X-linked lymphoproliferative disease (XLP) is a severe immunodeficiency associated with a marked reduction in circulating memory B cells. Our investigation of the B cell compartment of XLP patients revealed an increase in the frequency of a population of B cells distinct from those previously defined. This population displayed increased expression of CD10, CD24, and CD38, indicating that it could consist of circulating immature/transitional B cells. Supporting this possibility, CD10+CD24highCD38high B cells displayed other immature characteristics, including unmutated Ig V genes and elevated levels of surface IgM; they also lacked expression of Bcl-2 and a panel of activation molecules. The capacity of CD24highCD38high B cells to proliferate, secrete Ig, and migrate in vitro was greatly reduced compared with mature B cell populations. Moreover, CD24highCD38high B cells were increased in the peripheral blood of neonates, patients with common variable immunodeficiency, and patients recovering from hemopoietic stem cell transplant. Thus, an expansion of functionally immature B cells may contribute to the humoral immunodeficient state that is characteristic of neonates, as well as patients with XLP or common variable immunodeficiency, and those recovering from a stem cell transplant. Further investigation of transitional B cells will improve our understanding of human B cell development and how alterations to this process may precipitate immunodeficiency or autoimmunity.

B cell development occurs in the fetal liver and adult bone marrow (BM)3 and involves the sequential differentiation of stem cells into pro-B, pre-B, and then immature B cells (1, 2). Immature B cells exported from the BM enter a transitional phase during which further maturation events occur to produce mature cells (1, 2, 3, 4). The transitional B cell, therefore, represents an intermediate stage of development and, as such, is susceptible to positive and negative selection pressures (3, 4, 5, 6, 7, 8, 9). In the mouse, three populations of transitional B cells (T1, T2, T3) have been identified phenotypically (3, 4, 10). T1 B cells are CD24highIgMhighIgDCD21CD23CD93+, T2 B cells are CD24highIgMhighIgDhighCD21+CD23+CD93+, while T3 B cells are CD24highIgMlowIgDhighCD21+CD23+CD93+ (3). These cells can be resolved from mature splenic follicular and marginal zone B cells which are CD24lowIgMlowIgDhighCD21lowCD23+CD93 and CD24lowIgM+IgDCD21+CD23CD93, respectively (8, 11). Phenotypically distinct subsets of human B cells can also be identified in different lymphoid tissues. Thus, immature B cells are CD19+CD27CD10+IgM+IgD, while naive B cells are CD19+CD27IgMlowIgDhigh, and memory B cells are CD19+CD27+ and express IgM, IgG, or IgA (2, 12, 13, 14, 15). Human transitional B cells, however, remain poorly characterized, although their existence is suggested by the recent demonstration of a population of cells in peripheral blood (PB) distinguishable from mature B cells on the basis of a CD24highCD38high phenotype (16, 17).

B cell development is a tightly regulated process. If aberrations occur during this process, perturbations to B cell homeostasis may ensue. Indeed, pregerminal center (GC) or GC founder cells and plasma cells (PC) have been aberrantly detected in PB of patients with systemic lupus erythematosus (SLE) (18, 19), while in immunodeficiencies such as common variable immunodeficiency (CVID; Refs.20, 21, 22) and hyper-IgM syndrome (23, 24), as well as patients recovering from hemopoietic stem cell transplantation (HSCT) (25), there is a paucity of circulating memory (CD27+) B cells. Recently, we demonstrated a deficiency in the number of memory B cells in patients with X-linked lymphoproliferative disease (XLP; Refs.21, 26), an immunodeficiency caused by mutations in SH2D1A (27, 28, 29) and characterized by fulminant infectious mononucleosis, hypogammaglobulinemia, and malignant lymphoma (26).

Further investigation of naive (CD27) B cells from XLP patients revealed that a substantial proportion of them exhibited a phenotype (i.e., CD10+CD24highCD38highCD5+bcl-2) distinct from other defined B cell subsets. A similar population of B cells was also detectable in healthy individuals, albeit at a ∼5-fold lower frequency than XLP, as well as in normal BM and cord blood (CB). The phenotype of this population resembled that of cells recently proposed to be human transitional B cells (16, 17). In the current study, transitional B cells were found to display functional characteristics of immature B cells, such as the lack of expression of Bcl-2 and reduced survival, proliferation, differentiation, and chemotaxis compared with mature B cells; they also expressed unmutated Ig V region genes. Thus, in addition to a deficiency in memory B cells, circulating immature B cells–resembling putative transitional B cells–are substantially increased in XLP patients. Moreover, higher numbers of these B cells were found in the blood of neonates, some CVID patients, and patients recovering from HSCT. This latter finding confirmed that these cells are BM-derived transitional B cells. In other words, a common feature of immunodeficiency states characterized by impaired humoral immunity is the predominance of functionally immature cells in the peripheral B cell compartment. Such defects in B cell differentiation in vivo–decreased memory and increased transitional B cells–not only explain the hypogammaglobulinemia characteristic of these conditions, but also suggest that methods for enhancing their differentiation into mature effector cells in vivo may alleviate the hypogammaglobulinemic state of such individuals.

The following mAbs were used: FITC-anti-CD19, anti-CD20, and anti-CD27; PE-anti-CD5, anti-CD19, anti-HLA-DR and anti-Bcl-2; allophycocyanin-anti-CD19 (BD Immunocytometry Systems); PE-anti-CD21, anti-CD25, anti-CD27, anti-CD80, anti-CD86, anti-CD95, anti-IgM, anti-IgD, anti-CXCR4; allophycocyanin-anti-CD10; biotinylated-anti-CD44, anti-IgD, anti-IgM, and anti-IgG; streptavidin-PerCP (BD Pharmingen); FITC-anti-CD23, anti-CD24; PE-anti-CD20, anti-CD22, anti-CD23, anti-CD24, anti-CD38, anti-CD62L, anti-CD69; allophycocyanin-anti-CD20, anti-CD38; anti-CD38, isotype controls (Caltag Laboratories); PE-anti-CD40 (provided by J. Banchereau, Schering-Plough Laboratory of Immunological Research, Dardilly, France); PE-anti-CD9; biotinylated-anti-CD27 (eBioscience); biotinylated anti-IgA (Southern Biotechnology Associates); biotinylated anti-CXCR5 (R&D Systems); B cell activating factor of the TNF family (BAFF), anti-BAFF receptor (BAFF-R) (Biogen Idec; Ref.30); goat biotinylated anti-human transmembrane activator and calcium modulator and cyclophilin ligand interactor (TACI) antiserum (PeproTech).

Normal spleens were obtained from cadaveric organ donors (Australian Red Cross Blood Service). PB samples were collected from normal healthy donors, XLP and CVID patients, and patients recovering from HSCT, following informed consent. Patients were diagnosed with CVID if they had marked decreases in their serum levels of two of the three major Ig isotypes (IgM, IgG, IgA) and documented evidence of recurrent and/or opportunistic infections arising from deficient humoral and cellular immunity in the absence of any known genetic, environmental, or other medical cause (see 〈www.esid.org/〉). The XLP patients used in this study were between 12 and 49 years old, and have been previously described (XLP nos. 1–3, 10, 11, 15, 16; Ref.21). CB samples were collected from King George V Hospital for Mothers and Babies (Sydney, Australia). BM aspirates from healthy donors and tonsils and lymph node (LN) samples were collected from patients at Royal Prince Alfred Hospital (Sydney, Australia). Institutional human ethics review committees approved all studies described. Mononuclear cells (MNCs) were prepared as previously described (30) and cryopreserved in liquid nitrogen until required. Human B cells were isolated using a B Cell Negative Isolation kit (Dynal Biotech). B cell subsets were isolated by sorting on a FACSVantage (BD Biosciences) after labeling purified total B cells with anti-CD24 and anti-CD38 mAb to identify CD24highCD38high, CD24+CD38+ (“naive”), and CD24+CD38 (“memory”) B cells, respectively, or with anti-CD20 and anti-CD27 mAb to identify naive and memory B cells (31).

Cells were incubated with anti-CD24, anti-CD38, and anti-CD19 mAb, and mAb to molecules of interest, to allow phenotyping of CD24highCD38high B cells. Expression of intracellular Bcl-2 was determined as described (32). Flow cytometric acquisition was performed on a FACSCalibur (BD Biosciences) and was analyzed using FlowJo (Tree Star) software. Fluorescence was measured on a log10 scale.

The VH5 genes were amplified from cDNA prepared from sorted B cells by nested PCR as previously described (32). Nucleotide sequences were analyzed using the Sequencher version 4.5 program (Gene Codes), and comparisons were performed using the GenBank database.

To determine proliferation, 5 × 103 B cells were cultured in 125 μl in round-bottom, 96-well plates in B cell medium (31) or with rCD40L (31) and/or F(ab′)2 of goat anti-human Ig (Jackson ImmunoResearch Laboratories). Plates were pulsed with 1 μCi [3H]thymidine after various times of activation and harvested 8 h later. Scintillation counting was performed on a beta plate counter (Pharmacia-LKB). Sort-purified B cell populations (10 × 103 cells/200 μl) were cultured in round-bottom, 96-well plates in B cell medium alone, or with CD40L, IL-10 (100 U/ml; provided by Dr. R. de Waal Malefyt, DNAX Research Institute, Palo Alto, CA), and/or Staphylococcus aureus Cowan (SAC) particles (Calbiochem) (0.01%). After 14 days, supernatants were collected and the level of secreted Ig was determined by ELISA (30, 31).

Chemotaxis assays using human CXCL12 (100 ng/ml; PeproTech), CXCL13 (3 μg/ml; R&D Systems), or CCL21 (600 ng/ml; PeproTech) were performed as previously described (33). The migrated population was labeled with anti-CD19, anti-CD24, and anti-CD38 mAb to resolve B cell subsets. The absolute number of each migrated subset was calculated and expressed as the percent of input cells.

Data were analyzed using unpaired t tests and ANOVA with Prism software (GraphPad Software).

The frequency and absolute number of total PB B cells in XLP patients is normal (21, 26). However, XLP patients have a marked reduction in memory (CD27+) B cells (Fig. 1 a) (21, 26). It was recently reported that some patients with CVID not only lacked memory B cells, but also B cells with an immature phenotype were occasionally detected (20). This prompted us to analyze the B cell compartment of XLP patients in greater detail to determine whether their CD27 B cells were phenotypically similar to those in normal donors. To do this, we examined surface molecules (CD10, CD24, CD38) that are present at the pre-/immature B cell stage of development (2, 12, 34).

FIGURE 1.

XLP patients have an increased frequency and number of CD24highCD38high B cells. a and b, PBMC from a normal donor and an XLP patient were labeled with mAb specific for (a) CD19 and CD27 or (b) CD24 and CD38. a, Naive and memory B cells or (b) CD24highCD38high B cells were then detected. The values in b represent the percentage of B cells that exhibit a CD24highCD38high phenotype. c, PBMC from a normal donor or XLP patient were labeled with mAb specific for CD24, CD38 and CD10. Expression of CD10 on naive (dashed line), memory (solid line), and CD24highCD38high B cells (gray shading) was determined by gating on CD24+CD38+, CD24+CD38, and CD24highCD38high B cells, respectively. Note, expression of CD10 on memory B cells from XLP patients is not presented due to a lack of these cells from the PB of XLP patients. The gray outlined histogram represents the fluorescence of cells labeled with isotype control mAb. Fluorescence was measured on a log10 scale. d and e, PBMC from 15 normal donors and 7 XLP patients were labeled with anti-CD19, anti-CD24, and anti-CD38 mAb, and the (d) frequency and (e) number of CD24highCD38high B cells was determined. The graphs show data points for all donors and patients examined with the mean represented by horizontal lines. Significant differences are indicated (∗∗, p < 0.01; ∗∗∗, p < 0.001).

FIGURE 1.

XLP patients have an increased frequency and number of CD24highCD38high B cells. a and b, PBMC from a normal donor and an XLP patient were labeled with mAb specific for (a) CD19 and CD27 or (b) CD24 and CD38. a, Naive and memory B cells or (b) CD24highCD38high B cells were then detected. The values in b represent the percentage of B cells that exhibit a CD24highCD38high phenotype. c, PBMC from a normal donor or XLP patient were labeled with mAb specific for CD24, CD38 and CD10. Expression of CD10 on naive (dashed line), memory (solid line), and CD24highCD38high B cells (gray shading) was determined by gating on CD24+CD38+, CD24+CD38, and CD24highCD38high B cells, respectively. Note, expression of CD10 on memory B cells from XLP patients is not presented due to a lack of these cells from the PB of XLP patients. The gray outlined histogram represents the fluorescence of cells labeled with isotype control mAb. Fluorescence was measured on a log10 scale. d and e, PBMC from 15 normal donors and 7 XLP patients were labeled with anti-CD19, anti-CD24, and anti-CD38 mAb, and the (d) frequency and (e) number of CD24highCD38high B cells was determined. The graphs show data points for all donors and patients examined with the mean represented by horizontal lines. Significant differences are indicated (∗∗, p < 0.01; ∗∗∗, p < 0.001).

Close modal

By using CD24 and CD38, three B cell populations could be resolved in normal individuals: CD24+CD38 and CD24+CD38+ cells (Fig. 1,b), which accounted for >95% of B cells, and a CD24highCD38high subset (Fig. 1,b) which comprised only 2.56 ± 0.20% (mean ± SEM; n = 15) of B cells (Fig. 1,d). In contrast, the PB of XLP patients contained only two B cell populations: CD24+CD38+ and CD24highCD38high, with the latter comprising 15.3% of B cells (n = 7; Fig. 1,d; p < 0.001 compared with healthy donors). There was also a significant increase in the number of CD24highCD38high B cells in XLP patients compared with normal donors (∼3-fold; Fig. 1,e). CD24highCD38high B cells present in PB of both normal controls and XLP patients uniformly expressed CD10, while CD24+CD38+ and CD24+CD38 B cells did not (Fig. 1,c). It was recently proposed that PB B cells with a CD10+CD24highCD38high phenotype represent transitional, or newly emigrated, B cells (16, 17, 34). Thus, the B cell compartment of XLP patients is characterized not only by a reduction in memory B cells but also an expansion in the number of transitional B cells. The absence of CD24+CD38 B cells in XLP patients suggested that this population contains predominantly memory B cells (21), while the CD24+CD38+ cells most likely correspond to naive B cells (Fig. 1 b).

The phenotype of PB transitional B cells was compared with CD24+CD38+ and CD24+CD38 B cells, which, for convenience, will be referred to as naive and memory B cells, respectively. Pan B cell markers (CD19, CD21, CD22, CD40, CD62L, HLA-DR) were expressed at similar levels on all three B cell populations. However, CD20 was significantly higher on CD24highCD38high B cells (mean fluorescence intensity (MFI): 1466 ± 270; n = 3) compared with naive (MFI: 551 ± 132) and memory (MFI: 617 ± 214) B cells (Fig. 2). CD23 was on naive and transitional, but down-regulated on memory, cells, consistent with previous studies of naive and memory B cells present in human spleen and tonsils (13, 14, 15, 32).

FIGURE 2.

The phenotype of CD24high CD38high B cells indicates a distinct subset of B cells. PBMCs from a normal donor were labeled with anti-CD19, CD24, and CD38 mAb. Electronic gates were set on CD24+CD38+ (naive), CD24+CD38 (memory), and CD24highCD38high B cells, as depicted in Fig. 1 b, and expression of the indicated molecules on these subsets was determined using a fourth fluorescence channel. The fluorescence of cells incubated with an isotype control mAb is indicated by the unfilled histogram and expression of the molecule of interest is indicated by the shaded histogram. These results are representative of data obtained from at least three different donors.

FIGURE 2.

The phenotype of CD24high CD38high B cells indicates a distinct subset of B cells. PBMCs from a normal donor were labeled with anti-CD19, CD24, and CD38 mAb. Electronic gates were set on CD24+CD38+ (naive), CD24+CD38 (memory), and CD24highCD38high B cells, as depicted in Fig. 1 b, and expression of the indicated molecules on these subsets was determined using a fourth fluorescence channel. The fluorescence of cells incubated with an isotype control mAb is indicated by the unfilled histogram and expression of the molecule of interest is indicated by the shaded histogram. These results are representative of data obtained from at least three different donors.

Close modal

Naive and transitional B cells were similar with respect to expression of surface Ig isotypes inasmuch as >95% of these cells were IgM+IgD+ (Fig. 2). Despite this similarity, the level of IgM on transitional B cells was consistently higher (up to 5-fold) than that on naive B cells (Fig. 2). In contrast, ∼50% of memory (CD24+CD38) B cells expressed IgM and IgD, and Ig isotype-switched B cells were largely restricted to this subset. Specifically, ∼20% and ∼15% of the CD24+CD38 memory cells expressed IgG or IgA, respectively, while <5% of naive and <0.2% of transitional B cells expressed these switched Ig isotypes (Fig. 2). This is consistent with our previous studies of naive and memory B cells, identified in PB by the differential expression of CD27, where ∼40% of CD27+ B cells expressed IgG or IgA, while <4% of CD27 B cells had this phenotype (21, 61). The B cell activation markers CD27, CD25, CD69, CD80, CD86, and CD95 were not expressed by CD24highCD38high B cells or by CD24+CD38+ B cells (Fig. 2). In contrast, CD27 was expressed on most CD24+CD38 B cells (Fig. 2), confirming their designation as memory B cells. CD24+CD38 B cells also expressed low but detectable levels of CD25, CD80, CD86, and CD95 (Fig. 2), again consistent with the phenotype of memory B cells in human spleen and tonsils (13, 14, 32). The adhesion molecule CD44 (Fig. 2) was expressed at lower levels on CD24highCD38high B cells (MFI CD44: 110 ± 11) than on mature B cells (naive: 271 ± 14.2; memory: 424 ± 30). However, CD9, which is highly expressed on PC (33), was significantly up-regulated on CD24highCD38high B cells relative to its expression on mature B cells (Fig. 2).

Several other molecules whose expression changes during B cell maturation were also examined. BAFF bound to all B cell subsets (Fig. 2). Binding of BAFF to transitional and naive B cells was most likely mediated by BAFF-R, while binding to memory B cells may involve both BAFF-R and TACI (Fig. 2). Strikingly, the antiapoptotic molecule Bcl-2 was virtually absent from CD24highCD38high B cells compared with both mature B cell subsets (Fig. 2). Bcl-2 expression increased as maturation progressed from the CD24highCD38high→naive→memory stages (Fig. 2), consistent with previous studies which showed an increase in Bcl-2 during B cell differentiation (32, 35, 36). Lastly, CD5, which is commonly regarded as a marker of murine B1 B cells (37), was uniformly expressed by all CD24highCD38high B cells, whereas it was detected on only ∼20% of naive and <10% of memory B cells (Fig. 2). Thus, CD24highCD38high B cells have a phenotype that distinguishes them from other well-defined peripheral B cell populations. When the phenotype of the CD24highCD38high B cells in the PB of XLP patients was examined, it was found to be identical to that of such cells present in normal healthy donors (data not shown). This demonstrates that the population of B cells that are expanded in XLP are bona fide transitional B cells, rather than another subset of peripheral B cells.

The presence of transitional B cells in different lymphoid tissues was next investigated. BM and CB contained an increased frequency of CD24highCD38high B cells compared with adult PB (Fig. 3,a). In contrast, the frequency of CD24highCD38high B cells in secondary lymphoid tissues was less than in PB; 1.8 ± 0.7% (n = 11) of splenic and 1.41 ± 0.7% (n = 5) of tonsillar B cells had this phenotype, while they were virtually absent in LN (∼0.01%; Fig. 3 a).

FIGURE 3.

Identification of CD24high CD38high B cells in immune tissues of healthy individuals. a, MNCs from BM, CB, PB, spleen, tonsil, and LN were labeled with anti-CD19, CD24, and CD38 mAb to identify and quantitate CD24highCD38high B cells. The values indicate the mean frequency (± SEM) of CD24highCD38high B cells, defined by the illustrated gate, detected in the indicated number of tissue samples. b, MNCs from cord blood were labeled with anti-CD19, CD24 and CD38 mAb. The CD24highCD38high and CD24+CD38+ B cell populations were identified using the indicated gates (left panel) and expression of the indicated molecules on both populations was then determined (right panel). Fluorescence was measured on a log10 scale.

FIGURE 3.

Identification of CD24high CD38high B cells in immune tissues of healthy individuals. a, MNCs from BM, CB, PB, spleen, tonsil, and LN were labeled with anti-CD19, CD24, and CD38 mAb to identify and quantitate CD24highCD38high B cells. The values indicate the mean frequency (± SEM) of CD24highCD38high B cells, defined by the illustrated gate, detected in the indicated number of tissue samples. b, MNCs from cord blood were labeled with anti-CD19, CD24 and CD38 mAb. The CD24highCD38high and CD24+CD38+ B cell populations were identified using the indicated gates (left panel) and expression of the indicated molecules on both populations was then determined (right panel). Fluorescence was measured on a log10 scale.

Close modal

Due to the increased frequency of CD24highCD38high B cells in CB compared with adult PB (Fig. 3,a) and the expression of CD5 on PB CD24highCD38high (Fig. 2) and CB B cells (38), it was important to distinguish transitional B cells from CD5+ B cells in CB. This was achieved by two complementary approaches. First, we hypothesized that if transitional B cells are distinct from CB B cells, then they should exhibit a distinct phenotype. For this experiment, expression of CD5, CD9, CD44, IgM, IgD, and Bcl-2 on CB CD24+CD38+ and CD24highCD38high B cells was determined because they were differentially expressed by transitional and mature PB B cells (see Fig. 2). CD24highCD38high CB B cells were CD5highCD9+CD44low IgMhighBcl-2 (Fig. 3,b), similar to those in PB (Fig. 2), while the CD24+CD38+ CB B cells exhibited a different phenotype (CD5+CD9lowCD44highIgM+Bcl-2+; Fig. 3 b).

Second, we assumed that if CD24highCD38high cells are transitional, they would arise in the BM and some CD5+ cells would be detectable within the CD24highCD38high BM population. We found a similar proportion of BM CD24highCD38high B cells expressing both CD5 and IgD (∼15.2%; data not shown). These cells are likely to be the BM counterparts of PB transitional B cells. In contrast, CD5IgDCD24highCD38high B cells, which comprised the majority of BM CD24highCD38high B cells (data not shown), presumably correspond to progenitor or immature B cells (16), and could give rise to transitional B cells. Taken together, these findings indicate that the population of CD24highCD38highCD5high B cells found in PB, CB, and BM appears to be distinct from B1 B cells, traditionally distinguished by CD5 expression.

The accumulation of mutations in Ig V region genes has been used to define different B cell subsets. Immature and naive B cells express unmutated Ig V region genes, whereas those expressed by memory B cells display a high frequency of somatic hypermutation (14, 15, 39, 40, 41). We investigated the mutational status of Ig V region genes in sort-purified CD24highCD38high B cells (>98% purity) by cloning and sequencing genes belonging to the Ig VH5 gene family (14, 32). Forty percent (8 of 20) of sequences from CD24highCD38high B cells from two healthy individuals were unmutated and another 25% (5 of 20) contained only one mutation (Fig. 4,a). Six of the remaining sequences contained two or three mutations, while one sequence had five mutations, which could have been derived from contaminating memory B cells, or represent errors introduced by the polymerase used in this process. Fifty percent of mutations detected in transitional B cells were silent mutations, and 95% of sequences contained no mutations in either of the CDRs (Fig. 4,a). On average, there were 1.2 mutations/sequence, representing a mutation frequency of 0.4% (Fig. 4,a). This was similar to the mutation rate we (0.8 mutations/sequence; 0.3%) and others have observed for naive B cells in PB (15 , 19 , 61) and immature B cells in BM (40). This level of mutation may be a minor overestimate due to the nested PCR–some mutations may occur due to the endogenous error rate of the polymerase over the large number of amplification cycles used. Despite this possibility, the mutation frequency of transitional B cells was significantly less than that observed for memory B cells present in PB (6.9 mutations/sequence; mean ± SD: 2.4 ± 2.1%, Fig. 4 b; and Ref.15 : 3.8 ± 1%), tonsil (2.5 ± 2.1%; Ref.41), BM (3.0 ± 2.4%; Ref.40), and spleen (2.54 ± 1.8%; Refs.14 and 39). These data suggest that human transitional B cells express predominantly unmutated Ig V region genes, a characteristic of immature/Ag-inexperienced cells.

FIGURE 4.

CD24highCD38high B cells predominantly express unmutated Ig V region genes. Ig VH5 genes were amplified from (a) CD24highCD38high and (b) CD24+CD38 memory B cell cDNA, cloned, sequenced, and compared with known germline sequences (VH5-32 and VH5-251/73). Each line represents a single Ig VH5 gene obtained from two healthy donors (donor 1: sequences 1–10; donor 2: sequences 11–20). Vertical bars represent silent mutations; vertical bars with enclosed circles represent replacement mutations. The total number of mutations within each VH5 gene sequence are shown at the end of the sequence line. Framework regions (FR) and CDRs are indicated.

FIGURE 4.

CD24highCD38high B cells predominantly express unmutated Ig V region genes. Ig VH5 genes were amplified from (a) CD24highCD38high and (b) CD24+CD38 memory B cell cDNA, cloned, sequenced, and compared with known germline sequences (VH5-32 and VH5-251/73). Each line represents a single Ig VH5 gene obtained from two healthy donors (donor 1: sequences 1–10; donor 2: sequences 11–20). Vertical bars represent silent mutations; vertical bars with enclosed circles represent replacement mutations. The total number of mutations within each VH5 gene sequence are shown at the end of the sequence line. Framework regions (FR) and CDRs are indicated.

Close modal

The functional characteristics of CD24highCD38high B cells were investigated by analyzing their proliferative potential, Ig secretion, survival, and migration relative to mature B cells.

CD24highCD38high B cells have reduced proliferative capacity compared with naive B cells.

Proliferation of naive and CD24highCD38high B cells was examined by culturing them in the absence or presence of CD40L, anti-Ig, or both for 5 days. Although neither B cell population proliferated when unstimulated or in response to anti-Ig alone, proliferation was induced by CD40L, and was augmented by anti-Ig (Fig. 5,a). However, the response of naive B cells to CD40L, with and without anti-Ig, was significantly greater than that of CD24highCD38high B cells (p < 0.001; Fig. 5,a). To exclude the possibility that differences in proliferation reflected differences in the kinetics of the responses of the individual B cell populations, a time course of the response was performed (Fig. 5,b). Proliferation of CD24highCD38high B cells was maximal on day 4, whereas the peak response of naive B cells occurred after 5 days. Despite this difference, CD24highCD38high B cells proliferated 50–90% less than naive B cells at all times examined (Fig. 5 b).

FIGURE 5.

CD24highCD38high B cells display poorer functional capabilities than mature B cell subsets. a, Sort-purified naive (□) and transitional (▧) B cells were cultured in vitro for 5 days either in medium alone, or in the presence of CD40L, anti-Ig or CD40L plus anti-Ig. Cell proliferation was measured by determining the incorporation of [3H]thymidine into newly synthesized DNA during the final 8 h of culture. Significant differences are indicated (∗∗∗, p < 0.001). b, Sort-purified naive (▪) and transitional (□) B cells were cultured in vitro for 8 days with CD40L and anti-Ig. Proliferation was assessed at the times indicated. Significant differences are indicated (∗, p < 0.05; ∗∗, p < 0.01). These results are representative of four (a) or two (b) experiments performed using cells from different donors. c, Purified B cells (106) were loaded into the upper chamber of a transwell in duplicate and either medium alone (basal, ▦) or CXCL12 (100 ng/ml), CXCL13 (3 mg/ml), or CCL21 (600 ng/ml) (▪) were added to the lower wells. Cells were allowed to migrate for 4 h, after which time migrated cells were harvested from the lower wells and stained with mAb specific for CD19, CD24, and CD38, to enable resolution of transitional (T), naive (N), and memory (M) B cell populations. The number of migrated B cells was determined by flow cytometry and results show the mean percentage of input cells from duplicate wells that migrated toward each chemokine.

FIGURE 5.

CD24highCD38high B cells display poorer functional capabilities than mature B cell subsets. a, Sort-purified naive (□) and transitional (▧) B cells were cultured in vitro for 5 days either in medium alone, or in the presence of CD40L, anti-Ig or CD40L plus anti-Ig. Cell proliferation was measured by determining the incorporation of [3H]thymidine into newly synthesized DNA during the final 8 h of culture. Significant differences are indicated (∗∗∗, p < 0.001). b, Sort-purified naive (▪) and transitional (□) B cells were cultured in vitro for 8 days with CD40L and anti-Ig. Proliferation was assessed at the times indicated. Significant differences are indicated (∗, p < 0.05; ∗∗, p < 0.01). These results are representative of four (a) or two (b) experiments performed using cells from different donors. c, Purified B cells (106) were loaded into the upper chamber of a transwell in duplicate and either medium alone (basal, ▦) or CXCL12 (100 ng/ml), CXCL13 (3 mg/ml), or CCL21 (600 ng/ml) (▪) were added to the lower wells. Cells were allowed to migrate for 4 h, after which time migrated cells were harvested from the lower wells and stained with mAb specific for CD19, CD24, and CD38, to enable resolution of transitional (T), naive (N), and memory (M) B cell populations. The number of migrated B cells was determined by flow cytometry and results show the mean percentage of input cells from duplicate wells that migrated toward each chemokine.

Close modal

CD24highCD38high B cells produce low amounts of Ig.

Ig secretion by naive, memory, and CD24highCD38high B cells was next assessed. Culturing cells in medium or with CD40L alone failed to promote any Ig secretion by CD24highCD38high B cells and induced only low levels from naive and memory B cells (Table I). Addition of SAC induced some Ig secretion by all B cell populations, with the highest levels being observed with CD40L, IL-10, and SAC (Table I). Under these conditions, the amount of Ig secreted by CD24highCD38high B cells was 2- to 10-fold less than that produced by naive B cells and 5- to 50-fold lower than memory B cells (Table I). Thus, transitional B cells are unable to respond as efficiently as mature B cells.

Table I.

Impaired Ig secretion by activated transitional B cells in vitroa

StimulusIgM Secretion (ng/ml) (mean ± SD)IgG Secretion (ng/ml) (mean ± SD)IgA Secretion (ng/ml) (mean ± SD)
TransNaiveMemoryTransNaiveMemoryTransNaiveMemory
Nil <1 <1 300 ± 18 <1 <1 5 ± 1 <1 <1 9.0 ± 1 
CD40L 37 ± 4.0 114 ± 13 3,445 ± 62 <0.01 <0.01 170 ± 29 <0.01 5 ± 1 633 ± 180 
CD40L + SAC 864 ± 51 6,127 ± 976 9,912 ± 920 5 ± 0.1 37 ± 2 492 ± 22 2.0 ± 0.3 20 ± 5 1,116 ± 184 
CD40L + IL-10 + SAC 13,761 ± 2,367 26,389 ± 1,615 84,448 ± 17,528 428 ± 106 1,718 ± 480 2,983 ± 1,372 459 ± 123 1,749 ± 657 44,238 ± 2,629 
StimulusIgM Secretion (ng/ml) (mean ± SD)IgG Secretion (ng/ml) (mean ± SD)IgA Secretion (ng/ml) (mean ± SD)
TransNaiveMemoryTransNaiveMemoryTransNaiveMemory
Nil <1 <1 300 ± 18 <1 <1 5 ± 1 <1 <1 9.0 ± 1 
CD40L 37 ± 4.0 114 ± 13 3,445 ± 62 <0.01 <0.01 170 ± 29 <0.01 5 ± 1 633 ± 180 
CD40L + SAC 864 ± 51 6,127 ± 976 9,912 ± 920 5 ± 0.1 37 ± 2 492 ± 22 2.0 ± 0.3 20 ± 5 1,116 ± 184 
CD40L + IL-10 + SAC 13,761 ± 2,367 26,389 ± 1,615 84,448 ± 17,528 428 ± 106 1,718 ± 480 2,983 ± 1,372 459 ± 123 1,749 ± 657 44,238 ± 2,629 
a

Transitional (CD24high CD38high), naive (CD24+CD38+), and memory (CD24+CD38) B cells were sort-purified and then cultured (10 × 103 cells/200 μl) for 2 wk with the indicated stimuli. The amount of Ig secreted was then determined by Ig H chain-specific ELISAs. The values are the mean ± SD of quadruplicate cultures. Similar results were obtained in a second experiment using cells from a different donor.

B cell survival.

Reduced proliferation and Ig secretion by transitional B cells may reflect their impaired survival, as demonstrated recently (17). Indeed, when viability was assessed after 4 days of in vitro culture in the absence of any exogenous stimuli, 50% of both the naive and memory B cells were dead, and >75% of CD24highCD38high B cells were lost from the starting population. Thus, transitional B cells have reduced survival in vitro compared with mature B cells. Interestingly, despite expressing detectable levels of BAFF-R (Fig. 2), supplementing the cultures with exogenous BAFF did not alleviate the rate of death in cultures of CD24highCD38high B cells (data not shown). To determine whether other stimuli may influence the survival of transitional B cells, isolated B cell populations were cultured with CD40L and anti-Ig. Under these conditions, the survival of transitional B cells after 4 days was improved by ∼25% compared with cultures of unstimulated cells. However, the number of surviving transitional B cells in cultures stimulated with CD40L and anti-Ig was still less than that recovered from cultures of naive or memory B cells (data not shown). These results are consistent with the greater level of proliferation observed by naive B cells compared with transitional B cells when stimulated with CD40L and anti-Ig (see Fig. 5, a and b). This is also consistent with the finding by Sims et al. (17) that even though culture with IL-4 or stromal cells improved the survival of transitional B cells relative to unstimulated conditions, there were still fewer surviving transitional B cells when compared with mature B cells.

Migration of CD24highCD38high B cells is decreased.

Expression of chemokine receptors and responses to their ligands increase during B cell development; this correlates with differential expression of chemokine receptors on developing B cells (42). Examination of human PB B cells revealed that transitional B cells expressed lower levels of CXCR4 than naive B cells (MFI: 12.7 ± 1.5 vs 39.3 ± 5.0) and slightly lower levels of CXCR5 (MFI: 78.2 ± 5.0 vs 95.2 ± 10.5; Fig. 2). The reduced expression of chemokine receptors on transitional B cells appeared to have functional consequences because their migration toward the chemokines CXCL12, CXCL13, and CCL21 tended to be less than that of naive and memory B cells (Fig. 5,c). The reduced responsiveness of transitional B cells to chemokines, when coupled with reduced expression of homing molecules (CD44, CD62L; see Ref.17), may contribute to the reduced frequencies of these cells in secondary lymphoid tissues (Fig. 3).

In addition to XLP, several other conditions are characterized by a deficiency of memory B cells and hypogammaglobulinemia, including CVID (20, 22, 43) and post-HSCT (25, 44). Humoral immune responses are also reduced in neonates (45). Based on these observations, we examined different groups of immunocompromised individuals for the presence of transitional B cells.

The PB B cell compartment of a 9-mo-old child contained ∼25% CD24highCD38high B cells, while 10.3% of PB B cells from two 30-mo-old children had this phenotype (Fig. 6,a). When PB samples from CVID patients were examined, the frequency of both total B cells (mean 11.2 ± 1.4%, range 0–41%, n = 44) and B cells with a memory phenotype (i.e., CD27+; 24.0 ± 3.2%, range 0.87–76.8%) were not significantly different from those of normal controls (total B: 15.0 ± 2.2%, range 6.7–21.7%; memory B: 27.7 ± 3.0%, range 11–45%; n = 10). However, in ∼20% of patients, <5% of B cells were of a memory phenotype, as reported in previous studies (20, 21, 22). Quantitation of transitional B cells revealed a broad distribution of frequencies ranging from 0.1 to 35% (mean ± sem: 4.6 ± 0.9%, n = 44), with individual figures being either comparable to, less than, or greater than normal donors (selected examples presented in Fig. 6,b; all data points presented in Fig. 6,c). Although the mean frequency was not significantly different from that of normal controls (2.06 ± 0.27%, n = 10), there was clearly a cohort of patients (9 of 44; ∼21%) in whom the frequency of transitional B cells was increased at least 3-fold (12.5 ± 3.0%) compared with normal donors. (Fig. 6,c). Interestingly, this figure is similar to that observed for XLP patients (Fig. 1). We also assessed whether there was an inverse relationship between the frequencies of transitional and memory B cells in CVID patients. Although a few patients exhibited a “low memory/high transitional” B cell compartment (e.g., CVID no. 5: 2.25% memory/14.7% transitional; CVID no. 21: 2.48% memory/7.1% transitional; CVID no. 39: 10% memory/35.6% transitional), this was not a significant correlation for all patients.

FIGURE 6.

Transitional B cells are expanded in conditions of hypogammaglobulinemia. PBMCs from different donors or patients were labeled with anti-CD19, anti-CD24, and anti-CD38 mAb to determine the frequency of CD24highCD38high B cells (a) throughout development or (b and c) in CVID (n = 44). For b, data from several CVID patients are presented to illustrate the heterogeneity observed within this cohort of patients. For c, the frequency of transitional B cells was calculated for all CVID patients.

FIGURE 6.

Transitional B cells are expanded in conditions of hypogammaglobulinemia. PBMCs from different donors or patients were labeled with anti-CD19, anti-CD24, and anti-CD38 mAb to determine the frequency of CD24highCD38high B cells (a) throughout development or (b and c) in CVID (n = 44). For b, data from several CVID patients are presented to illustrate the heterogeneity observed within this cohort of patients. For c, the frequency of transitional B cells was calculated for all CVID patients.

Close modal

Lastly, we examined reconstitution of the B cell compartment in patients recovering from HSCT. At the earliest time points examined (2–3 mo), transitional cells comprised 10–15% of peripheral B cells–this represents a 4- to 5-fold increase compared with normal donors (Fig. 7, a and b). Thereafter, the frequency of transitional B cells declined progressively with time until it laid within levels of normal donors (2.5%; Fig. 7, a and b). In contrast, <5% of B cells were memory cells 2–3 mo post-HSCT, while at later times, and coincident with the decline in transitional cells, the memory B cell compartment expanded (Fig. 7, a and c). Despite increasing, the frequency of memory B cells remained significantly lower than that of normal controls. Unlike transitional and memory B cells, naive B cells remained constant throughout the reconstitution period, comprising ∼80% of B cells (data not shown), with the remaining ∼20% being a mix of the other B cell subsets (Fig. 7). Thus, in posttransplant patients and neonates, transitional B cells are generated early and are replaced over time by memory B cells.

FIGURE 7.

Transitional B cells appear early following hemopoietic stem cell transplant, and are gradually replaced by memory B cells. PBMCs from patients recovering from autologous (n = 2) or allogeneic (n = 2) HSCT were labeled with anti-CD19, CD24, and CD38 or CD20 and CD27 mAb to identify transitional, naive, and memory B cell populations. a, The frequency (± SEM) of transitional and memory B cells in the peripheral blood of HSCT recipients at different times posttransplant was determined. The number of samples analyzed at the different times are as follows: 2 and 16 mo, n = 1; 3, 6, 9, and 12 mo, n = 4. The right panel indicates the frequencies of these cells in normal donors (n = 4) analyzed concomitantly. b and c, Representative plots of the appearance of (b) transitional (CD24highCD38high) and (c) memory (CD20+CD27+) B cells in the peripheral blood of one patient at the indicated times post-HSCT. Corresponding plots from a normal donor are shown for comparative purposes. The values indicate the percentage of B cells with a transitional (b) or memory (c) phenotype in this experiment.

FIGURE 7.

Transitional B cells appear early following hemopoietic stem cell transplant, and are gradually replaced by memory B cells. PBMCs from patients recovering from autologous (n = 2) or allogeneic (n = 2) HSCT were labeled with anti-CD19, CD24, and CD38 or CD20 and CD27 mAb to identify transitional, naive, and memory B cell populations. a, The frequency (± SEM) of transitional and memory B cells in the peripheral blood of HSCT recipients at different times posttransplant was determined. The number of samples analyzed at the different times are as follows: 2 and 16 mo, n = 1; 3, 6, 9, and 12 mo, n = 4. The right panel indicates the frequencies of these cells in normal donors (n = 4) analyzed concomitantly. b and c, Representative plots of the appearance of (b) transitional (CD24highCD38high) and (c) memory (CD20+CD27+) B cells in the peripheral blood of one patient at the indicated times post-HSCT. Corresponding plots from a normal donor are shown for comparative purposes. The values indicate the percentage of B cells with a transitional (b) or memory (c) phenotype in this experiment.

Close modal

Based on our findings of a predominance of transitional B cells in various immunodeficient states, we predicted that there would be a net reduction in Ig production by CD27 B cells from such immunodeficient individuals compared with normal donors. Thus, CD27 B cells were sort-purified from a normal donor and an XLP patient, who had 2- to 3-fold more transitional B cells than normal, and cultured in vitro. CD27 B cells from an XLP patient, which were enriched for transitional B cells, secreted 20-fold less IgM than CD27 B cells from a normal donor, which were predominantly naive, in response to CD40L and IL-10, and failed to produce switched Ig isotypes (Table II). Thus, the presence of an increased proportion of transitional B cells in XLP patients represents a manifestation of the hypogammaglobulinemic state of these patients.

Table II.

Differential production of Ig by activated CD27 B cells from XLP patients and normal donorsa

StimulusIgM (ng/ml)IgGIgA
NormalXLPNormalXLPNormalXLP
Unstimulated <1 <1 <1 <1 <1 <1 
CD40L 11.6 ± 7.1 12.4 ± 5.7 <1 <1 <1 <1 
CD40L, IL-10 1,030 ± 571 42.1 ± 41 32.6 ± 6.1 <1 40.4 ± 24.0 <1 
StimulusIgM (ng/ml)IgGIgA
NormalXLPNormalXLPNormalXLP
Unstimulated <1 <1 <1 <1 <1 <1 
CD40L 11.6 ± 7.1 12.4 ± 5.7 <1 <1 <1 <1 
CD40L, IL-10 1,030 ± 571 42.1 ± 41 32.6 ± 6.1 <1 40.4 ± 24.0 <1 
a

CD27 B cells were sort purified from a normal donor and then cultured (10 × 103 cells/200 μl) for 2 wk with the indicated stimuli. The amount of Ig secreted was then determined by Ig H chain-specific ELISAs. The values are the mean ± SD of triplicate or quadruplicate cultures.

Mature human B cells can be divided into subsets corresponding to naive, memory, or PC. This has been achieved by examining their phenotype, function, and anatomical localization (13, 14, 15, 31, 32, 39, 41, 46). Although these same B cell populations can be resolved in murine lymphoid tissues, albeit using different phenotypic criteria, transitional B cells remain incompletely characterized in humans (16, 17, 34). By investigating CD27 B cells in XLP patients, our study revealed a unique subset with a CD10+CD24highCD38high phenotype that accounted for a significant proportion of circulating B cells in these patients. B cells with this phenotype were also detected in normal donors, however, at much reduced frequencies compared with XLP patients. It was recently suggested that CD24highCD38high B cells correspond to human transitional B cells (16, 17). The current study confirms many of the phenotypic and functional features of transitional B cells described recently by Lipsky’s group (17), such as their frequency in PB, and reduced proliferation and survival in vitro (Fig. 5). However, we have substantially extended these findings by analyzing CD24highCD38high B cells in different lymphoid tissues and immunodeficient states, as well as by establishing the expression and function of chemokine receptors, survival molecules, and their functional competency with respect to Ig secretion. Importantly, for the first time we have revealed an aberration in the generation and/or maturation of these cells in a genetically defined immunodeficiency, namely XLP, and demonstrated their BM origin by examining recipients of HSCT.

By comparing the phenotype and function of B cells with a transitional phenotype to those of naive and memory cells from normal donors, it was clear that CD24highCD38high B cells were distinct from mature B cells. They were CD10+ and expressed higher levels of surface IgM, CD20, CD5, and CD9, and lower levels of CD44, CXCR4, and Bcl-2 than mature B cells. Furthermore, CD24highCD38high B cells exhibited less proliferation, differentiation, and chemotaxis in vitro than mature B cells. The impaired responses of CD24highCD38high B cells may reflect their reduced survival, which is probably attributable to a lack of expression of bcl-2, and possibly other survival molecules. This is consistent with previous studies that reported murine transitional cells do not undergo significant proliferation in vitro (3, 8) and display poorer survival than mature B cells (47), perhaps due to reduced expression of the antiapoptotic molecules Bcl-xL and A1 (4). Murine transitional B cells also migrate less than mature B cells, and this correlated with their lower expression of chemokine receptors compared with mature B cells (42). These functional similarities between murine transitional and human CD24highCD38high B cells support the proposal that the latter are indeed human transitional B cells. Because expression of Bcl-2, IgD, and CXCR4 is developmentally regulated (2, 12, 36, 42), it is likely that within the sequence of human B cell development IgD+Bcl-2CXCR4+ transitional B cells lie between IgDBcl-2CXCR4low immature B cells and IgD+Bcl-2+ CXCR4high mature B cells.

Murine transitional B cells can be divided into three distinct subsets. In contrast, the overall characteristics of human CD24highCD38high B cells incorporate features of all subsets of murine transitional B cells. For instance, the detection of CD24highCD38high B cells in PB, their lack of response to BAFF (4, 17), reduced response to BcR stimulation (3, 47), and reduced expression of prosurvival molecules (4, 8, 9) are features of mouse T1 B cells. Conversely, human transitional B cells were found to express CD21, CD23, and IgD, and in this respect resemble murine T2/T3 B cells which reside in the spleen (8, 9). Thus, maturation of human B cells may comprise only a single transitional stage. This is consistent with the finding of the uniform expression pattern of many of the cell surface molecules examined in this study (see Fig. 2). This is most notable for the absence of prosurvival molecules, such as bcl-2, in human transitional B cells, that are induced in murine B cells at the T2 stage (4). In contrast, the finding of a single population of human transitional B cells by phenotype does not eliminate the possibility of heterogeneity within this population. In other words, if subsets of human transitional B cells exist, they may be within the CD24highCD38high phenotype. Indeed, Sims et al. (17) separated the CD24highCD38high B cell population into different subsets on the basis of expression of CD38 and IgD and the minimal gradation of other markers such as CD24, and accordingly described type 1 and type 2 human transitional B cells. Similarly, our finding of broad expression of CD44 may be another means of dividing human transitional B cells into distinct subsets. Further analysis of human transitional B cells will require identification of molecules differentially expressed by these cells that may allow the delineation of transitional B cells into phenotypically resolvable subpopulations.

Although there were clear differences in phenotype and function of CD24highCD38high B cells compared with mature B cells, the elevated expression of CD5 on the former population raised the possibility that these cells could either belong to the B1 lineage or represent activated B cells. It is unlikely that CD24highCD38high cells are B1 cells (defined by expression of CD5). First, the frequency of B cells in adult PB and tonsils that are CD5+ is ∼30% and ∼10%, respectively, yet they are very infrequent in the BM (38, 48, 49). In contrast, CD24highCD38high B cells comprised only 2.5% of total PB B cells and were virtually absent from lymphoid tissues, yet abundant in BM (Fig. 3). Second, CD24highCD38high B cells lacked Bcl-2 expression (Fig. 2), while tonsillar B1 (CD5+) B cells are Bcl-2+ (49). It is also unlikely that CD24highCD38high B cells express CD5 due to activation in vivo because they are small cells, and do not express the activation markers CD25, CD69, CD80, CD86, and CD95. Thus, CD24highCD38high B cells appear to represent a unique population of human B cells, with morphological, phenotypic, and functional characteristics that distinguish them from mature B cells and would be consistent with their designation as transitional B cells. Interestingly, expression of RAG-1 and RAG-2 by circulating human B cells was recently shown to be associated with CD5 expression (50). Thus, it is possible that the CD5+ B cells examined (50) were predominantly transitional B cells that continue to express RAG proteins following their export from the BM (51), akin to murine transitional B cells (8).

A population of B cells in human tonsil has been described as pre-GC or GC founder cells (19, 52, 53). Several studies have suggested that these B cells can also be detected in the PB of normal individuals, as well as patients with SLE (19, 53). Interestingly, the frequency of such cells in normal individuals is similar to that of transitional B cells; ∼2–3% (53). GC founder cells were defined as IgD+CD38high; in tonsils, these are large cells that express CD10, CD27, CD77, and CD95, and ∼50% of them are IgM+ (2, 52). Remarkably, the proposed GC founder B cells in PB are smaller than tonsil GC B cells, and are CD27CD77CD95 (53). These morphological and phenotypic features of PB “GC founder cells” are dramatically different from those in tonsils (19, 52, 53). Furthermore, the frequency of mutation of Ig V region genes expressed by PB GC founder B cells (0%; Ref.53) from normal individuals was substantially less than that of corresponding cells in tonsil (∼1.3%; (52)). Taken together, it appears that the IgM+IgD+CD10+CD27CD38high B cells previously purported to be circulating GC founder B cells are more likely to be transitional B cells. Our study highlights both the requirement for a more extensive investigation of a multitude of surface molecules before reporting a distinct cell population, as well as the limitations of assuming surface phenotype as a definitive characterization of a subset of cells.

An important result of the current paper was the finding that transitional B cells are increased in XLP. By phenotyping the CD24highCD38high B cells in XLP patients, we demonstrated that these cells indeed constituted a population of transitional B cells. Although a recent study reported increased frequencies in SLE, the absolute number of transitional cells was normal because SLE patients are lymphopenic (17). Thus, XLP is the first human disease where transitional B cells are overrepresented in the B cell compartment. We also found an expansion of transitional B cells in neonates, some CVID patients, and patients recovering from HSCT, conditions that are characterized by hypogammaglobulinemia and an impaired ability to mount efficient humoral immune responses (43, 44, 45). Of particular note, the B cell compartment of these individuals resembled that of XLP patients, with not only an increase in transitional B cells but a decrease in memory B cells as well. Interestingly, studies from the 1980s reported that CD5+ B cells were detected at a greater frequency than conventional CD5 B cells in patients post-HSCT (54, 55). The CD5+ B cells detected in these patients expressed higher levels of IgM and CD20, but similar levels of CD19 and IgD, to CD5 B cells (54). Based on our finding that transitional B cells are CD5+IgMhighCD20high, it is highly likely that these earlier studies (54, 55) actually identified transitional B cells, rather than B1 cells, as was reported at the time. Interestingly, several studies have reported that patients infected with HIV have a significantly decreased frequency of memory B cells (56) as well as an increased frequency of CD10+ B cells in their PB (57). Furthermore, CD27 B cells in HIV patients express higher levels of CD38, lower levels of bcl-2, and are more prone to apoptosis in vitro compared with CD27 B cells from normal donors (58). Thus, HIV infection represents another immune-deficient state associated with a decreased proportion of memory B cells and an increased proportion of circulating “immature” B cells (56, 57, 58) that are most likely transitional B cells. These findings raise the question of the contribution of transitional B cells to the hypogammaglobulinemia characteristic of neonates and these different groups of patients. A substantial increase in the number of transitional B cells, coupled with a deficit in memory B cells, could certainly contribute to their immunodeficient state because transitional B cells produced less Ig than mature B cells. Indeed, this was demonstrated experimentally, as CD27 XLP B cells produced substantially less IgM, and failed to secrete detectable levels of isotype-switched Ig, in vitro compared with CD27 B cells from normal donors, where the frequency of transitional B cells is significantly less. It is presently unclear why transitional B cells are increased in XLP. It was interesting to observe that during reconstitution of the B cell compartment in HSCT patients, as well as in normal children, memory B cells appeared to replace the transitional B cells over time, while the naive population remained static (Fig. 7). This raises the possibility that in XLP, transitional B cells “fill the space” in the peripheral B cell compartment due to the absence of memory B cells as a consequence of compensatory mechanisms of the primary immunodeficiency (21).

Our findings may also have practical benefit. As well as detecting an increase in transitional B cells in XLP, we previously noted an absence of memory B cells and NKT cells in this disease (21, 26). Thus, enumerating these cell types may facilitate improved diagnosis of XLP. Similarly, monitoring the frequency of transitional, as well as memory, B cells may provide a means of assessing the immunocompetence of CVID or HIV-infected patients or individuals post-HSCT. Interestingly, transitional B cells appear to represent a checkpoint where autoreactive B cells are removed from the peripheral population (34). In other words, aberrations at this stage of B cell development may contribute to the appearance of circulating autoreactive B cells (59, 60), which may be one explanation for the increase in the frequency of transitional B cells in SLE (17). Overall, this study has characterized a B cell subset that corresponds to a transitional cell occupying an intermediate stage in differentiation between immature and mature B cells. Further investigation of these cells will improve our understanding of the molecular, cellular, and biological processes underlying human B cell development, and how alterations to these processes may precipitate immunodeficiency or autoimmunity.

We thank Stephen Adelstein, Don Anderson, Frank Alvaro, Barbara Fazekas, John Gibson, David Fulcher, Joy Ho, Amy Klion, Monique Parkin, Sean Riminton, Ron Walls, Andrew Williams, Melaine Wong, and the Australian Red Cross Blood Service for providing patient and tissue samples; Rene de Waal Malefyt and Susan Kalled for reagents; Dr. Grant Shoebridge and Nathan Hare for preparing the human CD40L; Dr. Adrian Smith and Vivienne Moore for cell sorting; and Prof. Tony Basten and Drs. Pam Schwartzberg and Tri Phan for critical review of this manuscript.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by the National Health and Medical Research Council (NHMRC) of Australia. S.G.T. is the recipient of an R. D. Wright Biomedical Career Development Award from the NHMRC.

3

Abbreviations used in this paper: BM, bone marrow; PB, peripheral blood; GC, germinal center; PC, plasma cell; SLE, systemic lupus erythematosus; CVID, common variable immunodeficiency; HSCT, hemopoietic stem cell transplant; XLP, X-linked lymphoproliferative disease; CB, cord blood; BAFF, B cell activating factor of the TNF family; BAFF-R, BAFF receptor; TACI, transmembrane activator and calcium modulator and cyclophilin ligand interactor; SAC, Staphylococcus aureus Cowan; MFI, mean fluorescence intensity; LN, lymph node; MNC, mononuclear cell.

1
Rolink, A., F. Melchers.
1993
. B lymphopoiesis in the mouse.
Adv. Immunol.
53
:
123
-156.
2
Banchereau, J., F. Rousset.
1992
. Human B lymphocytes: phenotype, proliferation, and differentiation.
Adv. Immunol.
52
:
125
-262.
3
Allman, D., R. C. Lindsley, W. DeMuth, K. Rudd, S. A. Shinton, R. R. Hardy.
2001
. Resolution of three nonproliferative immature splenic B cell subsets reveals multiple selection points during peripheral B cell maturation.
J. Immunol.
167
:
6834
-6840.
4
Allman, D., B. Srivastava, R. C. Lindsley.
2004
. Alternative routes to maturity: branch points and pathways for generating follicular and marginal zone B cells.
Immunol. Rev.
197
:
147
-160.
5
Allman, D. M., S. E. Ferguson, M. P. Cancro.
1992
. Peripheral B cell maturation. I. Immature peripheral B cells in adults are heat-stable antigenhigh and exhibit unique signaling characteristics.
J. Immunol.
149
:
2533
-2540.
6
Allman, D. M., S. E. Ferguson, V. M. Lentz, M. P. Cancro.
1993
. Peripheral B cell maturation. II. Heat-stable antigenhigh splenic B cells are an immature developmental intermediate in the production of long-lived marrow-derived B cells.
J. Immunol.
151
:
4431
-4444.
7
Carsetti, R., G. Kohler, M. C. Lamers.
1995
. Transitional B cells are the target of negative selection in the B cell compartment.
J. Exp. Med.
181
:
2129
-2140.
8
Chung, J. B., M. Silverman, J. G. Monroe.
2003
. Transitional B cells: step by step towards immune competence.
Trends Immunol.
24
:
343
-349.
9
Loder, F., B. Mutschler, R. J. Ray, C. J. Paige, P. Sideras, R. Torres, M. C. Lamers, R. Carsetti.
1999
. B cell development in the spleen takes place in discrete steps and is determined by the quality of B cell receptor-derived signals.
J. Exp. Med.
190
:
75
-89.
10
Rolink, A. G., J. Andersson, F. Melchers.
1998
. Characterization of immature B cells by a novel monoclonal antibody, by turnover and by mitogen reactivity.
Eur. J. Immunol.
28
:
3738
-3748.
11
Oliver, A. M., F. Martin, G. L. Gartland, R. H. Carter, J. F. Kearney.
1997
. Marginal zone B cells exhibit unique activation, proliferative and immunoglobulin secretory responses.
Eur. J. Immunol.
27
:
2366
-2374.
12
Uckun, F. M..
1990
. Regulation of human B-cell ontogeny.
Blood
76
:
1908
-1923.
13
Liu, Y. J., C. Barthelemy, O. de Bouteiller, C. Arpin, I. Durand, J. Banchereau.
1995
. Memory B cells from human tonsils colonize mucosal epithelium and directly present antigen to T cells by rapid up-regulation of B7-1 and B7-2.
Immunity
2
:
239
-248.
14
Tangye, S. G., Y. J. Liu, G. Aversa, J. H. Phillips, J. E. de Vries.
1998
. Identification of functional human splenic memory B cells by expression of CD148 and CD27.
J. Exp. Med.
188
:
1691
-1703.
15
Klein, U., K. Rajewsky, R. Kuppers.
1998
. Human immunoglobulin (Ig)M+IgD+ peripheral blood B cells expressing the CD27 cell surface antigen carry somatically mutated variable region genes: CD27 as a general marker for somatically mutated (memory) B cells.
J. Exp. Med.
188
:
1679
-1689.
16
Carsetti, R., M. M. Rosado, H. Wardmann.
2004
. Peripheral development of B cells in mouse and man.
Immunol. Rev.
197
:
179
-191.
17
Sims, G. P., R. Ettinger, Y. Shirota, C. H. Yarboro, G. G. Illei, P. E. Lipsky.
2005
. Identification and characterization of circulating human transitional B cells.
Blood
105
:
4390
-4398.
18
Odendahl, M., A. Jacobi, A. Hansen, E. Feist, F. Hiepe, G. R. Burmester, P. E. Lipsky, A. Radbruch, T. Dorner.
2000
. Disturbed peripheral B lymphocyte homeostasis in systemic lupus erythematosus.
J. Immunol.
165
:
5970
-5979.
19
Arce, E., D. G. Jackson, M. A. Gill, L. B. Bennett, J. Banchereau, V. Pascual.
2001
. Increased frequency of pre-germinal center B cells and plasma cell precursors in the blood of children with systemic lupus erythematosus.
J. Immunol.
167
:
2361
-2369.
20
Warnatz, K., A. Denz, R. Drager, M. Braun, C. Groth, G. Wolff-Vorbeck, H. Eibel, M. Schlesier, H. H. Peter.
2002
. Severe deficiency of switched memory B cells (CD27+IgMIgD) in subgroups of patients with common variable immunodeficiency: a new approach to classify a heterogeneous disease.
Blood
99
:
1544
-1551.
21
Ma, C. S., N. J. Hare, K. E. Nichols, L. Dupre, G. Andolfi, M. G. Roncarolo, S. Adelstein, P. D. Hodgkin, S. G. Tangye.
2005
. Impaired humoral immunity in X-linked lymphoproliferative disease is associated with defective IL-10 production by CD4+ T cells.
J. Clin. Invest.
115
:
1049
-1059.
22
Brouet, J. C., A. Chedeville, J. P. Fermand, B. Royer.
2000
. Study of the B cell memory compartment in common variable immunodeficiency.
Eur. J. Immunol.
30
:
2516
-2520.
23
Agematsu, K., H. Nagumo, K. Shinozaki, S. Hokibara, K. Yasui, K. Terada, N. Kawamura, T. Toba, S. Nonoyama, H. D. Ochs, A. Komiyama.
1998
. Absence of IgD-CD27+ memory B cell population in X-linked hyper-IgM syndrome.
J. Clin. Invest.
102
:
853
-860.
24
Weller, S., A. Faili, C. Garcia, M. C. Braun, F. F. Le Deist, G. G. de Saint Basile, O. Hermine, A. Fischer, C. A. Reynaud, J. C. Weill.
2001
. CD40-CD40L independent Ig gene hypermutation suggests a second B cell diversification pathway in humans.
Proc. Natl. Acad. Sci. USA
98
:
1166
-1170.
25
Avanzini, M. A., F. Locatelli, C. Dos Santos, R. Maccario, E. Lenta, M. Oliveri, S. Giebel, P. De Stefano, F. Rossi, G. Giorgiani, et al
2005
. B lymphocyte reconstitution after hematopoietic stem cell transplantation: functional immaturity and slow recovery of memory CD27+ B cells.
Exp. Hematol.
33
:
480
-486.
26
Nichols, K. E., C. S. Ma, J. L. Cannons, P. L. Schwartzberg, S. G. Tangye.
2005
. Molecular and cellular pathogenesis of X-linked lymphoproliferative disease.
Immunol. Rev.
203
:
180
-199.
27
Nichols, K. E., D. P. Harkin, S. Levitz, M. Krainer, K. A. Kolquist, C. Genovese, A. Bernard, M. Ferguson, L. Zuo, E. Snyder, et al
1998
. Inactivating mutations in an SH2 domain-encoding gene in X-linked lymphoproliferative syndrome.
Proc. Natl. Acad. Sci. USA
95
:
13765
-13770.
28
Coffey, A. J., R. A. Brooksbank, O. Brandau, T. Oohashi, G. R. Howell, J. M. Bye, A. P. Cahn, J. Durham, P. Heath, P. Wray, et al
1998
. Host response to EBV infection in X-linked lymphoproliferative disease results from mutations in an SH2-domain encoding gene.
Nat. Genet.
20
:
129
-135.
29
Sayos, J., C. Wu, M. Morra, N. Wang, X. Zhang, D. Allen, S. van Schaik, L. Notarangelo, R. Geha, M. G. Roncarolo, et al
1998
. The X-linked lymphoproliferative-disease gene product SAP regulates signals induced through the co-receptor SLAM.
Nature
395
:
462
-469.
30
Avery, D. T., S. L. Kalled, J. I. Ellyard, C. Ambrose, S. A. Bixler, M. Thien, R. Brink, F. Mackay, P. D. Hodgkin, S. G. Tangye.
2003
. BAFF selectively enhances the survival of plasmablasts generated from human memory B cells.
J. Clin. Invest.
112
:
286
-297.
31
Tangye, S. G., D. T. Avery, P. D. Hodgkin.
2003
. A division-linked mechanism for the rapid generation of Ig-secreting cells from human memory B cells.
J. Immunol.
170
:
261
-269.
32
Ellyard, J. I., D. T. Avery, T. G. Phan, N. J. Hare, P. D. Hodgkin, S. G. Tangye.
2004
. Antigen-selected, immunoglobulin-secreting cells persist in human spleen and bone marrow.
Blood
103
:
3805
-3812.
33
Ellyard, J. I., D. T. Avery, C. R. Mackay, S. G. Tangye.
2005
. Contribution of stromal cells to the migration, function and retention of plasma cells in human spleen: potential roles of CXCL12, IL-6 and CD54.
Eur. J. Immunol.
35
:
699
-708.
34
Wardemann, H., S. Yurasov, A. Schaefer, J. W. Young, E. Meffre, M. C. Nussenzweig.
2003
. Predominant autoantibody production by early human B cell precursors.
Science
301
:
1374
-1377.
35
Bovia, F., A. C. Nabili-Tehrani, C. Werner-Favre, M. Barnet, V. Kindler, R. H. Zubler.
1998
. Quiescent memory B cells in human peripheral blood co-express bcl-2 and bcl-xL anti-apoptotic proteins at high levels.
Eur. J. Immunol.
28
:
4418
-4423.
36
Merino, R., L. Ding, D. J. Veis, S. J. Korsmeyer, G. Nunez.
1994
. Developmental regulation of the Bcl-2 protein and susceptibility to cell death in B lymphocytes.
EMBO J.
13
:
683
-691.
37
Berland, R., H. H. Wortis.
2002
. Origins and functions of B-1 cells with notes on the role of CD5.
Annu. Rev. Immunol.
20
:
253
-300.
38
Bhat, N. M., A. B. Kantor, M. M. Bieber, A. M. Stall, L. A. Herzenberg, N. N. Teng.
1992
. The ontogeny and functional characteristics of human B-1 (CD5+ B) cells.
Int. Immunol.
4
:
243
-252.
39
Dunn-Walters, D. K., P. G. Isaacson, J. Spencer.
1995
. Analysis of mutations in immunoglobulin heavy chain variable region genes of microdissected marginal zone (MGZ) B cells suggests that the MGZ of human spleen is a reservoir of memory B cells.
J. Exp. Med.
182
:
559
-566.
40
Paramithiotis, E., M. D. Cooper.
1997
. Memory B lymphocytes migrate to bone marrow in humans.
Proc. Natl. Acad. Sci. USA
94
:
208
-212.
41
Pascual, V., Y. J. Liu, A. Magalski, O. de Bouteiller, J. Banchereau, J. D. Capra.
1994
. Analysis of somatic mutation in five B cell subsets of human tonsil.
J. Exp. Med.
180
:
329
-339.
42
Bowman, E. P., J. J. Campbell, D. Soler, Z. Dong, N. Manlongat, D. Picarella, R. R. Hardy, E. C. Butcher.
2000
. Developmental switches in chemokine response profiles during B cell differentiation and maturation.
J. Exp. Med.
191
:
1303
-1318.
43
Di Renzo, M., A. L. Pasqui, A. Auteri.
2004
. Common variable immunodeficiency: a review.
Clin. Exp. Med.
3
:
211
-217.
44
Parkman, R., K. I. Weinberg.
1997
. Immunological reconstitution following bone marrow transplantation.
Immunol. Rev.
157
:
73
-78.
45
Gathings, W. E., H. Kubagawa, M. D. Cooper.
1981
. A distinctive pattern of B cell immaturity in perinatal humans.
Immunol. Rev.
57
:
107
-126.
46
Arpin, C., J. Banchereau, Y. J. Liu.
1997
. Memory B cells are biased towards terminal differentiation: a strategy that may prevent repertoire freezing.
J. Exp. Med.
186
:
931
-940.
47
Su, T. T., D. J. Rawlings.
2002
. Transitional B lymphocyte subsets operate as distinct checkpoints in murine splenic B cell development.
J. Immunol.
168
:
2101
-2110.
48
Kasaian, M. T., H. Ikematsu, P. Casali.
1992
. Identification and analysis of a novel human surface CD5 B lymphocyte subset producing natural antibodies.
J. Immunol.
148
:
2690
-2702.
49
Abe, M., K. Tominaga, H. Wakasa.
1994
. Phenotypic characterization of human B-lymphocyte subpopulations, particularly human CD5+ B-lymphocyte subpopulation within the mantle zones of secondary follicles.
Leukemia
8
:
1039
-1044.
50
Hillion, S., A. Saraux, P. Youinou, C. Jamin.
2005
. Expression of RAGs in peripheral B cells outside germinal centers is associated with the expression of CD5.
J. Immunol.
174
:
5553
-5561.
51
Yu, W., H. Nagaoka, M. Jankovic, Z. Misulovin, H. Suh, A. Rolink, F. Melchers, E. Meffre, M. C. Nussenzweig.
1999
. Continued RAG expression in late stages of B cell development and no apparent re-induction after immunization.
Nature
400
:
682
-687.
52
Lebecque, S., O. de Bouteiller, C. Arpin, J. Banchereau, Y. J. Liu.
1997
. Germinal center founder cells display propensity for apoptosis before onset of somatic mutation.
J. Exp. Med.
185
:
563
-571.
53
Bohnhorst, J. O., M. B. Bjorgan, J. E. Thoen, J. B. Natvig, K. M. Thompson.
2001
. Bm1-Bm5 classification of peripheral blood B cells reveals circulating germinal center founder cells in healthy individuals and disturbance in the B cell subpopulations in patients with primary Sjogren’s syndrome.
J. Immunol.
167
:
3610
-3618.
54
Ault, K. A., J. H. Antin, D. Ginsburg, S. H. Orkin, J. M. Rappeport, M. L. Keohan, P. Martin, B. R. Smith.
1985
. Phenotype of recovering lymphoid cell populations after marrow transplantation.
J. Exp. Med.
161
:
1483
-1502.
55
Antin, J. H., K. A. Ault, J. M. Rappeport, B. R. Smith.
1987
. B lymphocyte reconstitution after human bone marrow transplantation: Leu-1 antigen defines a distinct population of B lymphocytes.
J. Clin. Invest.
80
:
325
-332.
56
De Milito, A., C. Morch, A. Sonnerborg, F. Chiodi.
2001
. Loss of memory (CD27) B lymphocytes in HIV-1 infection.
AIDS
15
:
957
-964.
57
Martinez-Maza, O., E. Crabb, R. T. Mitsuyasu, J. L. Fahey, J. V. Giorgi.
1987
. Infection with the human immunodeficiency virus (HIV) is associated with an in vivo increase in B lymphocyte activation and immaturity.
J. Immunol.
138
:
3720
-3724.
58
Chong, Y., H. Ikematsu, M. Yamamoto, M. Murata, K. Yamaji, M. Nishimura, S. Nabeshima, S. Kashiwagi, J. Hayashi.
2004
. Increased frequency of CD27 (naive) B cells and their phenotypic alteration in HIV type 1-infected patients.
AIDS Res. Hum. Retroviruses
20
:
621
-629.
59
Yurasov, S., H. Wardemann, J. Hammersen, M. Tsuiji, E. Meffre, V. Pascual, M. C. Nussenzweig.
2005
. Defective B cell tolerance checkpoints in systemic lupus erythematosus.
J. Exp. Med.
201
:
703
-711.
60
Samuels, J., Y. S. Ng, C. Coupillaud, D. Paget, E. Meffre.
2005
. Impaired early B cell tolerance in patients with rheumatoid arthritis.
J. Exp. Med.
201
:
1659
-1667.
61
Ma, C. S., S. Pittaluga, D. T. Avery, N. J. Hare, A. D. Klion, I. Maric, K. E. Nichols, and S. G. Tangye. Selective generation of functional somatically mutated IgM+CD27+, but not Ig isotype-switched, memory B cells in X-linked lymphoproliferative disease. J. Clin. Invest. In press.