The origin and nature of age-associated B cells (ABCs) in mice are poorly understood. In this article, we show that their emergence required MHC class II and CD40/CD40L interactions. Young donor B cells were adoptively transferred into congenic recipients and allowed to remain for 1 mo in the absence of external Ag. B cells expressing the T-bet transcription factor, a marker for ABCs, were generated after multiple cell divisions from C57BL/6 donors but not from MHC class II– or CD40-deficient donors. Furthermore, old CD154 (CD40L)-deficient mice did not accrue ABCs, confirming that they arise primarily through T-dependent interactions. To determine what Igs ABCs express, we sequenced VH and Vκ rearranged genes from unimmunized 22-mo-old C57BL/6 mice and showed that they had a heterogeneous repertoire, which was comparable to that seen in old follicular and marginal zone B cell subsets. However, in contrast to the follicular and marginal zone cells, ABCs displayed significant somatic hypermutation. The mutation frequency was lower than found in germinal center cells after deliberate immunization, suggesting that ABCs have undergone mild stimulation from endogenous Ags over time. These observations show that quiescent ABCs are Ag-experienced cells that accumulate during T cell–dependent responses to diverse Ags during the life of an individual.

Profound changes in the composition and dynamics of lymphoid populations occur with age, likely contributing to the decline in immune status, collectively termed immune senescence. For example, B cell production from bone marrow steadily decreases with age, yet the numbers of peripheral B cells remain relatively constant as a result of slowed turnover and altered representation of naive and Ag-experienced B cell subsets (18). A novel B cell subset that accumulates with age, termed age-associated B cells (ABCs), was identified recently (912). These cells have unique features that include preferential responsiveness to TLR7 and TLR9 ligands, surface markers consistent with prior Ag activation, and expression of the T-box transcription factor, Tbx21 (T-bet), which is required for their accumulation (13). Some ABCs also express Itgax (CD11c), an integrin that potentiates their ability to present Ag to T cells (14). ABCs are associated with the onset and severity of humoral autoimmunity in animal models and humans (10, 15, 16). Further, these cells play roles in age-associated immune dysfunctions, including elevated inflammatory cytokine levels and reduced B cell generation rates (11). Finally, a growing literature suggests that B cells with similar characteristics arise during some viral, bacterial, and parasitic infections (13, 1721), implying a role for ABCs in normal immune function.

Despite these observations, the origin and nature of ABCs remain poorly understood. In this study, we investigated their formation, Ig repertoire, and level of somatic hypermutation. The results indicate a polyclonal, Ag-experienced B cell population that arises primarily through T-dependent immune responses to diverse endogenous Ags.

All mice used for experiments were females on a C57BL/6 background. Old mice were obtained from the Charles River aged mouse colony at 18 mo of age and used at 22 mo. Cd154−/− (B6.129S2-Cd40lgtm1Imx/J) mice were purchased from the Jackson Laboratory and kept until 22 mo of age. Young (2–4 mo) CD45.1 and CD45.2 mice were obtained from the Jackson Laboratory. Young I-Ab−/− mice were from Terri Laufer (University of Pennsylvania), and Cd40−/− spleens from young mice were sent from M. Ford’s colony. Aid−/− mice were bred in the National Institute on Aging colony. Animal protocols were reviewed and approved by the Animal Care and Use Committees at the National Institute on Aging and the University of Pennsylvania.

CD23+ splenic B cells from 2-mo-old CD45.2 mice were enriched by positive selection using the MACS bead system (Miltenyi Biotec). Cells were labeled with CFSE (eBioscience), according to the manufacturer’s instructions, and 8 million cells were transferred into each CD45.1 congenic host by retro-orbital injection.

Single-cell suspensions were prepared from spleens and stained with fluorochrome-conjugated Abs. For flow cytometry of the adoptive-transfer and influenza experiments, we used Live/Dead Zombie Aqua, anti-CD45.1-AF700 (A20), anti-CD45.2-BV421 (104), anti-CD19-BV785 (6D5), anti-CD23 biotin (B3B4), and anti-CD11c (N418) (BioLegend). Anti-CD43–PE (S7) was from BD Biosciences. Cells were analyzed on an LSRII, and data were analyzed using FlowJo software (TreeStar). Intracellular stains for T-bet were performed with anti–T-bet–allophycocyanin (4B10; BioLegend) and a Foxp3 transcription factor kit (eBioscience), according to the manufacturer’s instructions. For FACS sorting to isolate subsets, we used anti-CD43–allophycocyanin (S7; BD Biosciences). Anti-CD23–PE Cy7 (B3B4), anti-CD21/CD35–eFluor 450 (4E3), anti-CD45R–FITC (B220, RA3-6B2), and anti-CD93 (AA4.1)-allophycocyanin were from eBioscience. Stained splenocytes were analyzed with a BD FACSCanto II or sorted using a BD FACSAria III, BD FACSAria Fusion, iCyt Reflection (Sony Biotechnology), or Beckman Coulter MoFlo. Follicular (FO) B cells were isolated as CD93 (AA4.1) CD43 B220+ CD21/35+ CD23+. Marginal zone (MZ) B cells were isolated as CD93 (AA4.1) CD43 B220+ CD21/35+ CD23lo. ABCs were isolated as CD93 (AA4.1) CD43 B220+ CD21/35 CD23. Analyses were done using FlowJo software.

Sorted cells were lysed in TRIzol reagent, and RNA was prepared. cDNA was synthesized using SuperScript III Reverse Transcriptase (Invitrogen). Ig H chain VDJ genes and Ig κ light (Igκ)-chain VJ genes were amplified using Taq polymerase (TaKaRa; Clontech) with 5′ degenerate primers specific to framework 1 of V genes and 3′ primers located in IgM or Igκ constant regions, as previously described (22). PCR products were cloned into StrataClone TA cloning vector (Agilent Technologies) and sequenced. Only sequences with unique VDJ or VJ joins were counted. The sequences were blasted against the mouse Ig loci using IgBLAST from the National Center for Biotechnology Information to identify V, D, and J gene segment usage and mutations. For mutational analysis of the JH4 intron, DNA was prepared, and a 492-bp intronic region downstream of JH4 from rearranged VHJ558 genes was amplified using nested PCR. The first round used forward primer J558 5′-AGCCTGACATCTGAGGAC-3′ and reverse primer V1.8NR4R 5′-TCCATACACATACTTCTGTGTTCCT-3′, and the second round used the same J558 forward primer listed above and reverse primer JH2827Bam 5′-CGCGGATCCGATGCCTTTCTCCCTTGACTC-3′. DNA was amplified using Herculase II Fusion DNA Polymerase (Agilent Technologies). The amplified PCR products were cloned into a StrataClone Blunt PCR Cloning vector (Agilent Technologies) and sequenced.

Four-month-old C57BL/6 mice were left uninfected or infected intranasally with 30 tissue culture infectious dose50 of influenza strain A/Puerto Rico/8/1934 (PR8), which was provided by Dr. S. Hensley (University of Pennsylvania). Both uninfected and infected mice were sacrificed 100 d later. To detect hemagglutinin (HA)-reactive B cells, we used a PE-labeled probe that recognizes HA of PR8 (HA-PE) (23). The probe was used at a concentration of 1:500, and data acquisition and analysis were performed as described (23).

We showed previously that ABCs could arise from FO B cells after in vivo expansion in adoptive hosts (9). This extensive division may reflect homeostatic expansion or could implicate Ag-driven activation involving T cell help and costimulation. To distinguish between these possibilities, we modified our adoptive-transfer model with CFSE-labeled donor B cells to use MHC class II– or CD40-deficient donor B cells. The rationale was that homeostatic expansion should be independent of Ag presentation and costimulation, whereas Ag-driven events should not. As shown in Fig. 1A, after 1 mo in the absence of immunization, a small proportion (∼0.2%) of C57BL/6 donor B cells underwent five to eight rounds of division, likely in response to stimulation by endogenous Ags. These extensively divided CFSElo cells were CD23 and T-bet+, which are markers for ABCs. Although the events occurred in only a month, they represent a snapshot of the slow accumulation of ABCs with time. In contrast, B cells from MHC class II–deficient (I-Ab−/−) and CD40-deficient (Cd40−/−) mice underwent fewer divisions with far less T-bet expression than did C57BL/6 cells of the same division cohort. Analyses of multiple mice (Fig. 1B) confirmed a significant increase in T-bet mean fluorescence intensity in CFSElo cells compared with CFSEhi cells from C57BL/6 donors, whereas cells from I-Ab−/− and Cd40−/− donors had no increase. The data suggest that ABCs arise from B cells involved in immune responses to T-dependent Ags, because cognate Ag-presenting capacity and competence to receive CD40 costimulation are required. This interpretation further predicts that CD154-deficient mice, which lack CD40L, should have reduced ABC accumulation. Consistent with this expectation, analysis of splenic B cells from 22-mo-old CD154-deficient mice revealed a paucity of ABCs (Fig. 1C), despite no change in FO and MZ compartments compared with controls (Fig. 1D). Collectively, these results show that ABCs are generated slowly after endogenous Ag presentation via MHC class II and costimulation via the CD40 receptor with CD40L on T cells. The notion that ABCs are derived from T-dependent immune responses raises questions about the breadth and nature of potential Ags involved in their generation and whether they bear hallmarks of germinal center participation. Accordingly, we interrogated Ig variable (V) gene usage and levels of somatic hypermutation among quiescent, naturally occurring ABCs from old mice.

FIGURE 1.

Interactions with MHC class II and CD40 drive the accumulation of ABCs. (A) CD23+ FO B cells from 2-mo-old donor mice (CD45.2) were labeled with CFSE and adoptively transferred into young congenic CD45.1 hosts. Recipient mice were analyzed 1 mo later. Shown are a representative plot of the gating strategy and representative dot plots of CFSE dilution in C57BL/6, I-Ab−/−, and Cd40−/− cells. Numbers depict the percentage of cells in each box. Cells with multiple rounds of proliferation (CFSElo) are boxed in red. Line graphs show intracellular staining for T-bet in CFSElo cells. (B) Analyses of T-bet change in mean fluorescence intensity (ΔMFI) are summarized in three independent experiments for a total of 12 mice for C57BL/6, 6 mice for I-Ab−/−, and 12 mice for Cd40−/−. (C) Spleen cells from 22-mo-old Cd154−/− mice were gated on live B220+ cells. A representative dot plot shows the absence of ABCs (CD23CD21). Numbers represent the percentage of B cells in each population. (D) Absolute B cell numbers of the indicated cell subset from old C57BL/6 and Cd154−/− mice. Error bars signify the SD of values from 31 C57BL/6 mice and 5 Cd154−/− mice. The p values were calculated by an unpaired, equal variance Student t test.

FIGURE 1.

Interactions with MHC class II and CD40 drive the accumulation of ABCs. (A) CD23+ FO B cells from 2-mo-old donor mice (CD45.2) were labeled with CFSE and adoptively transferred into young congenic CD45.1 hosts. Recipient mice were analyzed 1 mo later. Shown are a representative plot of the gating strategy and representative dot plots of CFSE dilution in C57BL/6, I-Ab−/−, and Cd40−/− cells. Numbers depict the percentage of cells in each box. Cells with multiple rounds of proliferation (CFSElo) are boxed in red. Line graphs show intracellular staining for T-bet in CFSElo cells. (B) Analyses of T-bet change in mean fluorescence intensity (ΔMFI) are summarized in three independent experiments for a total of 12 mice for C57BL/6, 6 mice for I-Ab−/−, and 12 mice for Cd40−/−. (C) Spleen cells from 22-mo-old Cd154−/− mice were gated on live B220+ cells. A representative dot plot shows the absence of ABCs (CD23CD21). Numbers represent the percentage of B cells in each population. (D) Absolute B cell numbers of the indicated cell subset from old C57BL/6 and Cd154−/− mice. Error bars signify the SD of values from 31 C57BL/6 mice and 5 Cd154−/− mice. The p values were calculated by an unpaired, equal variance Student t test.

Close modal

ABC accumulation may reflect the aggregate of immune responses to a large and diverse class of endogenous Ags and, thus, involve a broad array of clonotypic specificities. Alternatively, accumulation could be mediated by common exposure to a limited array of self- or environmental ligands that generate oligoclonal expansions with limited repertoire diversity. To differentiate between these possibilities, we sorted FO, MZ, and ABC B cell subsets from 22-mo-old mice and compared VH and Vκ gene segment usage. Because the majority of ABCs express IgM (9), sequencing analyses for H chain genes were done on cDNA amplified with a Cμ 3′ primer and degenerate VH 5′ primers. Likewise, κ L chain genes were identified by amplifying cDNA with a Cκ 3′ primer and degenerate Vκ 5′ primers. Some 2400 unique sequences for both loci were collected and analyzed. Overall, the usage of VH and Vκ gene segments was similar among all three subsets. For VH genes, 85 genes from 12 families were identified, and their frequencies were measured within the subsets. ABCs were compared separately with FO (Fig. 2A) and MZ (Fig. 2B) cells, and significant differences in over- or underutilization were seen in only two to four individual genes. For Vκ genes, 69 genes from 15 families were found; when ABCs were compared with FO (Fig. 3A) or MZ (Fig. 3B) cells, only three to five genes were significantly over- or underused. Thus, there was no evidence for strong repertoire skewing, arguing against a restricted Ag-driven response. We also did not observe significant selection for amino acid replacement changes in CDRs for Ig H chain and Igκ-chains from the ABC population (data not shown). These results suggest that ABCs develop in response to a broad range of Ags.

FIGURE 2.

Diverse VH gene segment usage in ABCs. The frequency of gene expression within the indicated B cell population in old mice (n = 18–24 mice for FO, MZ, and ABC subsets) was determined using RT-PCR. For each subset, ∼400 VH sequences were analyzed. V genes were grouped by family, which is indicated numerically below the graph. Significant differences in V gene usage between ABC and FO (A) or between ABC and MZ (B) subsets were calculated using the Fisher exact test and are shown below the gene name in the yellow bar. The p value heat map scale is shown.

FIGURE 2.

Diverse VH gene segment usage in ABCs. The frequency of gene expression within the indicated B cell population in old mice (n = 18–24 mice for FO, MZ, and ABC subsets) was determined using RT-PCR. For each subset, ∼400 VH sequences were analyzed. V genes were grouped by family, which is indicated numerically below the graph. Significant differences in V gene usage between ABC and FO (A) or between ABC and MZ (B) subsets were calculated using the Fisher exact test and are shown below the gene name in the yellow bar. The p value heat map scale is shown.

Close modal
FIGURE 3.

Diverse Vκ gene segment usage in ABCs. (A) ABC vs. FO subsets. (B) ABC vs. MZ subsets. Details are similar to Fig. 2 legend.

FIGURE 3.

Diverse Vκ gene segment usage in ABCs. (A) ABC vs. FO subsets. (B) ABC vs. MZ subsets. Details are similar to Fig. 2 legend.

Close modal

The requirement for CD40–CD154 interactions in ABC accumulation suggests that most ABCs are products of activation involving cognate T cell help. If so, the V genes of ABCs should contain increased frequencies of mutations compared with other subsets. To address this, we counted the number of mutations in VDJ heavy and VJ κ light exons amplified from FO, MZ, and ABC B cell subsets from 22-mo-old mice used for the repertoire analysis. Sequences of V, D, and J gene segments were compared with their germline counterparts to identify mutations. VDJ and VJ genes from ABCs had a significant 4-fold increase in mutations compared with FO cells and a significant 2-fold increase compared with MZ cells (Fig. 4A, 4B). As a control, V exons were sequenced from FO and MZ cells from young Aid−/− mice, which cannot undergo hypermutation because the activation-induced deaminase protein is absent. The mutation frequency was ∼2 × 10−3 mutations per base pair for activation-induced deaminase–deficient cells, which represents the background frequency of errors produced during cDNA synthesis and PCR amplification. The distribution of mutations per sequence is shown in Fig. 4C, which shows that two thirds of sequences from ABCs had mutations, indicating that most of these B cells have encountered some type of Ag during their existence. An examination of the types of nucleotide substitutions in the cadre of >2700 mutations from VDJ and VJ genes from the ABC sequences showed no difference compared with FO and MZ substitutions (data not shown). Because the error rate for sequencing cDNA clones from RNA is elevated due to errors from the low-fidelity reverse transcriptase used to make cDNA, we also analyzed mutations in the 492-bp JH4 intronic region directly from DNA, using a high-fidelity polymerase. FO, MZ, and ABC B cell subsets were sorted as described above, and the JH4 region was amplified from genomic DNA. As shown in Fig. 4D–F, there was a significant increase in mutation frequency from ABCs compared with those from FO and MZ cells, confirming that ABCs have undergone somatic hypermutation. As a control, introns were sequenced from germinal center B cells of young mice taken 4 wk after immunization with (4-hydroxy-3-nitrophenyl) acetyl–chicken γ globulin (24), and the frequency was 5-fold higher than in ABCs. This comparison places ABCs in the middle between naive and germinal center cells, suggesting that they undergo mild chronic stimulation by endogenous Ags versus acute stimulation by immunization.

FIGURE 4.

ABCs have increased somatic hypermutation. (AC) Exon sequences from Figs. 2 and 3 were analyzed for mutations in rearranged VDJ genes for the H chain and in rearranged VJ genes for the κ L chain. (A) Numbers of unique sequences, nucleotides, mutations, and frequencies. (B) Mean mutation frequencies (mutations/nucleotides) in the exons of each B cell subset were calculated. The dotted line indicates the mutation frequency in Aid−/− FO and MZ B cells from young mice. (C) Distribution of mutations per sequence. The number of sequences is shown in the center of each circle. Segments represent the proportion of sequences that contain the indicated number of mutations. (DF) JH4 intron sequences were analyzed from genomic DNA of 11–13 mice for each subset. (D). Numbers of sequences, nucleotides, mutations, and frequencies. (E) Mean mutation frequencies; the dotted line represents the frequency in germinal centers from young immunized mice. (F) Distribution of mutations per sequence. *p < 0.0001, χ2 test.

FIGURE 4.

ABCs have increased somatic hypermutation. (AC) Exon sequences from Figs. 2 and 3 were analyzed for mutations in rearranged VDJ genes for the H chain and in rearranged VJ genes for the κ L chain. (A) Numbers of unique sequences, nucleotides, mutations, and frequencies. (B) Mean mutation frequencies (mutations/nucleotides) in the exons of each B cell subset were calculated. The dotted line indicates the mutation frequency in Aid−/− FO and MZ B cells from young mice. (C) Distribution of mutations per sequence. The number of sequences is shown in the center of each circle. Segments represent the proportion of sequences that contain the indicated number of mutations. (DF) JH4 intron sequences were analyzed from genomic DNA of 11–13 mice for each subset. (D). Numbers of sequences, nucleotides, mutations, and frequencies. (E) Mean mutation frequencies; the dotted line represents the frequency in germinal centers from young immunized mice. (F) Distribution of mutations per sequence. *p < 0.0001, χ2 test.

Close modal

Our results suggest an Ag-driven origin for ABCs that, coupled with their continuous accumulation, BLyS independence, and resting state (9, 25), lends credence to the idea that they are an unusual subset of B cells. To further interrogate the provenance of ABCs, we compared their gene-expression profiles with those from FO B cells sorted from old and young mice. ABC uniqueness is shown by a subset of 70 genes with ≥5-fold higher expression in old ABCs compared with old and young FO B cells (Supplemental Fig. 1A, Supplemental Table I). T-bet and CD11c were overexpressed in ABCs, confirming previous reports (13, 14). Principal component analysis was used to visualize intersample variation among all of the genes from sorted subsets and illustrated that old ABCs have distinct gene-expression profiles compared with FO B cells from old and young mice (Supplemental Fig. 1B).

Based on their accumulation of somatic hypermutation, we hypothesized that ABCs represent a subset of Ag-experienced B cells whose accretion reflects the cumulative aggregate of challenges that drive their formation. To demonstrate that another subset of Ag-experienced B cells arising from deliberate infection also expresses T-bet and CD11c, we infected young mice with influenza. HA-specific B cells were tracked by binding to fluorescent labeled HA-PE. Prior to infection, the frequency of HA-reactive B cells was low (Fig. 5A), consistent with prior estimates of ∼1/50,000 splenic B cells (26). Following infection, mice displayed the expected weight loss and fully recovered 30 d later (Fig. 5B), indicating that the virus was cleared. At day 100 postinfection, HA-reactive B cells increased, and ∼25% of these were T-bet+CD11c+ (Fig. 5C, 5D). Collectively, these observations on influenza-infected mice support the analogy that some long-lived Ag-experienced cells express the phenotype associated with ABCs.

FIGURE 5.

B cells infected with influenza HA express T-bet and CD11c. Mice were infected intranasally with PR8, and spleens were harvested 100 d postinfection. (A) Representative profile of HA-PE–binding B cells (CD19+) in an uninfected spleen. Number shows the percentage of cells in highlighted box. (B) Weight loss and recovery postinfection. Error bars signify SD from eight mice. (C) Gating profile for HA-binding B cells in spleens at day 100 (upper panel). Number in box reflects the percentage of cells. A portion of HA-binding B cells express T-bet and CD11c (lower panel). (D) Number of B cells that bind HA. Total T-betCD11c and T-bet+CD11c+ cells (black circles) compared with only T-bet+CD11c+ cells (red circles). Data are from seven uninfected and eight d-100 infected mice collected in three independent experiments. *p < 0.01, uninfected versus infected groups, unpaired equal-variance Student t test.

FIGURE 5.

B cells infected with influenza HA express T-bet and CD11c. Mice were infected intranasally with PR8, and spleens were harvested 100 d postinfection. (A) Representative profile of HA-PE–binding B cells (CD19+) in an uninfected spleen. Number shows the percentage of cells in highlighted box. (B) Weight loss and recovery postinfection. Error bars signify SD from eight mice. (C) Gating profile for HA-binding B cells in spleens at day 100 (upper panel). Number in box reflects the percentage of cells. A portion of HA-binding B cells express T-bet and CD11c (lower panel). (D) Number of B cells that bind HA. Total T-betCD11c and T-bet+CD11c+ cells (black circles) compared with only T-bet+CD11c+ cells (red circles). Data are from seven uninfected and eight d-100 infected mice collected in three independent experiments. *p < 0.01, uninfected versus infected groups, unpaired equal-variance Student t test.

Close modal

These studies probe the origin and nature of ABCs, a B cell subset that steadily accumulates with age and whose surface phenotype and transcriptional signature were associated with humoral autoimmunity and antipathogen immune responses. We provide three lines of evidence that ABCs are a unique B cell subset. First, adoptive-transfer studies using MHC class II and CD40-deficient cells confirm our prior report that young FO B cells can give rise to ABCs after extensive division (9) and extend this observation in several ways. Notably, they demonstrate that ABCs can be generated with cognate T cell help, which is substantiated by the lack of ABCs in old CD154−/− mice. These findings also significantly connect T-bet expression with these extensively divided cells, a feature that is well established in ABC genesis (12, 13). In addition, our results indicate that homeostatic expansion is unlikely to be the major source of ABCs. The transfer experiments involved replete hosts, with minimal space for donor cells to fill by division, and the relatively few transferred cells from MHC class II– and CD40-deficient donors that divided did not express T-bet. Overall, our data support the hypothesis that most ABCs are the cumulative result of enduring environmental Ag stimulation through T-dependent mechanisms. However, the results do not exclude a TLR-mediated origin for some ABCs that may respond to viral or autoimmune stimuli (27).

Second, analyses of V gene segment use and somatic hypermutation indicate encounters with multiple Ags. The breadth of V-gene usage speaks against a monolithic origin in terms of Ag or epitope recognition and demonstrates that ABCs represent a cross-section of responses to a broad array of Ags. Numerous early reports suggested that the total B cell repertoire was restricted in old mice (5, 2831). However, our extensive analysis of 85 VH gene segments and 69 Vκ gene segments revealed a rich spectrum of V-gene usage by FO, MZ, and ABCs from old mice, indicating that global sequencing generates a more comprehensive view of the repertoire than do limited studies of Ag-specific cells. Only a handful of genes were over- or underused by ABCs; overall, there was no significant difference when ABCs were compared with FO or MZ repertoires. Such results would be expected if multiple heterogeneous Ags generated the diverse repertoire. Possible candidates are self-antigens or Ags of the microbiota environment. With regard to self-antigens, a previous report found that mice stimulated chronically with TLR7 agonists developed ABCs expressing high titers of anti-Smith autoimmune Abs (10). However, the few genes that were overused in ABCs in this study did not possess positively charged CDR3 regions in their rearranged sequences, which are common in self-reactive Abs (32). It appears that healthy old mice without deliberate immunization develop a heterogeneous repertoire, without propensity for autoimmune Abs. Although the repertoires of ABCs, FO cells, and MZ cells were similar, the results of the mutation analyses were strikingly different. ABCs showed clear evidence of mutation in the VH and Vκ exons compared with FO and MZ B cells. Although FO B cells had the lowest mutation frequency, consistent with their preimmune status, the frequency was 2-fold higher in MZ cells, which have likely encountered microbial Ags during circulation, and 4-fold higher in ABCs. In ABCs, mutations did not accumulate in CDRs, consistent with the lack of selection of certain V genes in the repertoire analysis. Furthermore, there was a significant increase in mutation frequency in ABCs in the noncoding JH4 intron, which is a broad substrate for hypermutation in the absence of selection (33). However, it remains uncertain whether all ABCs are the products of germinal center reactions. For example, most ABCs have IgM receptors and continue to express the surface receptor TACI (9), both of which are inconsistent with the germinal center B cell phenotypes (34). Moreover, the mutation frequency in ABCs was lower compared with germinal center B cells from immunized mice (24). There is precedent for mutated IgM-bearing cells occurring in the absence of germinal centers (35, 36). Alternatively, ABCs may represent early germinal center emigrants that exit before concerted selection (37, 38).

Third, microarray analyses of gene expression show that ABCs from old mice are unique in relation to FO cells from young and old mice. Transcription analysis was also performed by Rubtsov et al. (10) to compare old ABCs with old FO, old B1, and young B1 cells. However, the two analyses profile different sets of genes because the cells were isolated under different conditions. The ABCs in the study from Rubtsov et al. (10) were sorted for CD11b+ expression, and the ABCs analyzed in this study were sorted as CD21 CD23. Nonetheless, both analyses show that T-bet and CD11c are greatly increased in old ABCs relative to old FO cells. We found that this signature was also present in some long-lived B cells following influenza infection 100 d later, confirming our previous report (39). By analogy, ABCs are Ag experienced, because they have increased somatic hypermutation, and they require T cell interactions for their generation. The cells presumably arise from chronic stimulation by endogenous Ags, but it is important to note that ABCs are resting cells that persist over time. Whether they can undergo recall responses when they encounter cognate Ags remains to be determined.

We thank Robert Maul, Ranjan Sen, and Diana Castiblanco for stimulating conversations and advice, William Yang for technical assistance, and Kate Turlington, Haleigh Larson, and Arielle Kilner for experiments. Robert Wersto, Jade Scheers, Tonya Wallace, and Cuong Nguyen (National Institute on Aging Flow Cytometry Core Facility) assisted in cell sorting. Michael Paley and E. John Wherry (University of Pennsylvania) assisted with the ABC/FO microarray and initial analyses.

This work was supported by the Intramural Research Program of the National Institutes of Health, National Institute on Aging (to L.M.R.K. and P.J.G.) and by Grants PR130769 (Department of the Army) and R01 AG030227 (National Institute on Aging) (to M.P.C.). M.S.N. was supported in part by National Institute of Allergy and Infectious Diseases Grant T32 AI055428.

The online version of this article contains supplemental material.

Abbreviations used in this article:

ABC

age-associated B cell

AID

activation-induced deaminase

FO

follicular

HA

hemagglutinin

HA-PE

PE-labeled probe that recognizes HA of PR8

MZ

marginal zone

PR8

influenza strain A/Puerto Rico/8/1934.

1
Miller
J. P.
,
Allman
D.
.
2003
.
The decline in B lymphopoiesis in aged mice reflects loss of very early B-lineage precursors.
J. Immunol.
171
:
2326
2330
.
2
Cancro
M. P.
2005
.
B cells and aging: gauging the interplay of generative, selective, and homeostatic events.
Immunol. Rev.
205
:
48
59
.
3
Miller
J. P.
,
Cancro
M. P.
.
2007
.
B cells and aging: balancing the homeostatic equation.
Exp. Gerontol.
42
:
396
399
.
4
Kline
G. H.
,
Hayden
T. A.
,
Klinman
N. R.
.
1999
.
B cell maintenance in aged mice reflects both increased B cell longevity and decreased B cell generation.
J. Immunol.
162
:
3342
3349
.
5
Johnson
S. A.
,
Rozzo
S. J.
,
Cambier
J. C.
.
2002
.
Aging-dependent exclusion of antigen-inexperienced cells from the peripheral B cell repertoire.
J. Immunol.
168
:
5014
5023
.
6
Zheng
B.
,
Han
S.
,
Takahashi
Y.
,
Kelsoe
G.
.
1997
.
Immunosenescence and germinal center reaction.
Immunol. Rev.
160
:
63
77
.
7
Frasca
D.
,
Blomberg
B. B.
.
2011
.
Aging impairs murine B cell differentiation and function in primary and secondary lymphoid tissues.
Aging Dis.
2
:
361
373
.
8
Goenka
R.
,
Scholz
J. L.
,
Naradikian
M. S.
,
Cancro
M. P.
.
2014
.
Memory B cells form in aged mice despite impaired affinity maturation and germinal center kinetics.
Exp. Gerontol.
54
:
109
115
.
9
Hao
Y.
,
O’Neill
P.
,
Naradikian
M. S.
,
Scholz
J. L.
,
Cancro
M. P.
.
2011
.
A B-cell subset uniquely responsive to innate stimuli accumulates in aged mice.
Blood
118
:
1294
1304
.
10
Rubtsov
A. V.
,
Rubtsova
K.
,
Fischer
A.
,
Meehan
R. T.
,
Gillis
J. Z.
,
Kappler
J. W.
,
Marrack
P.
.
2011
.
Toll-like receptor 7 (TLR7)-driven accumulation of a novel CD11c+ B-cell population is important for the development of autoimmunity.
Blood
118
:
1305
1315
.
11
Ratliff
M.
,
Alter
S.
,
Frasca
D.
,
Blomberg
B. B.
,
Riley
R. L.
.
2013
.
In senescence, age-associated B cells secrete TNFα and inhibit survival of B-cell precursors.
Aging Cell
12
:
303
311
.
12
Rubtsova
K.
,
Rubtsov
A. V.
,
Cancro
M. P.
,
Marrack
P.
.
2015
.
Age-associated B cells: a T-bet-dependent effector with roles in protective and pathogenic immunity.
J. Immunol.
195
:
1933
1937
.
13
Rubtsova
K.
,
Rubtsov
A. V.
,
van Dyk
L. F.
,
Kappler
J. W.
,
Marrack
P.
.
2013
.
T-box transcription factor T-bet, a key player in a unique type of B-cell activation essential for effective viral clearance.
Proc. Natl. Acad. Sci. USA
110
:
E3216
E3224
.
14
Rubtsov
A. V.
,
Rubtsova
K.
,
Kappler
J. W.
,
Jacobelli
J.
,
Friedman
R. S.
,
Marrack
P.
.
2015
.
CD11c-expressing B cells are located at the T cell/B cell border in spleen and are potent APCs.
J. Immunol.
195
:
71
79
.
15
Ratliff
M.
,
Alter
S.
,
McAvoy
K.
,
Frasca
D.
,
Wright
J. A.
,
Zinkel
S. S.
,
Khan
W. N.
,
Blomberg
B. B.
,
Riley
R. L.
.
2015
.
In aged mice, low surrogate light chain promotes pro-B-cell apoptotic resistance, compromises the PreBCR checkpoint, and favors generation of autoreactive, phosphorylcholine-specific B cells.
Aging Cell
14
:
382
390
.
16
Claes
N.
,
Fraussen
J.
,
Vanheusden
M.
,
Hellings
N.
,
Stinissen
P.
,
Van Wijmeersch
B.
,
Hupperts
R.
,
Somers
V.
.
2016
.
Age-associated B cells with proinflammatory characteristics are expanded in a proportion of multiple sclerosis patients.
J. Immunol.
197
:
4576
4583
.
17
Moir
S.
,
Ho
J.
,
Malaspina
A.
,
Wang
W.
,
DiPoto
A. C.
,
O’Shea
M. A.
,
Roby
G.
,
Kottilil
S.
,
Arthos
J.
,
Proschan
M. A.
, et al
.
2008
.
Evidence for HIV-associated B cell exhaustion in a dysfunctional memory B cell compartment in HIV-infected viremic individuals.
J. Exp. Med.
205
:
1797
1805
.
18
Weiss
G. E.
,
Crompton
P. D.
,
Li
S.
,
Walsh
L. A.
,
Moir
S.
,
Traore
B.
,
Kayentao
K.
,
Ongoiba
A.
,
Doumbo
O. K.
,
Pierce
S. K.
.
2009
.
Atypical memory B cells are greatly expanded in individuals living in a malaria-endemic area.
J. Immunol.
183
:
2176
2182
.
19
Zinöcker
S.
,
Schindler
C. E.
,
Skinner
J.
,
Rogosch
T.
,
Waisberg
M.
,
Schickel
J. N.
,
Meffre
E.
,
Kayentao
K.
,
Ongoïba
A.
,
Traoré
B.
,
Pierce
S. K.
.
2015
.
The V gene repertoires of classical and atypical memory B cells in malaria-susceptible West African children.
J. Immunol.
194
:
929
939
.
20
Portugal
S.
,
Tipton
C. M.
,
Sohn
H.
,
Kone
Y.
,
Wang
J.
,
Li
S.
,
Skinner
J.
,
Virtaneva
K.
,
Sturdevant
D. E.
,
Porcella
S. F.
, et al
.
2015
.
Malaria-associated atypical memory B cells exhibit markedly reduced B cell receptor signaling and effector function.
ELife
8
:
4
.
21
Yates
J. L.
,
Racine
R.
,
McBride
K. M.
,
Winslow
G. M.
.
2013
.
T cell-dependent IgM memory B cells generated during bacterial infection are required for IgG responses to antigen challenge.
J. Immunol.
191
:
1240
1249
.
22
Wang
Z.
,
Raifu
M.
,
Howard
M.
,
Smith
L.
,
Hansen
D.
,
Goldsby
R.
,
Ratner
D.
.
2000
.
Universal PCR amplification of mouse immunoglobulin gene variable regions: the design of degenerate primers and an assessment of the effect of DNA polymerase 3′ to 5′ exonuclease activity.
J. Immunol. Methods
233
:
167
177
.
23
Whittle
J. R.
,
Wheatley
A. K.
,
Wu
L.
,
Lingwood
D.
,
Kanekiyo
M.
,
Ma
S. S.
,
Narpala
S. R.
,
Yassine
H. M.
,
Frank
G. M.
,
Yewdell
J. W.
, et al
.
2014
.
Flow cytometry reveals that H5N1 vaccination elicits cross-reactive stem-directed antibodies from multiple Ig heavy-chain lineages.
J. Virol.
88
:
4047
4057
.
24
Zanotti
K. J.
,
Maul
R. W.
,
Castiblanco
D. P.
,
Yang
W.
,
Choi
Y. J.
,
Fox
J. T.
,
Myung
K.
,
Saribasak
H.
,
Gearhart
P. J.
.
2015
.
ATAD5 deficiency decreases B cell division and Igh recombination.
J. Immunol.
194
:
35
42
.
25
Scholz
J. L.
,
Crowley
J. E.
,
Tomayko
M. M.
,
Steinel
N.
,
O’Neill
P. J.
,
Quinn
W. J.
 III
,
Goenka
R.
,
Miller
J. P.
,
Cho
Y. H.
,
Long
V.
, et al
.
2008
.
BLyS inhibition eliminates primary B cells but leaves natural and acquired humoral immunity intact.
Proc. Natl. Acad. Sci. USA
105
:
15517
15522
.
26
Cancro
M. P.
,
Gerhard
W.
,
Klinman
N. R.
.
1978
.
The diversity of the influenza-specific primary B-cell repertoire in BALB/c mice.
J. Exp. Med.
147
:
776
787
.
27
Naradikian
M. S.
,
Hao
Y.
,
Cancro
M. P.
.
2016
.
Age-associated B cells: key mediators of both protective and autoreactive humoral responses.
Immunol. Rev.
269
:
118
129
.
28
Viale
A. C.
,
Chies
J. A.
,
Huetz
F.
,
Malenchere
E.
,
Weksler
M.
,
Freitas
A. A.
,
Coutinho
A.
.
1994
.
VH-gene family dominance in ageing mice.
Scand. J. Immunol.
39
:
184
188
.
29
Ben-Yehuda
A.
,
Szabo
P.
,
LeMaoult
J.
,
Manavalan
J. S.
,
Weksler
M. E.
.
1998
.
Increased VH 11 and VH Q52 gene use by splenic B cells in old mice associated with oligoclonal expansions of CD5+ B cells.
Mech. Ageing Dev.
103
:
111
121
.
30
Riley
S. C.
,
Froscher
B. G.
,
Linton
P. J.
,
Zharhary
D.
,
Marcu
K.
,
Klinman
N. R.
.
1989
.
Altered VH gene segment utilization in the response to phosphorylcholine by aged mice.
J. Immunol.
143
:
3798
3805
.
31
LeMaoult
J.
,
Delassus
S.
,
Dyall
R.
,
Nikolić-Zugić
J.
,
Kourilsky
P.
,
Weksler
M. E.
.
1997
.
Clonal expansions of B lymphocytes in old mice.
J. Immunol.
159
:
3866
3874
.
32
Wardemann
H.
,
Yurasov
S.
,
Schaefer
A.
,
Young
J. W.
,
Meffre
E.
,
Nussenzweig
M. C.
.
2003
.
Predominant autoantibody production by early human B cell precursors.
Science
301
:
1374
1377
.
33
Lebecque
S. G.
,
Gearhart
P. J.
.
1990
.
Boundaries of somatic mutation in rearranged immunoglobulin genes: 5′ boundary is near the promoter, and 3′ boundary is approximately 1 kb from V(D)J gene.
J. Exp. Med.
172
:
1717
1727
.
34
Goenka
R.
,
Matthews
A. H.
,
Zhang
B.
,
O’Neill
P. J.
,
Scholz
J. L.
,
Migone
T. S.
,
Leonard
W. J.
,
Stohl
W.
,
Hershberg
U.
,
Cancro
M. P.
.
2014
.
Local BLyS production by T follicular cells mediates retention of high affinity B cells during affinity maturation.
J. Exp. Med.
211
:
45
56
.
35
Voigt
I.
,
Camacho
S. A.
,
de Boer
B. A.
,
Lipp
M.
,
Förster
R.
,
Berek
C.
.
2000
.
CXCR5-deficient mice develop functional germinal centers in the splenic T cell zone.
Eur. J. Immunol.
30
:
560
567
.
36
Di Niro
R.
,
Lee
S. J.
,
Vander Heiden
J. A.
,
Elsner
R. A.
,
Trivedi
N.
,
Bannock
J. M.
,
Gupta
N. T.
,
Kleinstein
S. H.
,
Vigneault
F.
,
Gilbert
T. J.
, et al
.
2015
.
Salmonella infection drives promiscuous B cell activation followed by extrafollicular affinity maturation.
Immunity
43
:
120
131
.
37
Pape
K. A.
,
Taylor
J. J.
,
Maul
R. W.
,
Gearhart
P. J.
,
Jenkins
M. K.
.
2011
.
Different B cell populations mediate early and late memory during an endogenous immune response.
Science
331
:
1203
1207
.
38
Weisel
F. J.
,
Zuccarino-Catania
G. V.
,
Chikina
M.
,
Shlomchik
M. J.
.
2016
.
A temporal switch in the germinal center determines differential output of memory B and plasma cells.
Immunity
44
:
116
130
.
39
Naradikian
M. S.
,
Myles
A.
,
Beiting
D. P.
,
Roberts
K. J.
,
Dawson
L.
,
Herati
R. S.
,
Bengsch
B.
,
Linderman
S. L.
,
Stelekati
E.
,
Spolski
R.
, et al
.
2016
.
Cutting edge: IL-4, IL-21, and IFN-γ interact to govern T-bet and CD11c expression in TLR-activated B cells.
J. Immunol.
197
:
1023
1028
.

The authors have no financial conflicts of interest.

Supplementary data