Precursor B cell production from bone marrow in mice and humans declines with age. Because the mechanisms behind are still unknown, we studied five precursor B cell subsets (ProB, PreBI, PreBII large, PreBII small, immature B) and their differentiation-stage characteristic gene expression profiles in healthy individual toddlers and middle-aged adults. Notably, the composition of the precursor B cell compartment did not change with age. The expression levels of several transcripts encoding V(D)J recombination factors were decreased in adults as compared with children: RAG1 expression was significantly reduced in ProB cells, and DNA-PKcs, Ku80, and XRCC4 were decreased in PreBI cells. In contrast, TdT was 3-fold upregulated in immature B cells of adults. Still, N-nucleotides, P-nucleotides, and deletions were similar for IGH and IGK junctions between children and adults. PreBII large cells in adults, but not in children, showed highly upregulated expression of the differentiation inhibitor, inhibitor of DNA binding 2 (ID2), in absence of changes in expression of the ID2-binding partner E2A. Further, we identified impaired Ig locus contraction in adult precursor B cells as a likely mechanism by which ID2-mediated blocking of E2A function results in reduced bone marrow B cell output in adults. The reduced B cell production was not compensated by increased proliferation in adult immature B cells, despite increased Ki67 expression. These findings demonstrate distinct regulatory mechanisms in B cell differentiation between adults and children with a central role for transcriptional regulation of ID2.
Commitment and differentiation of early precursor B cells to Ig-producing B lymphocytes is a key requirement for a competent adaptive immune system. Each differentiation step is tightly controlled by integrated activities of multiple transcription factors in a complex gene regulatory network (1). In fact, a handful of key transcription factors is used in multiple contexts and distinct combinations to initiate and maintain the commitment process throughout the lifespan of the B cell (2).
Production of B lymphocytes from bone marrow (BM) continues throughout life. Similar to T lymphocytes (3), the precursor B cell pool is decreasing with age in both humans (4, 5) and mice (6–8), whereas the production of other hematopoietic lineages seems to continue unchanged (9). With advancing age, the capacity to induce protective Ab responses decreases, leading to reduced ability to deal with new and previously encountered pathogens (10). Some studies in mice suggest that the impact of aging on lymphoid progenitors is first manifested in hematopoietic stem cells (11, 12). Other studies link the waning B cell production in mice to the inability of immature B cells to replenish the peripheral compartments (6), to decreased transition of ProB cells into PreB cells (6, 13), and/or to microenvironmental changes causing altered Ig gene rearrangements (14, 15). The question has also been raised whether specific transcriptionally regulated mechanisms obstruct B cell generation with age, such as changes in expression of inhibitor of DNA binding 2 (ID2) (16–20), a physiological regulator of the essential transcription factor E2A. E2A protein levels in mice have been reported to decrease with age (17), but, to date, no age-related changes were found for ID2 protein using in vitro expanded ProB/early PreB cells from aged and young mice (16).
In contrast to mouse studies, the effects of aging on human precursor B cell development have been scarcely studied to date. Global gene expression has only been studied in successive maturation stages of precursor B cells in children (21) and adults (22), separately. We have previously reported that the total pool of precursor B cells in human BM decreased rapidly during the first 2 y of life concomitantly with reduced expression of RAG1 (5). Still, these studies were unable to address the issue of decreased BM output of B cells. Therefore, we analyzed precursor B cell subsets and their gene expression profiles in BM from healthy young children and adults. Through differentiation stage-dependent analysis of five precursor B cell subsets and pairwise comparisons between children and adults, we identified several mechanisms that suggest tighter checkpoint control in adult precursor B cells.
Materials and Methods
We obtained BM samples from healthy children aged 18 ± 2 mo (mean ± range) and healthy adults aged 50 ± 5 y (mean ± range). The children were eligible for minor surgery, and the adults for elective orthopedic surgery. Both groups were hematologically healthy, and none of the middle-aged adults had active inflammatory disease requiring regular anti-inflammatory medication. Written informed consent was obtained using protocols approved by the Regional Medical Research Ethics Committee of Eastern Norway (REK Øst, accession number 473-02132) (https://helseforskning.etikkom.no/). The study was performed according to the Norwegian Health Regulations.
Isolation of CD10-positive cells
BM aspirates (∼20 ml from children and 120 ml from adults) were subjected to Ficoll density gradient centrifugation (Ficoll-Paque PLUS). CD10+Fig. 1A) after staining with the following Abs: IgM-FITC (G20-127), CD19-allophycocyanin (HIB19), CD20-PE (2H7), CD22-allophycocyanin (IS7), CD123-PE (6H6), CD10-PE-Cy7 (HI10a), and CD34-PerCP (8G12). The membrane marker CD123 was used in combination with CD22 to distinguish precursor B cells from basophilic progenitor cells (CD123+), which also appear in the lymphogate.
n = 39) for Bioanalyzer RIN indicating high RNA purity and integrity.
Microarray analyses and statistical analysis of data
Microarray analyses were performed on subsets from single individuals using the GeneChip Human Exon 1.0 (Affymetrix, Santa Clara, CA). A quantity amounting to 5 ng total RNA was used to generate cDNA with the OvationPico WTA System protocol (NuGEN). MinElute Spin Columns (Qiagen) were used for purification of amplified cDNA. Sense strand cDNA was generated from 3 μg cDNA with the WT-Ovation Exon Module Version 1.0 (NuGEN), according to the manufacturers' protocols for whole genome gene expression analysis. The resulting sense strand cDNA was fragmented and biotinylated using the Encore Biotin Module (NuGEN). The labeled cDNA was hybridized on the array, washed, and stained. The arrays were scanned using the Affymetrix Gene Chip Scanner 3000 7G. The scanned images were processed using the Affymetrix GeneChip Command Console Software, and the CEL files were imported into Partek Genomics Suite software (Partek). The Robust Multichip Analysis algorithm was applied for generation of signal values and normalization. On each array, 21,989 transcripts could be detected. Transcripts were analyzed in core mode (see http://www.affymetrix.com) using signal values of <22.6 across arrays as threshold to filter for low and nonexpressed genes yielding 15,830 expressed genes. For expression comparisons of different groups, profiles were compared using a one-way ANOVA model. The results were expressed as fold change. Gene lists were generated with the criteria of fold change ≥2 and p values < 0.05. For analysis of genes involved in the V(D)J process and in B cell commitment and differentiation, fold change <|2| was shown. The complete gene expression material is available online at ArrayExpress (http://www.ebi.ac.uk/arrayexpress/) with accession number E-MTAB-1422.
Ingenuity Pathway Analysis
Gene networks and canonical pathways representing key genes were identified using Ingenuity Pathways Analysis (www.ingenuity.com). Briefly, the data set containing gene identifiers and corresponding fold changes and p values was uploaded into the web-delivered application, and each gene identifier was mapped to its corresponding gene object in the Ingenuity Pathway Analysis software (www.ingenuity.com). The functional analysis identified the biological functions and/or diseases that were most significant to the data sets. Fisher’s exact test was performed to calculate a p value assigning probability to each biological function.
DNA probe preparation and three-dimensional immunofluorescence in situ hybridization
The three-dimensional DNA immunofluorescence in situ hybridization was performed, as described previously (23, 24), with fosmid clones 761A10 and 3777B2 (BACPAC Resources, Oakland, CA), BAC clones 47P23, 101G24 (BACPAC Resources), and 3087C18 (Open Biosystems, Huntsville, AL) recognizing regions within the human IGH
Approximately 100 μl 1 × 106 cells/ml suspension of freshly sorted cells was directly attached to poly(l
Image acquisition, distance calculations, and statistics
Pictures were captured with a Leica SP5 confocal microscope (Leica Microsystems). Using a ×63 lens (NA 1.4), we acquired images of ∼70 serial optical sections spaced by 0.15 μm. The data sets were deconvolved and analyzed with Huygens Professional software (Scientific Volume Imaging, Hilversum, The Netherlands). The three-dimensional coordinates of the center of mass of each probe were input into Microsoft Excel, and the distances separating each probe were calculated using the following equation: √(Xa − Xb)2 + (Ya − Yb)2 + (Za − Zb)2, where X, Y, Z are the coordinates of object a or b. Differences in distances between each two groups of cells were performed with the nonparametric Mann–Whitney U test using GraphPad Prism version 5.0.
Quantitative RT-PCR for key differentially expressed transcripts
t method (fold change = 2−ΔΔCt) using β2-microglobulin as endogenous control (25). Additional primer/probe sets were used for ID2 that covered the 5′ end of the transcript (Supplemental Table I); two sets for exon I; and one set for exon II, because the custom ID2 assay covered only the exon II and III boundary, which was partially outside the protein-coding area.
IGH and IGK junctional region analysis
DNA was isolated from sorted PreBI and small PreBII cells of healthy children or healthy adults using a GenElute genomic DNA extraction kit (Sigma-Aldrich, St. Louis, MO). Multiplex PCR was performed to amplify complete IGH gene rearrangements from PreBI cells and IGK gene rearrangements from small PreBII cells with V subgroup-specific forward primers and a consensus J primer, and cloned into the pGEM-T Easy vector (Promega Benelux BV, Leiden, The Netherlands) (21http://imgt.cines.fr/) (26). Statistical analyses were performed using the Mann–Whitney U test.
Replication history analysis using the recombination excision circles assay
The replication history of sorted precursor B cell subsets was determined with the κ-deleting recombination excision circles (KREC) assay, as described previously (27). Briefly, the amounts of coding and signal joints of the IGK27). To measure the number of cell divisions undergone by each population, we calculated the ratio between the number of coding joints and signal joints. The previously established control cell line U698 DB01 (InVivoScribe) contains one coding and one signal joint per genome and was used to correct for minor differences in efficiency of both real-time quantitative–PCR assays.
The relative distribution of precursor B cells in BM does not change with age
Within the total leukocyte populations in BM from children and adults, we found considerable variations in lymphocyte frequencies, but no age-related differences (mean ± 2 SD): adults 49.7% ± 8.9, and children 41.3% ± 16.6, respectively. In contrast, the number of isolated BM-restricted CD10+ cells (mean ± 2 SD) was significantly higher (p < 10−5) in children (8.2 × 106/ml ± 1.2 × 106/ml) than in adults (2.7 × 106/ml ± 2.2 × 106/ml).
To study whether the decrease in adults was due to impaired B cell differentiation, we analyzed progenitor B cell subsets using a panel of seven membrane markers (Fig. 1B). The relative sizes of the five major subsets, ProB, PreBI, PreBII large, and PreBII small and immature B cells, varied individually, but were not related to age (Fig. 1C). About 20% of the compartment consisted of the early CD34+ ProB and PreBI progenitor cells, whereas PreBII large and small cells were dominating (∼70%). Immature B cells constituted ∼12%. This indicates a seemingly similar profile of precursor B cell differentiation with age.
Stage-dependent global mRNA expression in precursor B cells from children and adults
To study whether transcriptional regulation of precursor B cell differentiation was affected by age, we analyzed the gene expression profiles of all five precursor B cell stages in four children and four adults. First, differences in gene expression levels were determined per cell stage in children and adults separately using cutoff p value <0.05 and fold change >±2 (Supplemental Table II). Between ProB and PreBI, PreBI and PreBII large, PreBII large and PreBII small, and PreBII small and immature B, respectively, a total of 141, 279, 31, and 697 transcripts had an ANOVA p value <0.05 in children (Fig. 2). These 0.2–4.4% of the transcripts describe the differentiation processes from one stage to the next. Corresponding numbers in adults were 294, 683, 174, and 525, representing 1.9, 4.3, 1.1, and 3.3% of all transcripts present on the array, respectively. Therefore, the total number of transcripts changing expression during differentiation from ProB to PreBI, PreBI to PreBII large, and further to PreBII small was ∼2- to 5-fold higher in adults than in children. Only in the last passage from PreBII small to immature B, ∼25% more transcripts were differentially expressed in children than in adults.
A striking difference was upregulation of 529 transcripts in the PreBI to PreBII large transition in adults versus 111 transcripts in children. Of these, ∼1/6 of the transcripts in adults and 2/3 of the transcripts in children were shared. The other cell transitions involved a more equal number of upregulated transcripts with age (Fig. 2A). The number of downregulated transcripts also showed marked age-related differences (Fig. 2B). Notably, in the last differentiation step, 36% more transcripts were downregulated in children (Fig. 2B). Common transcripts in this step represented ∼1/2 of the transcripts in children and 3/4 in the adult group.
Differences in transcripts essential in V(D)J recombination and their effects on Ig gene rearrangements
We first examined whether transcripts reported to be involved in precursor B cell commitment and differentiation showed similar expression in children and adults, in this study allowing fold change values <|2|. Of 164 transcripts previously described to be precursor B cell associated (21), 23 transcripts were differentially expressed between children and adults (Table I) (mean fold change 2.14, range 1.22–5.81; mean p value 0.02, range 7 × 10−6–0.04).
Among genes higher expressed in children were multiple genes involved in V(D)J recombination. The recombination-activating gene RAG1 was 1.5 higher expressed (p = 0.03) in pediatric ProB cells, whereas for the other subsets, we found no age-related differences. RAG1 and RAG2 mRNA were specifically upregulated in PreBI and small PreBII cells of both children and adults, fitting with their involvement in Ig H and L chain gene rearrangements (Fig. 3A). The nonhomologous end-joining factors Ku80 and DNA-PKcs (28) were both 1.5-fold higher (p = 0.01 and 0.02, respectively) in pediatric PreBI cells than in adults (Fig. 3B). Of transcripts encoding DNA ligation components, XRCC4 showed a 1.7-fold higher expression (p = 0.03) in PreBI cells in children, whereas LIG4 (encoding DNA ligase IV) was similarly expressed in all subsets in children and adults (Fig. 3C). IGK@, representing the Ig κ C region, was significantly higher expressed in adult ProB (3.6-fold up, p = 7 × 10−6) and PreBI (4.2-fold up, p = 4 × 10−6) cells. The template-independent DNA polymerase TdT, inserting random nucleotides during V(D)J recombination, was >3.2-fold higher expressed (p = 0.02) in adult immature B cells than in children (Fig. 3C).
To study whether the reduced RAG expression and increased TdT and IGK@ levels affected V(D)J recombination, we analyzed IGH and IGK gene rearrangements. To avoid the impact of selection, we analyzed IGH gene rearrangements in PreBI cells and IGK gene rearrangements in small PreBII cells (Fig. 3D) (21). Because IGH D-J junctions are formed in ProB cells (21), this analysis allowed us to study the Ig gene rearrangement process in three differentiation stages, as follows: ProB, PreBI, and PreBII small. The analyzed rearrangements did not show differences for deletions and P-nucleotides between children and adults (data not shown). Both D-J and V-D rearrangements showed high numbers of N-nucleotides, reflecting the high TdT levels in ProB and PreBI stages. N-nucleotide additions in IGK gene rearrangements were much lower and did not differ significantly between children and adults. Thus, the low levels of TdT in PreBII small as compared with ProB and PreBI cells result in fewer TdT insertions, but the difference in TdT levels between children and adults in PreBII small does not affect N-nucleotide insertions in PreBII small.
Transcripts encoding pre-BCR components and downstream signaling molecules
Several components of the (pre)BCR were significantly differentially expressed between children and adults. CD79A expression was significantly higher in childhood PreBI and PreBII small cells than in adults (Table I). Furthermore, IGLL1 (Ig λ-like polypeptide 1) was slightly higher expressed (fold change 1.3, p < 0.05) in ProB cells in children than in adults, but similar during further differentiation. VPREB1, however, was 1.7-fold higher expressed (p < 0.002) in adult immature B cells than in children, but otherwise showed no age differences. We also analyzed key transcripts in the pre-BCR signaling pathway (29) and found no age-related differences in the proliferation inhibiting transcripts B cell linker and Bruton´s tyrosine kinase or other components, except for a sole 1.3-fold higher expression (p < 0.05) of SYK (spleen tyrosine kinase) in PreBII large cells in children. These results strongly suggest that pre-BCR signaling in humans, as evaluated through these pathways, is predominantly unchanged with age.
Differences in transcripts essential in B cell commitment and differentiation and the effects on IGH locus contraction
Multiple transcription factors involved in B cell commitment and differentiation were differentially expressed between children and adults. These included transcription factors Helios (IKZF2) (30), which was 2-fold up in ProB (p = 0.01) and 2.2-fold up in PreBI (p = 0.01), and Aiolos (IKZF3), which was 2.1-fold up in PreBI cells (p = 0.0001; Fig. 4A). Furthermore, the differentiation inhibitor ID2 was upregulated in adult ProB (2.5-fold, p = 0.028) and PreBII large cells (5.8-fold, p < 9 × 10−5). In children, ID2 was expressed at relatively low levels with no significant change from one stage to another (Fig. 4B). In adults, a 3.1-fold increase (p = 0.01) in ID2 expression was seen in the transition from PreBI to PreBII large and a corresponding 3.0-fold decrease (p = 0.008) in the subsequent transition to PreBII small, followed by a further 2.4-fold decrease (p = 0.03) differentiating into immature B cells (Fig. 4B). The ID2 expressional pattern was confirmed with quantitative RT-PCR (Fig. 4E, 4F, Supplemental Table III). The target of ID2, E2A, was similarly expressed during differentiation in children and adults (Fig. 4B), although with a slightly higher expression, 20% (p = 0.02), in pediatric ProB cells and a similar, but not significant (p = 0.06) increase in large PreBII cells. Furthermore, other transcription factors involved in B cell differentiation showed strikingly similar expression in children and adults in all differentiation stages (Fig. 4C, 4D). These included the following: EBF1, PAX5, IFN regulatory factor (IRF) 4, IRF8, and POU2AF1.
The upregulation of ID2 in adults, without differences in E2A, PAX5 and EBF1 expression levels, suggests that, of these critical factors, only E2A function is potentially reduced. E2A is implicated in regulating Ig gene rearrangements (reviewed in Ref. 31). The IGH locus was previously found to contract prior to the initiation of complete V to DJ rearrangements in mouse and human PreBI cells (23, 24, 32–34). To study whether reduced E2A function might be contributing to reduced B cell production in adults as compared with children, we measured the spatial distances between three probe sets recognizing distal VH, proximal VH, and CH regions (Fig. 5A). Spatial distances between distal VH and CH regions, distal VH and proximal VH regions, and proximal VH and CH regions were short in childhood PreBI cells and significantly larger in small PreBII cells (Fig. 5B, 5C), confirming previous observations (24). The IGH locus of adult PreBI cells was also more contracted than in adult PreBII cells. Importantly, in both stages, the distances between distal VH and CH regions were larger than in their childhood counterpart (Fig. 5B, 5C). The distances between distal VH and proximal VH between childhood and adult cells were not different, but the distance between proximal VH and CH regions in adult PreBII cells was significantly larger than in children. Thus, the IGH locus was more contracted in childhood than in adult precursor B cells.
ID2-interacting molecules are involved in DNA replication, recombination and repair, cell cycle progression, and checkpoint control
We further analyzed functional annotations of other transcripts changing with ID2 in the PreBI to PreBII large transition in adults by building a network starting with ID2 and E2A using Ingenuity Pathway Analysis. The resulting network consisted of 36 transcripts involved in DNA replication, recombination, repair, and cell cycle progression (Fig. 6). Of these transcripts, 28 were upregulated concomitantly with ID2 and 6 downregulated. Eleven of the upregulated molecules were transcription factors, and 6 were kinases. The cell cycle regulators cyclin-dependent kinase (CDK) 1 and CDK2 were located centrally in the interacting hub. Together with cyclin E, these kinases regulate the G1/S phase transition, in collaboration with cyclin A the S/G2 phase traverse, and with cyclin B 1 and 2 the G2/M (mitosis) transition. Both CDK1 and CDK2 complexes are able to phosphorylate ID2, assumed to relieve its inhibitory effects on E2A. Notably, it has been suggested that the E2A protein level is modulated by both the abundance and phosphorylation status of ID2 (35). The network also includes the important regulator CHEK1, which is involved in several checkpoints of the cell cycle (Fig. 6). Other transcripts upregulated in the network are key factors in DNA replication, such as DNA polymerase α (POLA1) and δ (POLD3) and the primases PRIM1 and PRIM2, which add RNA primers to Okazaki fragments on the lagging strand during DNA synthesis (36). Downregulated transcripts included the cell cycle regulator cyclin D2, possibly caused by ID2, as previously documented (37). Altogether, it seems that with increased ID2 expression in adult PreBII large cells, a network of transcripts involved in DNA replication and cell cycle regulation is concomitantly activated or repressed. This is in contrast to children, in which only five transcripts in this network were significantly upregulated (LYN, IRF4, TFDP2, PTPN6, HDAC9) and four significantly downregulated (SLC22A16, ITPR1, ELK3, DNTT) (Supplemental Table II). Thus, the differential upregulation of ID2 in adult PreBII large cells ties in with an ensemble of other known cell cycle checkpoint regulators, supporting the hypothesis of the PreBII large stage as an important restriction step in adults.
Adult immature B cells do not show signs of compensatory proliferation
In addition to the upregulation of cell cycle regulators CDK1 and CDK2 in adult PreBII large cells, Ki67 was upregulated in adult immature B cells (Fig. 7A). Adults and children both showed strong upregulation in their PreBII cells of this marker for cells that are in cycle (38). However, immature B cells of adults showed significantly higher expression of Ki67 than children (2.6-fold up, p = 3 × 10−4) (39). The upregulation of proliferation markers could suggest that adult precursor B cells undergo proliferation to compensate for reduced B cell production. To study this, we determined the replication history of small PreBII and immature B cells of children and adults using the KREC assay (27). In line with previous observations (27), we did not observe any cell divisions in precursor B cells of children (Fig. 7B). Furthermore, PreBII small and immature B cells of adults did not show signs of proliferation either. Thus, despite the differences in gene expression levels of Ki67, adult immature B cells do not undergo proliferation to compensate for the reduced production of B cell progenitors.
B cell precursors are committed to differentiation or apoptosis. In this study, we characterized the differences in transcriptional activity of precursor B cells in adults and children, and identified differences in V(D)J recombination factors, ID2, and cell cycle genes that could underlie the marked reduction in the human precursor B cell compartment occurring with age (4, 5). Functional analysis of potentially involved mechanisms indicated that upregulation of ID2 in adults inhibited E2A-mediated Ig locus contraction to reduce efficient V(D)J recombination and precursor B cell differentiation.
Our cellular analysis of precursor B cells in BM revealed a 67% decrease in their absolute number in adults as compared with 2-y-old children. This decrease in BM precursor B cell numbers corresponded to the decline in the absolute size of the B cell population in peripheral blood of adults (40). Importantly, the relative composition of the various subsets was unchanged with age and in accordance with previous reports from both humans and mice (4, 5, 40). Thus, the reduced output from BM seemed the result of lower numbers of all B cell progenitors and not restricted to a specific differentiation stage.
The issue whether RAG1 and RAG2 expression decrease with age in precursor B cell subsets has been explored in several studies in mice (4, 14, 15, 41), but very few reports exist from humans. To our knowledge, only one human study (42) has reported, based on RT-PCR and agarose gel electrophoresis, that there was no change in transcription of RAG1 and RAG2 in precursor B cells from fetal and adult BM. This contrasts more recent experimental studies in mice (14, 15), adopting transgenic and knockin RAG2 reporter animals showing that the frequency of ProB cells expressing RAG2 was significantly lower in aged mice as compared with young. We found a 50% higher (p = 0.03) RAG1 expression in pediatric ProB cells as compared with adults, a result that should be confirmed due to the present small cohort size. Of other transcripts involved in the V(D)J rearrangement process, DNA-PKcs, Ku80, and XRCC4 showed a 50–70% higher expression (p = 0.01–0.03) in PreBI cells in children as compared with adults. TdT, in contrast, was 3.2-fold upregulated (p = 0.02) in adult immature B cells. Despite these differences in gene expression levels, we did not observe differences in IGH and IGK gene rearrangements regarding N-nucleotides, P-nucleotides, and deletions between children and adults. These suggest that processing and repair of RAG-induced dsDNA breaks are not affected by age. Still, the reduced RAG1 expression levels in adults could negatively affect the efficiency of Ig gene rearrangements. Because reduced RAG function is not apparent from junction analysis [reviewed in (43)], this may contribute to the lower output of precursor B cells from normal BM with age.
We then analyzed transcriptional activity of genes required for commitment and differentiation of precursor B cells. The most striking finding was the temporal increase in ID2 mRNA levels restricted to PreBII large cells in adults, and only once reported previously in mice (44). As an E2A inhibitory protein, ID2 is assumed to have a central role in modulating the E2A-dependent transcriptional regulatory networks, and hence inhibit B cell lineage commitment and differentiation (45, 46). ID2 has been shown to negatively regulate B cell differentiation in mice spleen (47, 48), and knockdown by shRNA in hematopoietic progenitor cells promoted B cell differentiation and induction of B cell lineage-specific genes (49). In aged, but not in young murine B cell precursors, reduced E2A protein expression was seen in the presence of maintained E2A mRNA levels (17, 18, 50). The reduced E2A protein levels in aged murine precursor B cells seemed to be due to accelerated degradation, because it was effectively blocked by proteasome inhibitors indicating involvement of the ubiquitin/proteasome pathway (18). Frasca et al. (16) were unable to detect ID2 protein using in vitro expanded aged and young precursor B cells, but found E2A protein levels decreased in adults. We found no age-related decrease in E2A mRNA expression in any B precursor subsets in agreement with the situation in mice (16). Notably, the above referred study (16) analyzed only ProB/early PreB cells for ID2 expression and no further maturation stages. In our study, a striking age-related difference in ID2 expression was seen first in PreBII large cells, a subpopulation more mature than the subsets analyzed in mice. We also found a smaller, but significantly higher expression in adult ProB cells, a difference that probably would have been blurred if ProB and PreBI subsets had been analyzed together (16). Another difference between the studies by Frasca et al. (16) and our study was that they analyzed ID2 protein levels in cultured precursor B cells, whereas we analyzed ID2 mRNA expression in freshly isolated precursor B cells. Bordon et al. (44), in contrast, demonstrated fluctuations in ID2 mRNA expression during precursor B cell differentiation in adult wild-type and transgenic mice overexpressing POU class 2-associating factor 1 (OBF1), concordant with our findings in humans.
To study whether the high ID2 gene expression levels indeed negatively affected E2A function, we analyzed one of the processes in which E2A is involved: Ig locus contraction. Indeed, we observed decreased contraction of IGH in PreBI and small PreBII cells in adults. These subsets were analyzed, because they are poised to undergo complete Ig gene rearrangements (21). ID2 transcripts were significantly higher in adults in ProB and large PreBII cells, the stages that directly precede the subsets analyzed for Ig locus contraction. Still, high ID2 transcript levels are likely to affect the subsequent differentiation stage, because the resulting protein products function to decrease E2A activity. Because E2A transcript levels are not upregulated with age, the effect of high ID2 transcripts is not compensated for. Thus, in addition to reduced RAG activity, the efficiency of V(D)J recombination in adult precursor B cells is further reduced by ID2 inhibition of Ig locus contraction.
With increased ID2 mRNA expression in adult PreBII large cells, we identified a concomitant upregulation of transcripts involved in cell cycle progression and control (51, 52), as previously reported. As PreBII large cells are characterized by a transient proliferative burst (21), we observed a similar transcriptional activity in this subset in children and adults by using Ki67 as a proliferation marker. In contrast, there was a clear upregulation of cell cycle and checkpoint-associated genes in the adult subset, concordant with a more stringent cell cycle regulation and control.
Only at the immature B cell stage we found differential expression of the mitotic marker Ki67 with a 2.6-fold (p = 3.4 × 10−4) higher expression in adults indicating a higher fraction of cycling cells. In 2005, Cancro (8) suggested a decreased turnover rate in the immature B cell subset in adult mice as a homeostatic mechanism counteracting the reduced production rate. Our expression data, however, rather suggest an expansion with age of the immature B cell subset, possibly to compensate for reduced production rate. This was not supported by our studies on the relative composition of the immature B cell subset and of the replication history of PreBII small and immature B cells. Thus, the increased levels of Ki67 indicate that more cells are in cell cycle, but these cells do not show signs of (extensive) proliferation to compensate for the reduced production of B cells in adults.
In conclusion, the human precursor B cell pool, although smaller, does not show major compositional variations with age, and comprises dynamic subpopulations held at steady state to support a lifelong production of B lymphocytes. Elevated mRNA levels of the differentiation inhibitory molecule ID2, along with a network of transcripts related to cell cycle checkpoint control in adult PreBII large cells, indicate restriction in precursor B cell differentiation. Moreover, the increased ID2 levels are a likely cause for the observed impaired Ig locus contraction due to limited E2A function. These findings demonstrate distinct regulatory mechanisms in B cell differentiation between adults and children with a central role for transcriptional regulation of ID2.
The authors have no financial conflicts of interest.
We thank all participating adults, children, and parents for trust and generous help, and S.J.W. Bartol and F.S. van de Bovenkamp for technical support.
This work was supported by grants from Torsteds legat, Rakel og Otto Kr. Bruuns legat, and Olav Raagholt og Gerd Meidel Raagholts stiftelse for forskning. This work was also supported by ZonMW/NWO Veni Grant 916.110.90 (to M.C.v.Z.).
The microarray data cited in this article were deposited in the ArrayExpress database (http://www.ebi.ac.uk/arrayexpress/) under accession number E-MTAB-1422.
The online version of this article contains supplemental material.
Abbreviations used in this article:
- bone marrow
- cyclin-dependent kinase
- inhibitor of DNA binding 2
- IFN regulatory factor
- recombination excision circles.
- Received December 31, 2012.
- Accepted May 29, 2013.
- Copyright © 2013 by The American Association of Immunologists, Inc.