry data were performed using GraphPad Prism v10. Paired Student t-tests were used to determine significance between groups where P value <0.05 was considered significant.
Results
Heme promotes PC differentiation kinetics
Heme has been shown to drive PC formation to both T-dependent and T-independent activating signals.10–12 To further explore how heme may regulate B-cell differentiation, a dose response and the kinetics of differentiation were phenotyped. B cells were isolated and purified from splenocytes of C57BL/6 mice and cultured with CD40L, IL-4, and IL-5 to mimic T-dependent differentiation conditions.12 Following 24 h, vehicle control or increasing concentrations of heme (5 to 60 µM) were added to the differentiation culture and the numbers of CD138^+^ PCs were evaluated by flow cytometry at 72 h. Heme treatment induced a dose-dependent increase in the frequency of PCs within the culture that peaked at the 60-μM dose with a 3.4-fold increase relative to vehicle control (Fig. 1A, B). This dose aligns with circulating concentrations of heme documented in individuals experiencing hemolytic disorders such as sickle cell disease.29 Additionally, significant increases were also observed at 10 μM, a concentration within the upper range reported for heme in healthy human plasma.30 Using the 60-μM heme concentration, a time course was performed to assess the kinetics of PC formation. At 48 h, a low frequency of PCs was observed, with no difference in the percentage of CD138^+^ PCs in the heme- or vehicle-treated cultures (Fig. 1C, D). However, the frequency of PCs was significantly augmented after 72 h of culture with 1.6-fold more PCs than vehicle-treated cells. At the 96-h time point, the frequency of PCs in heme-treated cultures contained 21.8% CD138^+^ cells, 1.9-fold more than vehicle control cultures. To determine if heme promoted the formation of PCs in response to other activating signals, B cells were stimulated with LPS, IL-2, and IL-5 as a T-independent antigen model.31 In this system, heme significantly augmented PC formation to similar levels as with CD40L (1.9-fold) compared to control, resulting in 33.1% CD138^+^ cells after 72 h (Fig. 1E, F). Thus, heme augments PC differentiation of naïve B cells irrespective of whether the signal is through T-dependent or -independent activation.
Heme promotes plasma cell (PC) formation in response to T-dependent and T-independent activating signals. (A) Naïve B cells were stimulated with T-dependent (CD40L, IL-4, and IL-5) signals ex vivo and the indicated concentration of heme or vehicle was added after 24 h. Representative flow cytometry analysis at 72 h. (B) Quantitation of the frequency of CD138+ PCs from (A). (C) Naïve B cells were stimulated with CD40L, IL-4, and IL-5 signals ex vivo, and 60 µM of heme or vehicle control was added after 24 h. Representative flow cytometry analysis at 48, 72, and 96 h. (D) Quantitation of the frequency of CD138+ PCs from (C). (E) Naïve B cells were stimulated with T-independent (LPS, IL-2, and IL-5) signals ex vivo, and 60 µM of heme or vehicle control was added after 24 h. Representative flow cytometry analysis at the indicated time points. (F) Quantitation of the frequency of CD138+ PCs from (E). Data represent the combination of at least 2 independent experiments with at least 3 mice each. Error bars represent the mean ± SD and each data point an individual mouse. Statistics were calculated using a paired Student t-test.
Heme changes the transcriptomic signature of naïve B cells
To identify potential mechanisms that heme uses to increase PC formation, B cells were stimulated ex vivo with LPS, IL-2, and IL-5, and heme or vehicle was added after 24 h. After 72 h, B220^+^GL7^+^CD138^–^ activated B cells (ActB) and CD138^+^ PCs were FACS isolated (Fig. S1A, B) and RNA-seq was performed. Principal component analysis (PCA) of all pairwise differentially expressed genes (DEGs) (absolute log_2_ fold change [FC] ≥1 and false discovery rate [FDR] <0.05) indicated that the samples within each group discretely clustered, with PC1 primarily differentiating among the 2 cell types and PC2 separating the heme- and vehicle-treated groups (Fig. 2A). Quantitation of the DEG revealed that heme treatment altered the expression of 715 genes in ActB (385 up and 330 down) and 385 genes in PCs (116 up and 269 down) compared to the vehicle control (Fig. 2B). From these DEGs, 195 were shared between ActB and PCs (Fig. 2C).
Heme imprints a unique transcriptomic signature in differentiating B cells. RNA-seq was performed on activated B cells (ActB; B220+GL7+CD138–) and plasma cells (PCs; CD138+) from naïve B cells activated with LPS, IL-2, and IL-5 for 72 h with 60 µM heme or vehicle added for the last 48 h. (A) Principal component analysis of DEGs (FDR <0.05, absolute log2FC >1) identified between pairwise comparisons of each experimental group. (B) Quantitation of the DEG between heme and vehicle control cultures in ActB and PCs. (C) Venn diagram of the overlap between ActB and PC DEGs from (B). (D) Heatmap comparing the level of expression of iron metabolism genes within ActB and PC samples. (E) Bar plots showing the reads per kilobase million (RPKM) normalized values for the indicated genes. Error bars represent mean ± SD and each circle represents data from an individual biological replicate. Asterisks () indicate statistically significant differences between the indicated groups. (F) Heatmap comparing the level of expression of previously defined bone marrow PC (BMPC) signature genes38 within ActB and PC samples. (G) GSEA comparing the heme vs vehicle control data between ActB and PCs using previously described gene sets.39,40 FDR-corrected P values are indicated for each comparison.*
Treatment with heme resulted in altered gene expression for essential iron metabolism and homeostasis genes that mange iron overload, with ActB and PCs displaying distinct changes for many genes. For example, Slc11a2, which encodes the divalent metal transporter 1 (DMT1) that transports iron through the extracellular membrane,32 was increased in ActB but reduced in PCs (Fig. 2D, E). PCs showed strong increased expression of Hmox1, Ftl1, and Slc40a1. Hmox1 encodes heme oxygenase, which catabolizes heme into ferrous iron, carbon monoxide, and biliverdin.33 Ftl1 (ferritin light chain) and Slc40a1 (ferroportin) function to store and export iron from the cell, respectively.34^,^35 In contrast, heme treatment resulted in downregulation of iron-deficiency genes, including Alas1 (5′aminolevulinate synthase 1), which is important for heme biosynthesis36 and Tfrc (transferrin receptor), which is necessary for iron uptake.37
Because heme can target BACH2 for degradation, relieving the repression that it normally exerts over BLIMP1,11^,^19 we explored whether the observed transcriptional differences were related to the BACH2/BLIMP1 regulatory axis. As expected, PCs showed no significant differences in expression for Bach2 or Prdm1 following heme treatment (Fig. 2E). Additionally, all PC groups displayed high expression of a core gene set defining bone marrow PCs38 compared to ActB samples (Fig. 2F), suggesting that heme treatment did not dramatically alter the transcriptome of differentiated PCs. However, GSEA22^,^23 revealed that in ActB the addition of heme resulted in a downregulation of Bach2 transcripts and a concomitant upregulation of Prdm1 expression. Furthermore, genes that are upregulated in BACH2-deficient B cells39 were significantly upregulated in both heme-treated ActB and PCs (Fig. 2G). Similarly, analysis of genes upregulated by BLIMP1 demonstrated an enrichment in heme-treated ActB and PCs, whereas genes repressed by BLIMP140 were enriched in vehicle-treated control cells. These results demonstrate that the addition of heme promoted a unique transcriptomic state in B cells, affecting iron homeostasis genes and a rebalancing of the BACH2/BLIMP1 transcription factor networks in ActB.
Heme promotes epigenetic alterations as B cells differentiate
Epigenetic remodeling of B cells during differentiation to PCs is essential and occurs systematically in a cell division–coupled manner.41^,^42 Furthermore, iron is an essential cofactor for epigenetic remodeling enzymes that remove methylation and acetylation modifications.43^,^44 Therefore, to assess whether heme influenced the epigenetic landscape of B cells, the assay for transposase-accessible chromatin sequencing (ATAC-seq) was performed on ActB and PCs from vehicle- or heme-treated cultures as described above (Fig. S1A, B). PCA of all differentially accessible regions (DARs) (absolute log_2_FC ≥1 and FDR <0.05) showed that similar to the RNA-seq data, the samples primarily separated based on cell type (PC1) and treatment (PC2) status (Fig. 3A). Overall, more loci lost accessibility with heme treatment than with the vehicle control group irrespective of the cell type analyzed (Fig. 3B). Specifically, 884 DARs were found in ActB (189 increased and 695 decreased), compared with 765 in PCs (24 increased and 741 decreased). Only 84 DAR were shared between ActB and PCs (Fig. 3C), suggesting heme-exerted cell type–specific effects on chromatin accessibility. Analysis of the accessibility patterns of each DAR grouping across the cell types showed that changes promoted by heme in ActB mirrored those patterns observed in PC, with sites losing accessibility in ActB also demonstrating lower signal in PCs. The reciprocal pattern was observed for sites gaining accessibility in heme-treated ActB. Regions that lost accessibility in heme-treated PCs were more uniquely changed in PCs (Fig 3D).
Heme promotes epigenetic remodeling of differentiating B cells. ATAC-seq was performed on activated B cells (ActB; B220+GL7+CD138–) and plasma cells (PC; CD138+) from naïve B cells activated with LPS, IL-2, and IL-5 for 72 h with heme or vehicle added for the last 48 h. (A) Principal component analysis of 51,178 differentially accessible regions (DARs) (FDR <0.05, absolute log2FC >1) identified between pairwise comparisons of each experimental group. Each circle represents data from an individual biological replicate. (B) Quantitation of the DARs between heme and vehicle control cultures in ActB and PCs. (C) Venn diagram of the overlap between ActB and PC DARs from B. (D) Heatmap showing the change in accessibility within each DAR group in each of the conditions. Data are normalized to reads per peak per million (rppm) and scaled across 2 kb surrounding each peak. Signal represents the mean of all samples within the indicated group. (E) Scatterplot comparing the log2FC of ATAC-seq data vs log2FC of RNA-seq data. Each peak was mapped to a unique gene based on the nearest transcription start site. NS, not significant. (F) Dot plot depicting the HOMER transcription factor motif enrichment of the indicated DAR set. Color values are mapped to the motif target to background (T: B) ratio and dot size is scaled to the -log10 of the P value.
To better understand if genes transcriptionally regulated by heme were also affected at the epigenetic level, the fold change in accessibility was correlated with the fold change in gene expression for the corresponding gene (Fig. 3E). Genes responsible for heme catabolism and iron storage were found both to be DEGs and to contain DARs in each cell type and included Hmox1, Ftl1, and the lysosome to cytoplasm heme transporter Slc48a1,45 which were all upregulated and contained DAR with increased accessibility. Conversely, the G1 to S cell cycle inhibitor Cdkn1a was downregulated consistently in ActB and PCs in the presence of heme. Other genes were found to be regulated only in one cell type, including Tfrc and Alas1 in PCs, while the RNA-editing deaminase Aicda and Icosl were downregulated in ActB. These results show that heme influences gene expression networks at the transcriptomic and epigenomic level, targeting genes that are important for the usage of iron in activated B cells, as well as those regulating proliferation and cell cycle control.
To address whether changes in accessibility corresponded to specific TF binding motifs and networks, HOMER analyses were performed on specific DAR sets for each condition. The heme-upregulated DAR contained too few loci, resulting in no statistical TF motif enrichment. Within the ActB and PC DARs that lost accessibility following heme treatment, E2A and REL transcription factors, which are important for Ig recombination46 and survival/proliferation of B lymphocytes, respectively,47 were enriched in both cell types (Fig. 3F). However, the majority of enriched TF motifs were only found in ActB. These included BATF, which has a relevant role in the generation of germinal centers and maturation of antibody-secreting cells48; IRF8, which acts by restraining PC differentiation49; and p53, a tumor-suppressor protein that regulates the G1 to S phase transition of the cell cycle.50
Heme controls cell cycle/proliferation-related genes expression
To further investigate how heme may influence PC formation, GSEA was performed to identify common gene pathways and functions enriched in the ActB and PC DEGs following heme treatment. As expected, both cell types treated with heme were enriched for pathways that correspond to cellular iron metabolism, transport, heme signaling, or ferroptosis (Fig. S1C). Genes associated with metabolic pathways necessary for PC differentiation were enriched with heme treatment in ActB, such as the unfolded protein response (UPR) that allows PCs to sustain high rates of protein synthesis and secretion,51 and mechanistic target of rapamycin (mTOR), which sustains early PC development by promoting the transcription of UPR-related genes.52 Additionally, heme-treated ActB cells were enriched in nuclear factor erythroid 2-related factor 2 (NRF2) pathway genes. NRF2 is a transcription factor that regulates the redox balance in a cell and has a direct effect on the expression of Hmox153 and cell cycle/proliferation regulators, including Cdkn1a.54 Strikingly, both cell types treated with heme expressed lower levels of genes related to cell cycle/proliferation compared to the vehicle control group. Of the different phases of the cell cycle, the enriched gene sets indicated that the G1 to S phase transition was the most impacted following heme treatment (Fig. 4A, Fig. S1C). This cell cycle checkpoint is tightly controlled by p53 and the retinoblastoma protein (Rb),55 whose respective targets were also found to be reduced when heme was added to the culture.50^,^56
Expression of cell cycle checkpoint genes are altered by heme treatment. RNA-seq was performed on activated B cells (ActB) and plasma cells (PCs) from naïve B cells activated with LPS, IL-2, and IL-5 for 72 h with heme or vehicle added for the last 48 h. (A) GSEA analysis of the indicated gene set using RNA-seq data from Fig. 2. GSEA analysis of p5356 and Rb50 targets were previously described. (B) Scatterplot showing the correlation in log2FC of the 195 shared DEGs between PCs and ActB from Fig. 2C with select pathways and corresponding genes highlighted. (C) Bar plots showing the RPKM normalized values for the indicated genes. Error bars represent mean ± SD and each circle represents data from an individual biological replicate. Asterisks () indicate statistically significant differences between the indicated groups. GOBP, Gene Ontology Biological Process; OAS, oligoadenylate synthase; RPKM, reads per kilobase million.*
To investigate further into how heme affects metabolic pathways to promote PC differentiation, the 195 DEGs shared between ActB and PCs were analyzed. Overall, these shared genes displayed a consistent expression pattern between the 2 cell types (Fig. 4B). Genes involved in heme metabolism were identified (Hmox1, Slc40a1, Tfrc, Alas) along with chemokine receptors, lymphocyte activation, oligoadenylate synthase family, lysosomal processes, glycolysis, mTORC1 pathway, and important G1 to S phase cell cycle regulators, such as Cdkn1a. Cdkn1a was downregulated in both ActB and PCs (Fig. 4C). The expression levels of Rb1 and Trp53, encoding p53, were reduced only in ActB as well. These results indicate that heme drives the expression of genes that participate in PC function during the ActB state and importantly regulates the expression of genes that control the G1 to S transition of the cell cycle.
Heme augments naïve B-cell proliferation to T-independent and T-dependent stimulation
Cell division is an essential process for B cells to become PCs, with multiple epigenetic, transcriptomic, and metabolic changes occurring at each division.42^,^57^,^58 Since the RNA- and ATAC-seq results above indicated that cell cycle–related genes were augmented in the presence of heme, we hypothesized that heme-treated cultures would display alterations to the cell cycle and/or the division when PCs are formed. To address this, splenic naïve B cells were purified, stained with the cell trace dye CTY, and cultured with CD40L, IL-4, and IL-5 ex vivo to induce differentiation using T cell–dependent conditions. Vehicle control or increasing concentrations of heme (5 to 60 µM) were added 24 h after stimulation and CTY dilution was analyzed by flow cytometry. The dilution of CTY was used to calculate the mean division number (MDN), which represents the average number of cell divisions each culture and condition has undergone. Heme treatment increased cell division and resulted in a higher MDN at all tested concentrations (5 to 60 µM) compared to vehicle control (Fig. 5A, B). For example, at the 15 µM concentration the MDN reached 4.3, compared to a maximum of 3.1 in vehicle-treated cultures. A time-course experiment was performed using the 60-µM heme dose to assay a window between 48 and 96 h when the peak of CD138^+^ PC formation occurs. At each time point analyzed, CD40L-stimulated, heme-treated cells were found at later divisions compared to the vehicle control (Fig. 5C, D and Fig. S1D). At 64 h, heme-treated cells divided 2 to 3 times, while the control group were primarily between divisions 1 and 2. By 70 h, more than 80% of the heme-treated cells were found between divisions 3–5, whereas the vehicle-treated cells were at divisions 2–3. The MDN of heme-treated cells increased between 88 and 94 h from 3.5 to 6.2 while the vehicle-treated cells only progressed to 5.3 (Fig. 5D). These results suggest that heme facilitates a higher proliferative capacity, allowing a higher percentage of the dividing population to reach later divisions.
Heme induces more proliferating cells at later divisions. (A) Naïve B cells were stimulated with T-dependent (CD40L, IL-4, and IL-5) signals ex vivo and the indicated concentration of heme or vehicle was added after 24 h. Representative flow cytometry histograms showing CTY dilution at 72 h for all B cells in culture. (B) Quantitation of the mean division number (MDN) for all cultures in (A). (C) Naïve B cells were stimulated with T-dependent (CD40L, IL-4, IL-5) signals ex vivo and 60 µM of heme or vehicle control added after 24 h. Representative flow cytometry histograms showing CTY dilution at the indicated time points for all B cells in culture. (D) Quantitation of the MDN of cell cultures in (C). (E) Representative flow cytometry analysis of CD138+ cells vs CTY dilution at the indicated time points. (F) Quantitation of the percentage of CD138+ cells within each division from (E). (G) Naïve B cells were stimulated with T-independent (LPS, IL-2, and IL-5) signals ex vivo, and 60 µM of heme or vehicle control was added after 24 h. Representative flow cytometry histograms showing CTY dilution at 72 h. (H) Quantitation of the percentage of CD138+ cells within each division from (G). All data represent the combination of at least 2 independent experiments. Data points in (B) and (D) represent individual mice with horizontal bars depicting the data mean. Error bars in (F) and (H) depict mean ± SD. Statistics were calculated using a paired Student t-test.
To determine whether PCs were forming at earlier cell divisions, CD138 expression was tracked in concordance with the division tracing dye CTY. Although no significant differences in overall PC numbers were observed at 48 hours, there were more CD138^+^ cells in later divisions of heme-treated versus the vehicle-control cultures (Fig. 5E, F). At 48 h, most of the CD138^+^ cells in heme-treated cultures had advanced to division 2 with significantly more cells also progressing to division 3, whereas the vehicle control group contained more cells in divisions 0 and 1. Similar effects were also observed at 72 h and 96 h, with most of the heme-treated CD138^+^ cells occurring in divisions 4 and 5, while the control group was delayed. To corroborate that heme has the same effect in B cells activated with a T-independent stimulation, purified naïve B cells were stimulated with LPS, IL-2 and IL-5, and heme was added after 24 h as above. At 72 h the cultures were phenotyped by flow cytometry, and the formation of PCs was analyzed with respect to CTY dilution. This revealed that CD138^+^ PCs occurred following heme treatment in significantly higher frequencies (16.8%) than with vehicle control (10.8%) (Fig. 5G). Analysis by cell division revealed a significant increase in CD138^+^ PCs as early as the second division following heme treatment, with progressive increases in CD138^+^ cells through divisions 2–5 compared to the vehicle control group (Fig. 5H). Together, these results show that heme acts on activated B cells to augment their proliferation capacity and accelerates PC differentiation at earlier divisions irrespective of the type of B-cell activation signal.
Heme augments S phase cells at later divisions
The cell cycle transition from G1 phase to S phase is characterized by DNA synthesis and genome replication in preparation for cell division in M phase.55 Iron is an essential cofactor of ribonucleotide reductase, the enzyme that catalyzes the reduction of ribonucleotides to deoxyribonucleotides during DNA replication.59^,^60 Because proliferation was augmented in response to heme after B-cell activation, and because heme altered the transcriptomics and epigenetics of G1 to S phase checkpoint control genes, we hypothesized that there would be more cells actively synthetizing DNA after heme treatment. To address this, BrdU incorporation was assessed in naïve B cells cultured for 72 h following CD40L or LPS stimulation with 60 µM heme or vehicle control added after 24 h. When B cells were activated with T-dependent stimuli, BrdU^+^ cells were significantly higher in the heme-treated group (60.4%) compared to the vehicle-control group (43.3%), indicating that there were more heme-treated cells actively replicating DNA compared with the vehicle control (Fig. 6A). Conversely, when cells were activated with LPS, the frequencies of BrdU^+^ cells remained unchanged (36.8% of control group, and 35.8% in the heme group) (Fig. 6B). However, analysis of BrdU^+^ cells with regard to cell division and MDN revealed that, in both cases, the replicating cells accumulated in later divisions. CD40L-stimulated cells increased their BrdU^+^ MDN following heme addition to 4.4 compared to the vehicle control group with 3.8, while heme-treated LPS-activated cells showed a significant increase of MDN of BrdU^+^ cells to 5.1 compared to 4 for the control group (Fig. 6C, D). These data confirm that heme enhances naïve B-cell proliferation in part by inducing more cells in S phase at later divisions during B-cell differentiation to both T cell–dependent and –independent activation models, suggesting that heme promotes the G1 to S phase transition during B-cell activation.
Heme augments S phase cells at later divisions. (A) Naïve B cells were stained with CTY and stimulated with T-dependent (CD40L, IL-4, and IL-5) signals ex vivo and 60 µM heme or vehicle control added after 24 h. BrdU was pulsed into the cultures 1 h prior to staining and analysis at 72 h. Representative flow cytometry histograms (left) and quantitation (right) showing CTY dilution vs BrdU incorporation. (B) Naïve B cells were stained with CTY and stimulated with T-independent (LPS, IL-2, and IL-5) signals ex vivo and 60 µM heme or vehicle control added after 24h. BrdU was pulsed into the cultures 1 h prior to staining and analysis at 72 h. Representative flow cytometry histograms (left) and quantitation (right) showing CTY dilution vs BrdU incorporation. (C) Representative histogram (left) or MDN quantitation (right) of BrdU+ cells from (A). (D) Representative histogram (left) or MDN quantitation (right) of BrdU+ cells from (B). Data represent the combination of at least 2 independent experiments. Error bars represent the mean ± SD and each data point an individual mouse. Statistics were calculated using a paired Student t-test.
Heme enhances proliferation of memory B cells
We previously showed that MBCs have higher expression of iron metabolism genes compared to naïve B cells and respond more significantly to heme addition by promoting PC differentiation.12 To determine whether heme also affects the proliferation of MBCs, mice were infected with influenza virus A/Puerto Rico/8/1934(H1N1) and memory B cells were allowed to form over 35 days. At the memory time point, splenocytes were isolated, and the class-switched IgG^+^ MBCs were magnetically enriched and then stimulated with CD40L, IL4, and IL5. As above, 60 µM heme or vehicle control was added to the cultures after 24 h and differentiation monitored by flow cytometry over a time course. Consistent with naïve B cells, heme addition resulted in an increase in the frequency of PCs beginning at the 72-h time point compared to the vehicle control, where 17.1% of the total cells expressed CD138 compared to 8.9%, respectively (Fig. S2A, B). Similarly, at 96 h, the vehicle control group contained 37.1% PC, while heme treatment contained 44%. This indicates that heme augments PC formation on activated MBCs in a time-dependent manner. To determine if heme influences the proliferation of MBCs, a detailed time-course spanning 64 to 94 h was performed using the same culture system as above. A significant decrease in CTY intensity in the heme-treated group was observed that began at 64 h (Fig. S2C). At 64 h, the peak of dividing heme-treated cells was observed at division 3, whereas the vehicle control peaked at division 2 (Fig. S2D). By 70 h, divisions 2–5 were all significantly different between heme and control groups, with most of the cells in the heme-treated group found in divisions 3 and 4, while in the control group they peaked in division 3. At 94 h, more than 50% of the heme-treated cells were found at division 6 in contrast to only 40% of the vehicle-treated cells. To quantitate the effect heme has on overall proliferation of MBCs, the MDN was calculated at for each time point. At each time point analyzed, heme treatment significantly increased the MDN for MBC cultures compared to vehicle control (Fig. S2E). These results show that heme augments proliferation of MBCs by inducing more cells at later divisions at later times.
Heme modulates the p21-Rb cell cycle axis
Progression through the G1 to S cell cycle transition depends on relieving the repressive activity of several factors that act as checkpoints. Entry into the cell cycle is initiated by the activity of cyclin-dependent kinases (CDKs) that phosphorylate Rb (encoded by Rb1), resulting in Rb degradation and licensing of the E2F transcription factors to promote the expression of DNA replication genes and entry into the S-phase of the cell cycle (Fig. 7A). This process is negatively regulated by the TF p53, which induces the expression of p21 (encoded by Cdkn1a). p21 binds to the CDK complexes and negatively regulates their activity to promote Rb stabilization. Rb negatively regulates E2F TF activity to prevent target gene expression and ultimately progression through the G1 to S transition55 (Fig. 7A). Our RNA-seq data showed that heme treatment resulted in reduced the expression of G1/S negative regulators including Trp53 (encoding p53), Cdkn1a, and Rb1 in ActB, suggesting that heme acted on this regulatory axis to enhance entry into the cell cycle. However, measurement of Trp53 mRNA and p53 protein levels showed a lack of major changes at 54 or 72 h in the CD40L-activated culture system following heme treatment (Fig. 7B–D). To investigate if there were other indicators of altered p53 activity in the presence of heme, the accessibility at predicted p53 binding motifs was analyzed using the above-mentioned ATAC-seq dataset. The analysis found reduced accessibility at p53 binding motifs following heme treatment in both ActB and PCs treated with heme (Fig. 7E). Following DNA damage in irradiated B cells, p53 directly binds to the promoter of Cdkn1a and induces its expression, resulting in increased cellular p21.61 Consistent with downregulation of p21 following heme treatment, a significant reduction in accessibility was observed at the Cdkn1a promoter region in heme-treated ActB and PCs compared to vehicle controls that coincided with a p53 binding site (Fig. 7F). Consistent with these accessibility changes, Cdkn1a transcript and protein levels were reduced as early as 54 h following heme treatment compared to vehicle control (Fig. 7G–I). Finally, these results suggested that Rb protein should also be reduced. Indeed, heme treatment also resulted in a decrease of Rb protein at 72 h (Fig. 7J, K). Altogether, these results demonstrate that heme modulates the transition from the G1 to S phase of the cell cycle of activated B cells that are differentiating to PCs by downmodulating the expression of the key cell cycle regulators p21 and Rb.
Heme represses the p21-Rb cell cycle regulatory axis. (A) Model depicting how p53, p21, and Rb function to regulate the G1 to S phase transition of the cell cycle under conditions of cell cycle arrest (left) or cell cycle progression (right). Naïve B cells were stimulated with CD40L, IL-4, and IL-5, and heme or vehicle control was added at 24 h. (B) Bar plot representing Trp53 transcript abundance at 54 h and 72 h measured by RT-qPCR. (C) Western blot for p53 protein and β-actin. Images for p53 protein were adjusted for brightness (+40%) and contrast (−40%). (D) Bar plot representing p53 expression relative to β-actin from (C). (E) Boxplot quantitating the ATAC-seq signal at p53 motifs in the indicated condition. (F) Genome plot of the Cdkn1a locus depicting the ATAC-seq signal and significant differential accessible region (box) in heme- vs vehicle-treated activated B cells (ActB) and plasma cells (PCs). The location of a previously identified p53 binding site is indicated in red.61 (G) Bar plot representing Cdkn1a transcript abundance at 54 h and 72 h measured by RT-qPCR. (H) Representative flow cytometry histogram showing intracellular staining for p21 protein in heme or vehicle control cultures after 72 h along with isotype staining and FMO controls. (I) Bar plot quantitating MFI of p21 expression from (H). (J) Representative flow cytometry histogram showing intracellular staining for Rb protein in heme or vehicle control cultures after 72 h along with isotype staining and FMO controls. (K) Bar plot quantitating MFI of Rb expression from (J). Data represent the combination of at least 2 independent experiments with the indicated P value calculated with paired Student t-tests. FMO, Fluorescence Minus One; MFI, median fluorescence intensity; RPPM, reads per peak per million.
Discussion
Iron is a transition metal that is used as a cofactor for essential cellular metabolic redox reactions, including ATP generation and nucleotide synthesis, epigenetic remodeling enzymes, and TF activity. In mammals, iron is predominantly present as heme within erythrocytes where it mediates oxygen transport. In circulation, ferric iron is bound to transferrin and is imported by B cells through the transferrin receptor CD71.62 Heme and iron metabolism have been implicated in B-cell differentiation, as exogenous heme promotes PC formation in response to both T cell-dependent and -independent antigens.10–12 Heme influences this process in part through direct regulation of the TF networks controlling differentiation and modulating metabolic modes that provide PCs with energy.11^,^19 Through genomic profiling of B cells differentiating in the presence of excess heme, we identified both shared and cell type–specific changes in gene expression and chromatin accessibility that indicated that heme more broadly impacted differentiation than was previously appreciated. Specifically, heme-treated ActB cells augmented many cell cycle genes, including the G1 to S phase checkpoint regulators. Heme facilitated the transition to S phase and augmented proliferation through negative regulation of the p21-Rb axis, resulting in enhanced numbers of CD138^+^ cells.
In B cells, the transcription factors BACH2 and BLIMP1 form a mutually repressive feedback loop, with BACH2 maintaining B-cell identity and BLIMP1 promoting the PC fate. Heme can directly bind BACH2, resulting its degradation and subsequent derepression of Prdm1.11^,^19 Consistent with this, we found that heme-treated ActB cells increased Prdm1 transcripts, encoding BLIMP1, and display concomitant changes in BLIMP1-target genes. Additionally, Bach2 expression was reduced, resulting in higher expression of BACH2-repressed target genes following heme treatment. Furthermore, heme treatment impacted the expression of genes in the iron metabolism pathway. Both the ATAC-seq and RNA-seq data sets displayed consistent changes, with genes involved in catabolizing heme and storing excess iron (Hmox1 and Ftl1) being upregulated, while heme import and biosynthesis (Tfrc and Alas1) were downregulated in both ActB and PCs. Importantly, both the PCs and ActB stages of differentiation displayed both unique and shared changes in gene expression and chromatin accessibility. The fact that BACH2 is not expressed in PCs, and that B cells do not uniformly differentiate into PCs upon addition of heme, indicate that heme is acting on additional pathways, in parallel to BACH2, to influence B-cell fate.
In this study, we found that heme promoted progression through the G1/S cell cycle checkpoint, resulting in enhanced proliferation. The transcription factor p53 blocks proliferation through facilitating expression of cell cycle inhibitors including Cdkn1a. Intriguingly, similar to BACH2, heme can directly bind to the C-terminal fragment of p53 resulting in its degradation.63 This indicates that heme promotes degradation of p53 to stimulate cell cycle progression. Consistent with this mechanism, heme-treated B cells significantly diminished accessibility globally at p53 binding motifs. Chromatin immunoprecipitation sequencing data mapping p53 localization in B cells exposed to UV-induced DNA damage revealed a strong binding site at the Cdkn1a promoter.61 Our data indicated that this site lost accessibility upon heme treatment, likely due to less p53 binding, and that Cdkn1a mRNA as well as protein levels were reduced. However, changes in p53 levels were not observed. This suggests that other mechanisms controlling its ability to regulate p21 are altered by heme. Considering that iron is an essential cofactor of the ribonucleotide reductase that synthesizes DNA nucleotides,14 this feedback mechanism may ensure that proper cellular iron levels exist before cell cycle entry. Activated, dividing B cells that respond to influenza vaccination can be identified by expression of the transferrin receptor CD71,18 indicating that heme import is a conserved component of B-cell activation and likely functionally important during this proliferative phase of differentiation.
Profiling of the accessible chromatin landscape following heme treatment revealed a significant number of regions that lost accessibility in both ActB and PCs compared to regions that gained accessibility. This contrasted with the balanced gains and losses observed in gene expression. Iron is an essential cofactor for histone demethylases and some deacetylases. The overabundance of accessibility loss could be explained by changes in transcription factor occupancy but also by enhanced activity of histone remodeling enzymes. Iron excess may facilitate the activity of these enzymes, promoting the rapid turnover of histone marks. In the context of iron deficiency, the ability to remove histone H3 lysine 9 trimethylation at Cyclin E1 was impacted, resulting in inefficient S phase entry.13 Additionally, removal of the repressive H3K27me3 modification at the Prdm1 promoter,41 presumably by the JMJD3/UTX demethylases,64 is required for optimal expression of Prdm1. The analysis of the accessibility patterns across all cell types indicated that heme-induced changes in ActB emulated those observed in PCs. For example, loci that lost accessibility in heme-treated ActB were lower in PCs, regardless of heme treatment, and vice versa for sites that gained accessibility in heme-treated ActB. This indicates that at these loci heme treatment resulted in an epigenetic programming closer to PCs and may have further supported the enhanced differentiation potential of heme-treated ActB cells.
An important aspect of B-cell memory is the enhanced reactivation potential of memory B cells, facilitating a more robust secondary immune response.65 Both human and mouse influenza-specific MBCs upregulate heme metabolism genes compared to antigen-inexperienced naïve B cells and ex vivo reactivation of MBC-enhanced PC formation.12 Here, heme addition to MBC cultures enhanced the proliferative capacity in a similar manner to naïve cells. Consistent with an importance for iron in human B cells, individuals with missense mutations in TFRC lack an ability to import heme and have reduced CD27^+^IgD^–^ MBC frequencies within the B-cell compartment, and their B cells fail to proliferate to T cell–dependent stimulation conditions.66 As with naïve B cells, MBCs also appear to require a proliferative burst before differentiation into PCs. Therefore, importing and maintaining higher levels of heme in MBCs could be a conserved mechanism to lower the intracellular levels of BACH2, shifting the balance toward Prdm1 expression and PC differentiation,67 but also supporting rapid proliferation and transition through G1 and into S phase.
Physiological levels of circulating free heme in healthy individuals have been measured in the low micromolar range (<10 µM),30 and this is the result of heme scavenging by plasma proteins like hemopexin and albumin.68 Our dose-response data revealed that heme levels as low as 5 µM were sufficient to enhance cell division, but differentiation required higher concentrations >10 µM, suggesting that B-cell proliferation is intrinsically sensitive to iron levels. During hemolysis, which can occur during infection or in the context of sickle cell disease, extracellular heme increases substantially beyond a healthy baseline, with up to 100 μM of free heme observed under conditions such as severe malaria or sickle cell disease.29^,^30 Other studies reporting the role of heme in the regulation of B-cell differentiation have used ranges from 20 to 30 μM.10^,^11 The maximal phenotype in our system was observed at 60 μM heme, with consistent changes also observed in the 20 to 30 μM range. Of note, all of these concentrations represent extraphysiological levels of heme that only occur during infection or immune dysregulation. Intriguingly, sickle cell disease is associated with enhanced immunoglobulin levels.69 This suggests that B-cell differentiation is sensitive to distinct ranges of heme, with proliferation and expansion responsive to lower levels than PC differentiation cues that modulate BACH2 levels.
Individuals with lower systemic levels of iron responded poorly to measles vaccination compared to individuals with normal iron levels,13 providing an important correlation to the data presented here. Given that B-cell differentiation requires cell division,8 our study provides important molecular data supporting a mechanistic role for iron in the cell cycle progression and altered epigenetic landscape of ActB. Iron appears to be a central metabolite promoting humoral immune responses. These data build a complex model of how iron influences B-cell differentiation by regulating distinct nodes, including BACH2/BLIMP1 networks, PC ATP production, epigenetic remodeling, and cell cycle checkpoint control. Further delineating the B cell–iron interactome may provide tools to promote antibody responses after infection or vaccination, or inhibit in the case of pathological humoral immune responses.
Supplementary Material
vkag025_Supplementary_Data
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Patterson DG , Kania AK, Zuo Z, Scharer CD, Boss JM. Epigenetic gene regulation in plasma cells. Immunol Rev. 2021;303:8–22.34010461 10.1111/imr.12975 PMC 8387415 · doi ↗ · pubmed ↗
- 2Nutt SL , Heavey B, Rolink AG, Busslinger M. Commitment to the B-lymphoid lineage depends on the transcription factor Pax 5. Nature. 1999;401:556–562.10524622 10.1038/44076 · doi ↗ · pubmed ↗
- 3Muto A et al Bach 2 represses plasma cell gene regulatory network in B cells to promote antibody class switch. EMBO J. 2010;29:4048–4061.20953163 10.1038/emboj.2010.257PMC 3020649 · doi ↗ · pubmed ↗
- 4Klein U et al Transcription factor IRF 4 controls plasma cell differentiation and class-switch recombination. Nat Immunol. 2006;7:773–782.16767092 10.1038/ni 1357 · doi ↗ · pubmed ↗
- 5Shaffer AL et al Blimp-1 orchestrates plasma cell differentiation by extinguishing the mature B cell gene expression program. Immunity. 2002;17:51–62.12150891 10.1016/s 1074-7613(02)00335-7 · doi ↗ · pubmed ↗
- 6Lam WY et al Mitochondrial pyruvate import promotes long-term survival of antibody-secreting plasma cells. Immunity. 2016;45:60–73.27396958 10.1016/j.immuni.2016.06.011PMC 4956536 · doi ↗ · pubmed ↗
- 7Price MJ , Patterson DG, Scharer CD, Boss JM. Progressive upregulation of oxidative metabolism facilitates plasmablast differentiation to a T-independent antigen. Cell Rep. 2018;23:3152–3159.29898388 10.1016/j.celrep.2018.05.053PMC 6092755 · doi ↗ · pubmed ↗
- 8Jelinek DF , Lipsky PE. The role of B cell proliferation in the generation of immunoglobulin-secreting cells in man. J Immunol. 1983;130:2597–2604.6189896 · pubmed ↗