Absolute quantification of BACH1 and BACH2 transcription factors in B and plasma cells reveals their dynamic changes and unique roles

Abstract

Changes in the absolute protein amounts of transcription factors are important for regulating gene expression during cell differentiation and in responses to changes in the cellular and extracellular environment. However, few studies have focused on the absolute quantification of mammalian transcription factors. In this study, we established an absolute quantification method for the transcription factors BACH1 and BACH2, which are expressed in B cells and regulated by direct heme binding. The method used purified recombinant proteins as controls in western blotting and was applied to mouse naïve B cells in the spleen, as well as activated B cells and plasma cells. BACH1 was present in naïve B cells at approximately half the levels of BACH2. In activated B cells, BACH1 decreased compared to naïve B cells, whilst BACH2 increased. In plasma cells, BACH1 increased back to the same extent as in naïve B cells, whilst BACH2 was not detected. Their target genes, Prdm1 and Hmox1, were highly induced in plasma cells. BACH1 was found to undergo degradation with lower concentrations of heme than BACH2. Therefore, BACH1 and BACH2 are similarly abundant in B cells but differ in heme sensitivity, potentially regulating gene expression differently depending on their heme responsiveness.

Graphical Abstract

Open in new tabDownload slide

Transcription factors bind to their respective DNA motifs to regulate gene expression, but they do not necessarily bind to all of their DNA motifs in the genome of a given cell (1–3,). This concept has audaciously been shown for many of the transcription factors in prokaryotic cells (4,). The DNA-binding pattern of a transcription factor determined experimentally, such as chromatin immunoprecipitation-sequencing, may strongly reflect its intra-cellular concentration (5,6). Therefore, the absolute amount of transcription factors in a given cell is one of the most important integrant to be considered when assessing how, where and when transcription factors regulate their target genes. Appropriate transcriptional regulation during differentiation processes and alterations in intra- and extra-cellular environment is often controlled by changes in the amount of transcription factors. Therefore, measuring absolute changes in the protein amount of a given transcription factor during differentiation and environmental responses is important for elucidating the transcriptional regulatory mechanisms of cells.

Transcription factors are classified into families based on structural similarities of their DNA-binding domains, and transcription factors within a family often bind to similar DNA motifs (7,8,). A group of transcription factors that bind to a similar DNA motif may regulate gene expression in a complementary manner, in which transcriptional output is similar irrespective of binding by one or the other transcription factors (9,). In another case, binding of different transcription factors to a DNA motif often creates distinct transcriptional outputs, like strong activation versus weak activation or repression (10,). Alternatively, related transcription factors may show distinct target gene specificity depending on their protein amount (11). Thus, knowledge of the absolute amount of transcription factors will clarify the total molecular number of multiple transcription factors that function complementarily as well as the quantitative relationship between transcription factors that function competitively.

BACH1 and BACH2 are both members of the CNC family of bZIP-type transcription factors and both function primarily as transcriptional repressors (12,). The members of the CNC family bind to the Maf recognition element (MARE), which is related to the AP-1 binding site (13–16,). The Bach1 gene is ubiquitously expressed in a variety of cells (12,). In contrast, the Bach2 gene is highly expressed in B cells, T cells and neurons, and barely expressed in other cell types (13,17,). BACH1 and BACH2 are intra-cellular environment-responsive transcription factors that lose their transcriptional repressive activity when bound by heme (18–21,); their DNA binding activity is reduced, nuclear export is increased and poly-ubiquitination and degradation are induced. Since the amount of heme is altered by the supply of iron (22,), these factors can sense extra- and intra-cellular iron conditions as well. Along B cell development, both BACH1 and BACH2 are expressed in pre-pro B cells, where they complementarily promote early B cell differentiation by repressing myeloid genes (9,). Additionally, BACH2 is upregulated in germinal centre (GC) B cells (23,), where it plays a critical role in promoting class-switch recombination (CSR) and somatic hypermutation (SHM) (24,). BACH2 promotes CSR and SHM by repressing Prdm1 gene expression (25–27,), which encodes BLIMP1, a transcription factor crucial for plasma cell differentiation (28). However, the absolute amounts of BACH1 and BACH2 in B cells and their absolute changes during differentiation processes, including plasma cell differentiation, and quantitative differences in their responses to heme are unknown.

Despite the importance of measuring the absolute amount of transcription factors and their changes for resolving the questions discussed above, few attempts have been made to measure the absolute levels of multiple mammalian transcription factors (29–31). In this study, we measured the absolute amounts of BACH1 and BACH2 in B cells and their quantitative changes during development, activation and differentiation to plasma cells. Their regulation in response to heme was also quantitatively compared.

Materials and Methods

Cell culture

Cell lines used were 38B9 (mouse Pro-B cell line), 18–81 (mouse Pre-B cell line), WEHI231 (mouse Immature B cell line), BAL-17 (mouse Mature B cell line) and X63/0 (mouse plasma cell line). These cells were cultured in Isocove’s modified Dulbecco’s medium (IMDM) (Gibco). For primary B cells derived from mouse spleen, the cells were cultured in a medium containing 10% foetal bovine serum (Nichirei Bioscience), 50 μM 2-mercaptoethanol (2-ME) (Gibco), 100 μg/ml streptomycin (Gibco), 100 μg/ml penicillin (Gibco), 10 mM HEPES (Gibco), 1 mM sodium pyruvate (Gibco) and 0.1 mM non-essential Amino acid (Gibco) in RPMI-1640 (Wako). The cells were cultured under 37°C, 5% CO₂.

Plasmid preparation

The recombinant protein of mouse BACH1 used the amino acid region of 174–417 (cDNA 693–1421) (Fig. 1A). The recombinant protein of mouse BACH2 used the amino acid region of 119–333 (cDNA 355–999) (Fig. 1A) For cloning of BACH1 and BACH2, polymerase chain reactions (PCRs) were performed using the primers listed below and Phusion DNA polymerase (NEB). PCR products and the restriction enzyme cleavage product of the protein expression vector pMAL-c6T in Escherichia coli were subjected to ligation reactions using the Fast-Link DNA Ligation and Screening Kit (AR BROWN).

Primers used

Bach1 Forward: 5’-GATGGGGCGGGCCGCTTGGAAAAGAAACGTG-3’.

Bach1 Reverse: 5’-TATCGGCGGCCGCCATTTGGGCAAGGCGAGTC-3’.

Bach2 Forward: 5’-GTCGACGGGATCCCACAATCTGGGAGGACTCC-3’.

Bach2 Reverse: 5’-GAATTCGGGATCCACTCCTGGACCTGTCCAG-3’.

Protein expression and purification

The plasmid was transformed into E. coli BL21. Transformants were grown until OD600 reached 0.5, to which 1 mM IPTG (Wako) was added and grown at 20°C overnight. Extracts from the transformants were purified on a nickel column (Cytiva Histrap HP), adsorbed on column using buffer of 20 mM Tris pH 8.0, 0.8 M NaCl, 20 mM Imidazole, 1 mM dithiothreitol (DTT) and eluted with 20 mM Tris pH 8.0, 0.5 M NaCl, 0.5 M imidazole 1 mM DTT. Protein concentrations were identified by Coomassie protein assay kit (Thermo Fisher Scientific).

Whole-cell extraction

All operations were performed on ice. RIPA buffer (50 mM Tris–HCl pH 7.4, 1 mM EDTA, 150 mM NaCl, 1% NP-40, 0.5% Na deoxycholate, 0.1% SDS) was added at a ratio of 20 μl per 1 × 10⁶ cells collected, stirred vigorously for 5 min, and kept on ice for 60 min. The supernatant was then centrifuged at 15,000 × g for 5 min and used as the total cell extract.

Nuclear fraction extraction

All operations were performed on ice. Collected cells were dissolved in hypotonic cytoplasmic extraction buffer (10 mM HEPES pH 7.9, 1.5 mM Magnesium chloride, 10 mM Potassium chloride, 0.5 mM DTT, 1.0 mM EDTA, 0.05% NP-40). After washing twice with cytoplasmic extraction buffer, the pellets were added to nuclear extraction buffer (5 mM HEPES, 1.5 mM Magnesium chloride, 300 mM Sodium chloride). After stirring and standing on ice for 60 min, the pellet was centrifuged at 15,000 × g for 5 min, and the supernatant was used as nuclear extract.

SDS-polyacrylamide gel electrophoresis (SDS-PAGE)

Whole-cell extracts were heat-denatured by adding one-fourth equivalent of 5 × SDS sample buffer (250 mM Tris–HCl (pH 6.8), 500 mM DTT, 10% SDS, 20% glycerol) and heat-treated at 95°C for 5 min. The heat-treated whole-cell extracts were then electrophoresed on 7.5% or 12% polyacrylamide gels or 5–20% gradient polyacrylamide gels (HOG-0520-13, Oriental Instruments) and developed according to their molecular weights.

Western blotting

After SDS-PAGE, the gel was transferred to poly (vinylidene fluoride) (PVDF, Millipore) membrane by the wet method. The transferred PVDF membrane was then subjected to an overnight reaction at 4°C in 3% skim milk (Wako) blocking solution (TBS containing 0.05% Tween20). Then, as a secondary antibody, HRP conjugated anti-rabbit IgG blotting reagents (GE Healthcare) diluted 5,000-fold in 3% skim milk blocking solution was reacted with the primary antibody-reacted PVDF membrane for 1 h at room temperature. The membrane was then reacted with Clarity Western ECL Substrate (Bio-Rad) for detection by chemiluminescence, and the luminescence was detected with a ChemiDoc MP Imaging System (Bio-Rad). Primary antibodies were homemade for BACH1 and BACH2: BACH1 (A1–5) (32), BACH2 (N1). Other primary antibodies were β-actin (Gene Tex, GTX109639), HO-1 (ENZ, ADI-SPA-896) and LSD1 (abcam, ab17721).

Absolute quantification of BACH1 and BACH2

A one-half dilution series of purified recombinant protein was prepared and appliqued at 25–800 pg (or 25–400 pg). Samples were applied to the same gel as the dilution series and transferred to the same membrane for detection. ImageJ was used to quantify the signal. The number of protein molecules in the sample was calculated from the approximate curve equation.

Addition of hemin and cycloheximide to cells

Hemin (Sigma-Aldrich) was dissolved in 0.1 N NaOH and a 5 mM stock was prepared and used. Cycloheximide (Sigma-Aldrich) was dissolved in DMSO and a stock of 100 mg/ml was prepared and used. The cell lines BAL-17 were cultured in medium with FBS reduced to 2.5%, to which hemin at several concentrations and cycloheximide at 20 μg/ml were added. For experiments comparing time-dependent degradation, hemin was used at 1–100 μM.

RNA extraction and cDNA synthesis

Total RNA was extracted from cells using the RNeasy Plus Mini Kit (Qiagen) according to the product instructions. The concentration of the extracted total RNA was measured using an absorbance metre (Nano Drop 1000 Spectrophotometer, Thermo Fisher Scientific), and then up to 2,000 ng of total RNA was used for reverse transcription, which was performed using the High-Capacity cDNA Reverse Transcription Kit with RNase Inhibitor (Applied Biosystems) and Random Primers (Thermo Fisher Scientific) using up to 2,000 ng of total RNA. Reverse transcription was performed by heat treatment at 37°C for 60 min and 93°C for 5 min. cDNA (complementary DNA) was stored frozen at −30°C until use.

Quantitative PCR

Quantitative PCR (qPCR) was performed by the intercalator method using SYBR Green I. cDNA from each sample and reagents (LightCycler Fast Start DNA Master SYBR Green I (Roche), and each gene-specific primer) were mixed. The mRNA expression levels of the target genes were measured using the LightCycler96 system (Roche) qPCR machine. The mRNA expression level of the target gene was evaluated by relative quantification, in which the expression level of the internal standard gene was corrected and compared amongst the samples. Actb, the gene for β-Actin, was used as the internal standard gene. The sequences of each primer are as follows:

Actb.

Forward: 5’-CGTTGACATCCGTAAAGACCTC-3’.

Reverse: 5’-AGCCACCGATCCACACAGA-3’.

Bach1.

Forward: 5’-GCCCGTATGCTTGTGTGATT-3’.

Reverse: 5’-CGTGAGAGGGAAATTATCCG-3’.

Bach2.

Forward: 5’-AGTTCATCCACGACATCC-3’.

Reverse: 5’-AGGTGATTCCTCTCCGAC-3’.

Prdm1.

Forward: 5’-CCCTCATCGGTGAAGTCTA-3’.

Reverse: 5’-ACGTAGCGCATCCAGTTG-3’.

Hmox1.

Forward: 5’-GGGTGACAGAAGAGGCTAAG-3’.

Reverse: 5’-GTGTCTGGGATGAGCTAGTG-3’.

Fth1.

Forward: 5’-CGAGGTGGCCGAATCTTCC-3’.

Reverse: 5’-TGCAGTTCCAGTAGTGACTGAT-3’.

Mouse

Mice were wild-type mice and Blimp-1 EGFP reporter mice (33) with a genetic background of the C57BL/6 J strain. Breeding was conducted at the animal experiment facility affiliated with Tohoku University Graduate School of Medicine.

Isolation of primary B cells from mouse spleen

B cells were isolated by negative sorting of single-cell suspensions obtained from mouse spleens by using a B cell isolation kit (Miltenyi Biotec). The isolated B cells were incubated with lipopolysaccharide (LPS) (E. coli. O111:B4 SIGMA) at 20 μg/ml for 4 days, and then subjected to FACS sorting.

Fluorescent-activated cell sorting (FACS)

FACS sorting was performed to isolate activated B cells and plasma cells; IgG3-positive and dead cells were excluded from the EGFP-positive population as antibodies during FACS. Biotin Anti-Mouse IgG3 antibody (BD Pharmingen) and Streptavidin Alexa Fluor 647 conjugate (Invitrogen) were used as antibodies. DAPI (BD Pharmingen) was used for dead cell detection. Cells were sorted by using FACS AriaII (BD).

Results

Establishment of absolute quantification of BACH1 and BACH2

In order to determine and to compare the absolute amounts of BACH1 and BACH2 proteins, we established a western blotting-based method using recombinant proteins of BACH1 and BACH2 as measurement standards. The full-length recombinant BACH2 was particularly challenging to purify (34,35). Therefore, we used protein fragments containing the epitopes of the anti-BACH1 or anti-BACH2 antiserum as recombinant proteins for measurement standards. Specifically, the mouse BACH1 protein fragment encompassed the amino acid region 174–415 (rBACH1), the epitope of the anti-BACH1 antiserum used here and the mouse BACH2 protein fragment encompassed the amino acid region of 119–318 (rBACH2), the epitope of the anti-BACH2 antiserum. These protein fragments were fused to His-MBP tags, were expressed in E. coli and purified using nickel columns (Fig. 1A). The concentrations of the purified proteins were detected and measured by SDS-PAGE and western blotting of rBACH1 and rBACH2 in the range of 25–800 pg. The chemiluminescence intensity of the bands obtained by western blotting was quantified using ImageJ, and the number of molecules was calculated using the equation (Fig. 1B). A calibration curve was then generated with the chemiluminescence intensity on the vertical axis and the number of molecules on the horizontal axis (Fig. 1B). The chemiluminescence intensity of the bands of the recombinant protein changed depending on the number of molecules of the recombinant protein, suggesting the successful generation of a calibration curve. This method was utilized to quantify the absolute amounts of BACH1 and BACH2 proteins in this study.

Quantification of BACH1 and BACH2 in B cell lines representing differentiation stages

The absolute amounts of BACH1 and BACH2 proteins at each stage of B cell differentiation were unknown. Therefore, we quantified the absolute amounts of BACH1 and BACH2 proteins in cell lines representing each stage of differentiation from pro-B cells to plasma cells. First, whole-cell extracts of 5 × 10⁵ cell equivalents of the mature B cell line BAL-17 were detected on the same membranes as the recombinant proteins of the measurement standards used to generate the calibration curve, and the number of molecules of BACH1 and BACH2 present in a single cell was calculated. The number of molecules in single cells was then calculated for the other cell lines in comparison to the BAL-17 cells (Fig. 2A, B). BACH1 was least abundant in 38B9 (pro B cell line), with ~2,400 molecules per cell, and most abundant in the X63/0 (plasma cell line), with ~5,200 molecules. In contrast, BACH2 was undetectable in X63/0 and was present in the other cell lines. The 18–81 (pre-B cell line) had the second lowest number of total BACH2 molecules next to plasma cells, ~1,000 molecules, and WEHI231 (immature B cell line) had the highest number, 4,600 molecules. It was suggested that around half or more of the total amount of BACH2 was present in a phosphorylated state in each cell line. BACH1 and BACH2 were present in roughly equal amounts in the 38B9 and WEHI231. Hence, BACH1 was found to be about three times more abundant in 18–81 and two times more abundant in BAL-17 (mature B cell line) than BACH2. Next, to determine whether the differences in the number of BACH1 and BACH2 molecules in these cells revealed by the absolute quantification corresponded to differences in mRNA, the relative expression levels of mRNA were measured by qPCR in these cell lines. The results showed that Bach1 mRNA expression was lowest in WEHI231; moderate in the 38B9, 18–81 and BAL-17; and high in X63/0. Bach2 mRNA expression was not detected in X63/0 and high in 38B9, 18–81, WEHI231 and BAL-17 (Fig. 2C). When taken together, the expression pattern of Bach1 and Bach2 mRNA in these cells did not always match the pattern of protein molecule numbers. These findings demonstrate the importance of measuring protein absolute amounts in elucidating the mechanism and outcomes of transcriptional regulation by quantitative changes in BACH1 and BACH2.

Next, to test whether quantitative changes in BACH1 and BACH2 correlated with the expression of their known target genes, we also performed qPCR analysis for those genes (Fig. 2D). The Prdm1 gene encoding BLIMP1, a transcription factor essential for plasma cell differentiation, is mainly repressed by BACH2 in B cells with a relatively small auxiliary repression contributed by BACH1 (9,34). The expression of Prdm1 is highly upregulated in X63/0, which did not contain BACH2 protein at the detectable level. The expression of the Hmox1 gene encoding heme oxygenase-1 (HO-1) and Fth1 gene encoding ferritin heavy chain are common target genes of BACH1 and BACH2; their expression was low in 38B9, 18–81, WEHI231 and high in BAL-17 and X63/0. These results suggested that there could be a threshold total number of BACH1 and BACH2 molecules that determines the expression of their common target genes and/or that the expression levels of other transcription factors also affected the expression of these genes in the basal state.

Absolute quantification in the nucleus

BACH1 and BACH2 are transcription factors and are therefore considered to function primarily in the nucleus. However, they possess cytoplasmic localization signals and can shuttle between the cytoplasm and the nucleus (19,36–39). Therefore, we next tried to determine the absolute amounts of BACH1 and BACH2 in nucleus. We fractionated BAL-17 and X63/0 cells into nuclear and cytoplasmic fractions and carried out absolute quantification of BACH1 and BACH2 (Fig. 3A and B). The results showed that, in the BAL-17 cells, there were ~1,000 molecules of BACH1 and 500 molecules of BACH2 present in the nucleus. The X63/0 was also found to have 2,000 molecules of BACH1 in the nucleus. These results indicate that, in BAL-17, ~25% of the total number of molecules of both BACH1 and BACH2 are present in the nucleus, suggesting no significant difference in their subcellular distribution. In contrast, inX63/0, where BACH2 is not expressed, 39% of BACH1 (2,000 molecules) was present in the nucleus, which is greater than the combined total of BACH1 and BACH2 molecules in the nucleus of BAL-17 (Fig. 3C). Therefore, it is likely that BACH1 is solely responsible for MARE-mediated transcriptional regulation in the plasma cell but is not sufficient for the repression of Prdm1.

Changes of BACH1 and BACH2 during plasma cell differentiation of primary naïve B cells

To determine whether the results obtained with the B cell lines reflect the physiological situation, we attempted to measure BACH1 and BACH2 in mouse primary B cells and their dynamic changes during plasma cell differentiation. Naïve B cells were isolated from the spleens of EGFP reporter mice for Prdm1 expression (33,), which is known to report plasma cell differentiation (27). The naïve B cells were stimulated in the presence of LPS for 4 days ex vivo, followed by FACS sorting to isolate activated B cells (EGFP-negative) and plasma cells (EGFP-positive, Fig. 4A). Absolute quantification of BACH1 and BACH2 was then performed on these cells by western blotting as established for the cell lines (Fig. 4B). BACH1 was present ~3,000 molecules in naïve B cells, 1,500 molecules in activated B cells and 3,000 molecules in plasma cells. BACH2 was present ~6,000 molecules in naïve B cells and 7,500 molecules in activated B cells, and it was not detected in plasma cells. The overall patterns of these factors were similar to those in the cell lines.

To investigate whether changes in the number of BACH1 and BACH2 protein molecules during plasma cell differentiation are consistent with changes in mRNA expression, Bach1 and Bach2 mRNA expression levels were measured. Bach1 mRNA expression levels decreased by half after activation and increased ~4-fold in plasma cells. Bach2 mRNA expression levels decreased to 60% after activation and further down to 15% of naïve B cells in plasma cells (Fig. 4D). Changes in the number of protein molecules of BACH1 were consistent with changes in Bach1 mRNA expression, but the change in the number of protein molecules of BACH2 did not match the change in the expression level of Bach2 mRNA. This may reflect the post-transcriptional regulation of BACH2 upon B cell activation (23).

Next, to test whether quantitative changes in BACH1 and BACH2 proteins correlated with the expression of their target genes, mRNAs of Prdm1, Hmox1 and Fth1 were quantified by qPCR (Fig. 4E). As reported previously (26,27) and consistent with the above results, Prdm1 mRNA was barely expressed in naïve and activated B cells and very highly expressed in plasma cells. Hmox1 showed a similar pattern to Prdm1. In contrast, mRNA of Fth1 expression remained constant in naïve and activated B cells and increased only up to 1.5-fold in plasma cells compared to naïve B cells. The mRNA expression pattern of Prdm1 negatively correlated with the pattern of BACH2 protein molecule numbers. The expression pattern of Fth1 mRNA negatively correlates with the total number of BACH1 and BACH2 molecules. Hmox1, a common target gene of BACH1 and BACH2, unexpectedly correlated negatively with BACH2 protein, as did Prdm1. Thus, the expression of these genes may be regulated with unique patterns during plasma cell differentiation of mouse-derived B cells in accordance with quantitative changes in BACH1 and/or BACH2.

Differential sensitivity to heme

BACH1 and BACH2 have reduced transcriptional repressor capacity and undergo induced degradation upon heme binding (18–21,40,41,). Additionally, BACH1 and BACH2 can respond to intra- and extra-cellular environment since their activity is regulated by cysteine oxidation (37,40,42). Because BACH1 and BACH2 may maintain heme and iron homeostasis in response to heme or oxidative stress that can be induced by heme, testing for their differences in heme sensitivity was considered important to understand how BACH1 and BACH2 contribute to the homeostasis of iron and heme metabolism in B cells. No comparative analysis of the heme response of BACH1 and BACH2 in the same cells has been reported so far. In this experiment, heme was added to the culture medium of BAL-17 cells at concentrations ranging from 1 to 100 μM with a fixed treatment time in the presence of cycloheximide, and 50% effective concentration (EC50), in which BACH1 and BACH2 protein molecules became half-reduced, was calculated. EC50 of BACH1 was ~4.5 μM, whereas that of BACH2 was >100 μM (Fig. 5A, B). These results indicate that BACH1 is more sensitive to heme than BACH2 in B cells in terms of protein degradation.

Discussion

The absolute quantification method established in this study offers an experimental approach to quantify protein amounts more easily than other methods like those using mass spectrometry. Whilst this study targeted transcription factors with partially redundant functions, future research incorporating transcription factors that act in an opposite manner will provide a comprehensive understanding of transcriptional regulation by factors binding to the same DNA motif. For instance, NRF2 (NFE2L2), a transcription activator in the CNC family, counteracts the repressors BACH1 and BACH2 (10,16). BACH1 and BACH2 compete with NRF2 for the expression of their common target genes, such as Hmox1, influencing their expression level. Understanding the competitive relationships amongst these transcription factors, all of which respond to intracellular environments, is crucial to fully grasp the systematic regulation of oxidative stress response. Therefore, the next important step is to include quantification of NRF2 in terms of molecular numbers per cells in diverse cells.

The absolute quantification revealed that naïve mouse B cells contain ~3,000 and 6,000 molecules of BACH1 and BACH2, respectively. Mass spectrometry studies suggest that BACH2 is present at ~6,000 molecules in mouse-derived follicular B cells (31,), confirming the accuracy of the absolute quantification method used in this study. Considering previous reports estimating ~10³–10⁵ transcription factors in a mammalian cell (29,), BACH1 and BACH2, with the relatively low absolute abundance, may possess higher DNA binding affinity than other CNC or non-CNC bZIP factors to compete for the target DNA binding. Considering that plasma cells are continuously under oxidative stress due to oxidative antibody protein folding in the endoplasmic reticulum (43–45), the competitive regulation by BACH1 and NRF2 may be critical in these cells.

Comparing the results of primary B cell experiments and the cell line experiments, BACH1 was more abundant than BACH2 in the mature B cell line (Fig. 2B), whereas BACH2 was more abundant than BACH1 in primary B cells. This discrepancy may reflect adaptation of the cell lines to immortalization. Regarding the quantitative changes of BACH1 and BACH2 during plasma cell differentiation of primary B cells, BACH1 was more abundant in naïve B and plasma cells than in activated B cells. In contrast, BACH2 was transiently increased in activated B cells but became barely detectable in plasma cells. These dynamic patterns suggest two interesting possibilities. First, in activated B cells, BACH2 may play critical roles that cannot be contributed by BACH1. One of such is the repression of the Prdm1 gene to transiently suppress plasma cell differentiation for class switch recombination and somatic hypermutation of antibody genes. Likewise, some of the genes regulated by BACH2 in activated B cells (23,) may be specifically repressed by BACH2 rather than BACH1. Second, in plasma cells, BACH1 is the main repressor of MARE. As discussed above, BACH1 may tune oxidative stress response. Because BACH1 promotes ferroptosis (10,46,47), its contribution to the death and lifespan of plasma cells will be an interesting future issue.

For the common target genes of BACH1 and BACH2, changes in total protein levels (i.e. sum) of BACH1 and BACH2 did not always correlate with changes in their target gene expression (Fig. 4E). This result suggests that even common target genes may have different dependencies on BACH1 and BACH2. Whilst Fth1 was not sensitive to the reduction of total BACH1 and BACH2 in plasma cells, Prdm1 and Hmox1 showed much bigger, opposite increase in their expression. BACH1 is known to repress the expression of Hmox1 in diverse types of cells (21,48,). In contrast to those cells, it appears that BACH1 is not effective in its repression of Hmox1 in plasma cells. The de-repression of Hmox1 in plasma cells may be related to the increased function of other transcription factors in plasma cells. For example, NRF2 is upregulated and undergoes nuclear translocation after B cell activation (43,). In addition, the AP-1 transcription factors FRA2 and c-FOS may promote plasma cell differentiation and have elevated functions in plasma cells (49,50,). Since NRF2 and AP-1 transcription factors transcriptionally activate the Hmox1 gene, these factors may cause the de-repression of Hmox1 in plasma cells. As a further possibility, this could be due to a reduction in co-repressors of BACH1 like histone deacetylase-1 (51). It is also possible that the repressor activity of BACH1 may depend on other DNA-binding factors that are not expressed in plasma cells. Knowing that Hmox1 is de-repressed in plasma cells even with the presence of BACH1, it will be important to identify critical downstream genes of BACH1 in plasma cells. This line of research will pave the way towards understanding the quality control and longevity of plasma cells as well as the role of BACH1 in plasma cells.

The dynamic changes in the number of BACH1 and BACH2 molecules during plasma cell differentiation may reflect their cross-regulation. BACH1 protein was transiently decreased in activated B cells in which BACH2 protein was most abundant. Chromatin-immunoprecipitation-sequencing of BACH2 shows that BACH2 directly binds to Bach1 locus (Kurasawa, T et al, unpublished observation). These observations suggest that the Bach1 gene is transiently repressed by BACH2, whilst its expression is increased again upon the disappearance of BACH2 protein in plasma cells. The significance of these findings should be verified in future studies using Bach2-deficient B cells. Additionally, investigating dynamic changes in BACH1 and BACH2 in response to different stimuli beyond LPS stimulation, such as B cell receptor activation by antigen, will provide insights into their overlapping and unique roles.

Although both BACH1 and BACH2 activity is regulated by direct binding to heme, their differences in heme sensitivity have not been quantitively analysed in the same cell. This study showed that BACH1 is more sensitive to heme than BACH2. One possibility is that the difference is due to the differences in the number of CP motifs present in these proteins (six and five CP motifs in mouse BACH1 and BACH2, respectively). However, we examined only protein amount in this study. The heme responsibilities of other events like DNA binding or nuclear export of these proteins require further studies. In addition, the responses of BACH1 and BACH2 to intra- and extra-cellular environmental factors other than heme, such as cysteine oxidation and serine/threonine phosphorylation (37–40,42,52,), should be further examined to understand their differential usage in B cells and under different environments. For example, multiple serine and threonine residues of BACH1 and BACH2 are phosphorylated (38,52), but the extent of their phosphorylation may be different under the same conditions.

Heterozygous Bach2-deficient mice exhibit defective lymphocyte maturation, and haploinsufficiency of BACH2 in humans causes immunodeficiency and autoimmune diseases (53,). Therefore, the absolute quantification method developed in this study could potentially be used to diagnose immunodeficiency and autoimmune diseases by measuring the decrease in BACH2 protein in clinical specimens. Additionally, measuring absolute levels of BACH1 and BACH2 could be used to diagnose or stratify malignant diseases. BACH1 promotes cancer progression in diverse cell types (54–56,). In the B cell lineage, BACH2 acts both as a suppressor (57,) and a promoter (58,59,) of lymphoma and leukaemia. Thus, alterations in the amounts of these factors may be involved in the progression of cancer cells. Whilst BACH1 was more abundant than BACH2 in the mature B cell line, BACH2 was more abundant in naïve B cells from mouse spleen. This difference may reflect the pro-cancer functions of BACH1. Additionally, the proportion of phosphorylated BACH2 was slightly higher in early B cell lines. Since BACH2 is inhibited by phosphorylation, its function is likely suppressed in early B cell lines. Excessive phosphorylation of BACH2, which acts as a tumour suppressor especially in early B cell lymphoma (57), may occur during the development of early B-cell lymphomas. Thus, measuring the ratio of BACH1 to BACH2 or unphosphorylated BACH2 to phosphorylated BACH2 abundance in clinical specimens using an absolute quantification method may bring better understanding of their roles in lymphoma development.

In conclusion, this study revealed for the first time that BACH1 and BACH2 are present in equal amounts in B cells, but they show distinct quantitative changes in B cell activation and plasma cell differentiation. These observations strongly suggest that they may play complementary as well as unique roles in the B cell development and immune responses.

Acknowledgement

We are deeply grateful to members of the Departments of Biochemistry, Tohoku University Graduate School of Medicine for discussions and support; the Biomedical Research Core of Tohoku University Graduate School of Medicine for technical support; and Institute for Animal Experimentation of Tohoku University Graduate School of Medicine for breeding mice.

Funding

This work was supported in part by Grants-in-Aid from the Japan Society for the Promotion of Science 22H00443, 20KK0176, 18H04021 (to K.I.), 22 K07129 (to A.M.) and 23H02671 (to K.O.) and a Research Grant in the Natural Sciences from the Mitsubishi Foundation (to K.I.).

Author Contributions

Writing of original draft: T.K. Conceptualization and methodology: T.K., K.M. and K.I. Major investigation: T.K. Supportive investigation and technical supports: A.M., M.M., K.O. and K.M. Review and editing: T.K. and K.I. Comments: A.M., K.O. and K.M. Supervision: A.M, K.O. and K.I.

Conflict of Interest

The authors declare that they have no conflicts of interest with the contents of this article.

Ethics Statement

Animal experiments in this study were performed in accordance with the American Veterinary Medical Association (AVMA) Guidelines for the Euthanasia of Animals. This study was approved by the Institutional Animal Care and Use Committee of the Tohoku University Environmental & Safety Committee.

References