B Cell Receptor-Mediated Calcium Signaling Is Impaired in B Lymphocytes of Type Ia Patients with Common Variable Immunodeficiency Free

Common variable immunodeficiency (CVID) is the most frequent symptomatic primary immunodeficiency in human adults comprising diverse forms of primary Ab deficiencies (1). Less than 10% of CVID cases are characterized by a monogenetic defect (2, 3). In the last few years, the analysis of the peripheral B cell phenotype has allowed classifying this heterogeneous disease into distinct groups (4–6), but it does not distinguish between B cell intrinsic and other forms of defects in CVID. Altered BCR-induced upregulation of activation markers like CD86 and CD70 (7, 8), proliferation, and Ig synthesis in vitro (9) strongly suggest B cell-intrinsic defects underlying the immunodeficiency in several patients, rendering the identification of these defects a major pathogenically relevant goal.

One important pathway in the activation process of B cells is the mobilization of calcium (10). After crosslinking of the Ig receptors by Ag, phospholipase Cγ2 (PLCγ2) is phosphorylated by Syk and Btk and induces the generation of inositol trisphosphate (IP3) (11). IP3 mediates a transient calcium release from intracellular stores, which leads to a sustained influx of calcium through Ca²⁺ channels in the plasma membrane, a process termed store-operated calcium entry (12–14). PLCγ2-deficient mice have a partial block in B cell development at the transitional B cell stage and show reduced Ab response (15–17). Interestingly, targeted deletion of PLCγ2 in germinal center B cells demonstrated an additional essential role in the maintenance of memory B cells (18), a function severely impaired in CVID (4). In humans, impaired calcium signaling in B cells has been demonstrated in immunodeficiencies like CD19 deficiency, X-linked agammaglobulinemia, and Wiskott-Aldrich syndrome (WAS), all of which are characterized by hypogammaglobulinemia (19–21).

In this study, we describe for the first time impaired calcium signaling in B cells from a subgroup of patients with CVID corresponding to the CVID class Ia of the Freiburg classification (4). This patient population is defined by a severe reduction in IgD⁻IgM⁻CD27⁺ class-switched memory B cells and the expansion of CD21^low B cells (CD19^hiCD21⁻CD27⁻CD38^low). Clinically, the patients are prone to develop secondary complications, such as splenomegaly, granuloma, and autoimmune cytopenia (22). In this study, we show that Ca²⁺ responses to BCR stimulation are reduced in mature B cells of patients with class Ia CVID, whereas transitional B cells exhibit normal Ca²⁺ signals. In the CD21^low B cell population of patients with CVID Ia but also healthy controls, additional mechanisms reduce the Ca²⁺ signal even further. Whereas BCR proximal signaling and release of Ca²⁺ from endoplasmic reticulum (ER) stores are normal in B cells from patients with CVID Ia, Ca²⁺ influx across the plasma membrane is reduced, a defect that is accompanied by an elevated expression of CD22, a negative regulator of signaling in B cells. Impaired calcium influx in B cells from patients with CVID Ia is associated with autoimmune cytopenia and lymphadenopathy in these patients.

Patients with CVID diagnosed in line with the European Society for Immunodeficiencies/Pan-American Group for Immunodeficiency criteria were grouped according to the Freiburg classification, and 53 patients including 21 Freiburg type Ia patients and 39 healthy donors (HDs) were screened. Informed written consent to the internal ethics review board-approved clinical study protocol (University Hospital Freiburg 239/99) was obtained from each individual before participation in the study, in accordance with the Declaration of Helsinki. Splenomegaly was defined by ultrasound with a spleen size exceeding 4.7 cm in diameter or 11 cm in length.

For flow cytometric analyses, the following Abs were used: CD4-APC, CD20-PerCP-Cy5.5, CD27-FITC, CD38-APC, CD86-PE, phosphorylated phospholipase Cγ2 (pPLCγ2)-Alexa 647 (BD Biosciences, Heidelberg, Germany), CD8-PE, CD19-PE-Cy7, CD21-FITC, CD22-PE CD38-PE, CD45R0-FITC (Beckman Coulter, Krefeld, Germany), CD27-FITC, CD27-PE, Ki67-FITC (DakoCytomation, Hamburg, Germany), IgD-PE (Southern Biotechnology Associates, Birmingham, AL), IgM-PerCP-Cy5.5 (Dianova, Hamburg, Germany). The analysis was performed on an LSR2 (BD Biosciences) for calcium flux assays and on an FACSCanto II (BD Biosciences) for all other assays. Data were analyzed using FlowJo (Tree Star, Ashland, OR) software.

Ficoll-isolated PBMCs were loaded with 4.5 μM Indo-1 AM calcium dye (Sigma-Aldrich, Steinheim, Germany) and 0.045% Pluronic F-127 (Invitrogen, Carlsbad, CA) in RPMI 1640 supplemented with 16.8% FCS and incubated for 45 min at room temperature. Loaded cells were washed twice and subsequently stained with labeling Abs for 15 min at room temperature. B cell subpopulations were distinguished separating naive B cells and CD21^low B cells with CD19-PE-Cy7, CD21-FITC, CD27-PE, CD38-APC, naive B cells, IgM⁺CD27⁺ B cells, switched memory B cells, transitional B cells, and plasmablasts with CD19-PE-Cy7, CD27-FITC, CD38-APC, and IgD-PE (Supplemental Fig. 1). Cells were washed and resuspended at 4 × 10⁶ cells/ml in 10% RPMI. A baseline was read over the first minute, after which B cells were stimulated with 15 μg/ml F(ab′)₂ anti-IgM (Southern Biotechnology Associates), and the resulting calcium release was recorded for 8 min. For the assays distinguishing store depletion and Ca²⁺ entry from the extracellular space, B cells were stimulated with 20 μg/ml F(ab′)₂ anti-IgM in Ca²⁺-free Ringer’s solution supplemented with 0.5 mM EGTA, and after 4 min, CaCl₂ was added to a final concentration of 4 mM and recorded for 3 min. A total of 2 μg/ml ionomycin (Sigma-Aldrich) served to elicit the maximum response over the last minute in all assays. The relative concentration of intracellular free Ca²⁺ was measured as the median fluorescence ratio of Indo-1 bound/unbound (405/485 nm) over time.

For statistical comparison, the calcium response was normalized and expressed as the ratio of the maximum peak poststimulation to baseline. Kinetic plots were generated on live gated (Indo-1–labeled) cells using FlowJo software (Tree Star).

Of note is that staining of B cells with anti-IgD Abs to differentiate between class-switched and nonswitched memory B cells caused a slight decrease in calcium responses compared with non-IgD–labeled samples (Supplemental Fig. 2A), whereas the relative differences between Ca²⁺ responses in HD, non-Ia, and Ia patients were preserved after anti-IgD staining.

For analysis of responding CD21^low B cells of healthy controls, cells were gated first on the CD21^low B cell population according to surface markers. CD21^low B cells of healthy controls contain a substantial percentage of class-switched B cells, which, due to the absence of anti-IgD in this assay, could not be excluded by surface staining. Because by definition class-switched B cells would not respond to anti-IgM stimulation, the Ca²⁺ signal was analyzed only in B cells with a detectable Ca²⁺ signal.

After Ficoll isolation, PBMCs of six Ia, five non-Ia patients with CVID, and seven HDs were allowed to rest for 1.5 h at 37°C. Samples of 1 × 10⁶ cells were then incubated at 37°C in 10% RPMI 1640 alone or stimulated with 12.5 μg/ml F(ab′)₂ anti-IgM (Southern Biotechnology Associates) for 3 min. Cells were immediately fixed and permeabilized according to manufacturer’s protocol (BD Phosflow, BD Biosciences). Samples were stained with CD20-PerCP-Cy5.5, CD21-FITC, CD38-PE, and pPLCγ2-Alexa 647 to separate naive B cells, CD21^low B cells, and transitional B cells and with CD20-PerCP-Cy5.5, CD27-FITC, IgD-PE, and pPLCγ2-Alexa 647 to separate IgM⁺CD27⁺ B cells and switched memory B cells, respectively.

Activation assays were performed for nine Ia, six non-Ia patients with CVID, and eight HDs. A total of 1 × 10⁶ PBMCs were incubated at 37°C in 10% RPMI 1640 alone or with 12.5 μg/ml anti-IgM (MP Biomedicals, Solon, OH), 12.5 μg/ml anti-IgM plus 1 mM EDTA, 12.5 μg/ml anti-IgM plus 10 μg/ml anti-CD40 (R&D Systems, Wiesbaden, Germany), or 1 μg/ml ionomycin (Sigma-Aldrich).

After 3 d, cells were washed and stained for CD19-PE-Cy7 and CD86-PE. Stained cells were fixed, permeabilized using IntraPrep (Beckman Coulter), and stained for Ki67 expression, which served as a marker for the initiation of proliferation.

B cells of six Ia, five non-Ia patients with CVID, and five HDs were isolated by negative magnetic bead selection using the MACS B Cell Isolation Kit II (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer’s protocol. Purity of isolated B cells was above 95%. B cells were stimulated for 1 h at 37°C in RPMI 1640 plus 10% FCS medium either in the presence or absence of 12.5 μg/ml anti-IgM (MP Biomedicals), 12.5 μg/ml anti-IgM plus 3 mM EDTA, or 1 μg/ml ionomycin. After appropriate stimulations, total RNA was extracted using Trizol (Invitrogen) and followed by DNase digestion using RQ1-DNase (Promega, Madison, WI) for 30 min at 37°C. Reverse transcription was performed with oligo(dT) primers and Superscript II Reverse Transcriptase (Invitrogen). Quantitative PCR assays based on SYBR Green I detection were run using the following primer sequences (Apara Bioscience, Denzlingen, Germany): MIP-1α: 5′-CAG GTC TCC ACT GCT GCC-3′ (sense) and 5′-CAC TCA GCT CCA GGT CGC T-3′ (antisense); and CD19: 5′-CGG GAG TGG GCC CAG AAG AAG AGG-3′ (sense) and 5′-CCA GGC TGG CCC CGA ATG GAG-3′ (antisense). Amplification conditions were: activation at 95°C for 5 min and 45 cycles of amplification at 95°C for 10 s, 61.8°C for 20 s, and 72°C for 30 s, followed by subsequent melting curve analysis. CD19 gene was used as endogenous internal controls in samples.

Differences in the medians of calcium responses between HD and patient groups were analyzed by the Kruskal-Wallis test followed by the Mann-Whitney U or Wilcoxon test, where appropriate. Means of the functional assays were compared by Student t tests. The two-tailed Pearson correlation was used to analyze the relationship between calcium response and proliferation of cells. Statistical analyses were performed with Prism 5.0 (GraphPad Software, La Jolla, CA).

To investigate calcium signaling in B cells of patients with CVID, intracellular calcium levels were measured by flow cytometry pre- and poststimulation of the IgM receptor in 53 well-characterized patients with CVID and 39 HDs. After gating on total CD19⁺ B cells, all patients with CVID responded with an instant increase of intracellular calcium, but some showed a distinct reduction of the maximum peak. By subdividing the analyzed patients according to the Freiburg classification (4), we observed a pronounced decrease in Ca²⁺ responses in the subgroup corresponding to patients with CVID Ia (1.8 ± 0.2 ratio of peak/baseline Ca²⁺) compared with HD (2.1 ± 0.2) and non-Ia patients (2.2 ± 0.2) (both p < 0.001) (Fig. 1).

Patients with CVID class Ia are characterized by strongly reduced numbers of IgD⁻IgM⁻CD27⁺ class-switched memory B cells and an expansion of CD21^low cells. We asked whether the defect in Ca²⁺ signaling reflects only the altered composition of B cell populations or if it affects all B cell subtypes in patients with CVID. We analyzed naive, IgM⁺CD27⁺, transitional, and CD21^low B cells from HDs and patients with CVID for calcium responses after BCR stimulation. Naive B cells (CD19⁺CD21⁺CD27⁻CD38^interIgD⁺) as well as IgM⁺CD27⁺ B cells (CD19⁺CD21⁺CD27⁺IgD⁺) from patients with CVID Ia revealed a reduced calcium response compared with healthy individuals and non-Ia patients (p < 0.001 and p < 0.001, respectively), whereas transitional B cells (CD19⁺CD21^interCD27⁻CD38^hiIgD⁺) responded equally to BCR engagement in all groups (Fig. 2). Importantly, the strongest reduction in calcium flux of all IgM⁺ B cells was observed in the CD21^low B cell population (CD19^hiCD21⁻CD27⁻CD38^low). Their response (1.5 ± 0.2 ratio of peak/baseline Ca²⁺) was significantly reduced compared with naive B cells of the same patients (2.1 ± 0.3; p < 0.001) as well as of healthy controls (2.4 ± 0.2; p < 0.001) and non-Ia patients (2.4 ± 0.2; p < 0.001). Interestingly, CD21^low B cells of healthy donors also elicit a reduced calcium response (Supplemental Fig. 2B, 2C), suggesting that CD21^low B cells have physiologically lower Ca²⁺ signaling responses than other B cell populations.

Our data indicate that BCR-induced Ca²⁺ signals are reduced in mature B cell populations from patients with CVID class Ia. In addition, the Ca²⁺ response is strongly reduced in CD21^low B cells from both patients with CVID Ia and HDs, suggesting a cell type-specific dysregulation.

To investigate the cause of reduced Ca²⁺ mobilization in B cells from patients with CVID Ia, we analyzed BCR-mediated signaling events that are required to induce a Ca²⁺ response. Because activation of PLCγ2 is essential for the production of IP3 and Ca²⁺ release from ER stores, we tested the phosphorylation status of PLCγ2 in B lymphocytes before and after 3 min of BCR stimulation with anti-IgM (Fig. 3). No differences in the phosphorylation of PLCγ2 at tyrosine residue Y759 were observed in naive and CD21^low B cells from six CVID Ia, five non-Ia patients, and seven healthy controls, indicating that reduced Ca²⁺ responses in patients with CVID Ia are not due to a proximal BCR signaling defect. Freshly isolated, unstimulated naïve, and especially CD21^low B cells of patients with CVID Ia contained even elevated levels of pPLCγ2 compared with other B cell subsets (p < 0.001).

To determine whether the low Ca²⁺ signal results from impaired Ca²⁺ release from intracellular stores or reduced Ca²⁺ influx from the extracellular space, we next examined the Ca²⁺ response of B cells in Ca²⁺-depleted extracellular medium (Fig. 4A, 4B). The Ca²⁺ signal resulting from store depletion following anti-IgM stimulation of naive and CD21^low B cells from patients with CVID Ia was comparable to that in B cells from HDs. By contrast, Ca²⁺ influx after the readdition of CaCl₂ to the culture medium was significantly reduced in naive and CD21^low B cells from patients with CVID Ia compared with HDs (p < 0.05), suggesting that the defect in Ca²⁺ signaling in B cells from patients with CVID Ia predominantly resides in the mechanisms regulating Ca²⁺ influx from the extracellular space downstream of PLCγ2 activation and IP3 receptor-mediated release of Ca²⁺ from ER stores. This idea is corroborated by the fact that Ca²⁺ influx in CD21^low B cells, but not other mature B cell populations, from patients with CVID Ia in response to direct depletion of ER Ca²⁺ stores and activation of store-operated Ca²⁺ channels with ionomycin is significantly reduced (p < 0.01) (Fig. 4C).

Several negative regulators of Ca²⁺ signaling in B cells are known, including CD22, a member of the sialic acid-binding Ig-like lectin (SIGLEC) family, and FcγRIIB (23, 24). We asked whether decreased Ca²⁺ responses in B cells from patients with CVID Ia could be due to altered CD22 expression. Naive and especially CD21^low B cells from patients with CVID Ia showed higher expression of CD22 compared with CD21-positive B cells from HDs (p < 0.05 and p < 0.001, respectively; Fig. 4D). Interestingly, transitional B cells of patients with CVID Ia, comparable to transitional B cells from HDs, expressed lower levels of CD22 than other circulating B cell populations (data not shown). These findings are consistent with the normal Ca²⁺ responses we observed in transitional B cells from patients with CVID Ia and suggest that the increased expression of CD22 may be responsible for the reduced Ca²⁺ responses in the more mature B cell populations of patients with CVID Ia.

To determine the consequences of the altered Ca²⁺ response on potential downstream function of B cells, we investigated early events of activation and proliferation after BCR stimulation. First, we tested the upregulation of CD86 and induction of Ki67 as early activation and proliferation events in B cells from patients with CVID and HDs (Fig. 5A, 5B). Only CD86^high B cells from HDs expressed Ki67 as a sign of entering cell cycle. As expected, all patients with CVID showed a dramatically impaired activation upon crosslinking of the BCR with decreased upregulation of CD86 and nearly absent induction of Ki67, compatible with our previous finding of severely reduced proliferative response in B cells of patients with CVID Ia (25). Deficits in proliferation correlated significantly with the measured amplitude of calcium flux (p < 0.01) (Fig. 5B). Furthermore, stimulation of B cells from patients with CVID Ia with ionomycin only moderately increased CD86 expression and failed to induce cell cycle entry compared with HDs (Supplemental Fig. 3). B cells stimulated in the presence of 1 mM EDTA showed neither activation nor proliferation but mostly died in culture. Costimulation of B cells with anti-IgM and anti-CD40 did not restore activation or proliferation of CVID Ia patients’ cells (Supplemental Fig. 3). Taken together, these data demonstrate a severe defect in the activation of B cells from patients with CVID Ia that cannot be rescued by bypassing BCR proximal signaling events using ionomycin, consistent with a defect in Ca²⁺ signaling downstream of Ca²⁺ store depletion.

In addition, we analyzed BCR-induced expression of MIP-1α in B cells from five HDs, five non-Ia, and six Ia patients by quantitative RT-PCR because MIP-1α has previously been described to be expressed in a calcium-dependent manner (26–28). In patients with CVID Ia, MIP-1α expression was induced only by ∼5-fold upon stimulation with anti-IgM compared with a 19- and 12-fold induction in HDs (p = 0.03) and non-Ia patients (p = 0.05), respectively (Fig. 5C).

Noteworthy, reduced calcium signals in B cells from patients with CVID Ia correlated significantly with clinical manifestations like autoimmune cytopenia and lymphadenopathy (both p < 0.05) that indicate immunological dysregulation (Table I).

In this study, we describe for the first time a significant defect of calcium influx in B cells of the class Ia subgroup of patients with CVID, a subgroup characterized by the expansion of CD21^low B cells. Because the amplitude of the calcium response after BCR stimulation depends on the differentiation stage of B cells (29, 30) and varies between different human B cell populations (K. Warnatz, unpublished observation), we analyzed Ca²⁺ influx in each B cell subpopulation separately. This analysis revealed a defect in naive, IgM⁺CD27⁺, and CD21^low B cells but not in the more immature transitional B cells. It is thus tempting to speculate that B cells from patients with CVID Ia have a defect in a mechanism regulating Ca²⁺ influx that is specific to mature, posttransitional B cell stages.

Ca²⁺ responses in CD21^low B cells were even lower than in other mature circulating B cell populations, suggesting that additional regulatory mechanisms are present in this B cell type. The interpretation of cell type-specific regulatory mechanisms was supported by the fact that the low calcium response was not restricted to CD21^low B cells of patients with CVID but also found in CD21^low B cells of HDs. This further decrease in the calcium response is not due to the low expression of CD21 itself because B cells from a patient with a mutation in CD21 that abolishes CD21 protein expression had normal calcium influx (J. Thiel, L. Kimmig, U. Salzer, M. Grudzien, D. Lebrecht, T. Hagena, R. Draeger, N. Voelxen, A. Aichem, H. Illges, J.P. Hannan, H. Eibel, H.-H. Peter, K. Warnatz, B. Grimbacher, J.-A. Rump, and M. Schlesier, submitted for publication), indicating that CD21 is not itself required for generating a Ca²⁺ signal. In addition, in contrast to CD19 deficiency (19), which is associated with a low Ca²⁺ signal in B cells, the important signaling component of the B cell coreceptor CD19 (31) is even overexpressed on CD21^low cells (25).

Ca²⁺ responses are regulated positively and negatively by several mechanisms. Two steps lead to the rapid increase of intracellular Ca²⁺ levels after B cell activation. Following BCR engagement, activation of kinases, such as Syk and Btk, results in the phosphorylation and recruitment of PLCγ2 into the BCR signaling complex (13). We found that phosphorylation of the PLCγ2 tyrosine residue at position Y759 upon BCR stimulation is normal in naive and CD21^low B cells from patients with CVID Ia. Baseline levels of PLCγ2 phosphorylation were even significantly elevated in B cells from patients with CVID Ia, especially in the CD21^low B cell population, consistent with the previously observed preactivated status of circulating CD21^low B cells (25). PLCγ2 activation results in the production of IP3 and Ca²⁺ release from intracellular stores (13, 32). This temporary increase in intracellular calcium then leads to the opening of Ca²⁺ release-activated channels (CRAC) in the plasma membrane and calcium influx from the extracellular space (13, 32). Distinguishing both steps by activation of B cells in the presence or absence of external calcium revealed a comparable Ca²⁺ store release in patients with CVID Ia and controls consistent with normal PLCγ2 activation. By contrast, Ca²⁺ influx from the extracellular space in response to BCR stimulation was reduced. Bypassing BCR proximal signaling steps by depleting ER Ca²⁺ stores and activating CRAC channels directly with ionomycin resulted in reduced Ca²⁺ influx in CD21^low but not in naive or CD27⁺IgM⁺ B cells of type Ia patients. The discrepancy between a normal ionomycin and reduced BCR-induced Ca²⁺ influx in naive B cells from patients with CVID Ia is likely due to the fact that ionomycin results in more pronounced depletion of ER Ca²⁺ stores than BCR crosslinking. Ionomycin therefore mediates a stronger and more sustained activation of plasma membrane Ca²⁺ channels, thereby obliterating potential differences in Ca²⁺ signaling in B cells from HDs and controls. The additional negative regulatory mechanisms in CD21^low B cells of patients with CVID Ia impair the ionomycin-induced Ca²⁺ signal, pointing toward a defect at the level of CRAC channel activation.

CRAC channels are arguably the predominant plasma membrane Ca²⁺ channels in B cells. They are encoded by the gene ORAI1 and activated by stromal interaction molecule 1 (STIM1), which senses the filling state of Ca²⁺ in the ER and binds to ORAI1 upon store depletion. Lack of ORAI1 or STIM1 results in severely impaired Ca²⁺ influx in T cells and B cells from human patients with combined immunodeficiency (33–35) and mice (36–38). The predominance of impaired T cell function and normal serum Ig levels in patients lacking functional ORAI1 or STIM1 (39) and the preserved Ca²⁺ influx after ionomycin stimulation of CD21⁺ B cells of patients with CVID Ia severe ORAI1 and STIM1 deficiency are unlikely to explain the phenotype in patients with CVID Ia.

An alternative explanation for the reduced Ca²⁺ influx in B cells from patients with CVID Ia is excessive negative regulation through inhibitory B cell coreceptors, such as FcγRIIB, or members of the SIGLEC family, such as CD22 (23, 24). Interestingly, naive and especially CD21^low B cells from patients with CVID Ia showed high expression of CD22 compared with CD21-positive B cells from HDs. Transitional B cells of patients with CVID Ia express low levels of CD22 comparable to those of HDs in agreement with our finding of normal Ca²⁺ responses in transitional B cells of patients with CVID Ia. It is of note that mature follicular B cells but not transitional stage 1 B cells from CD22^−/− mice showed enhanced Ca²⁺ influx (40), indicating that the inhibitory CD22-Lyn-Src homology region 2 domain-containing phosphatase pathway is functional in mature but not immature B cells. Different pathways were discussed how CD22 regulates Ca²⁺ signals (41). One mechanism involves Src homology region 2 domain-containing phosphatase 1-mediated dephosphorylation of Vav attenuating the Ca²⁺ response upstream of Ca²⁺ release from ER stores (42). Given normal Ca²⁺ store release in naive and CD21^low B cells from patients with CVID Ia, it is unlikely that enhanced CD22 expression in patients’ B cells modulates Ca²⁺ responses through this mechanism. Alternatively, CD22 was shown to mediate Ca²⁺ efflux from the cytoplasm by activating a Ca²⁺ pump called plasma membrane Ca²⁺ ATPase-4 (43, 44), resulting in reduced intracellular Ca²⁺ concentrations (45). Future experiments will clarify the role SIGLECs play in the attenuation of the BCR-induced Ca²⁺ response in patients with CVID Ia.

So far, impaired Ca²⁺ signaling has been associated with at least three humoral immunodeficiencies. Although the pathogeneses of disturbed Ca²⁺ signaling in patients with Btk deficiency (20) and CD19 deficiency (19) are unrelated to the observed defect in patients with CVID Ia, the decreased Ca²⁺ influx in B cells from patients with WAS (21) might be related. The role of the mutated WAS protein in calcium signaling is not well understood, but a WAS-related protein, WAVE2, was shown to regulate actin reorganization and calcium influx downstream of Ca²⁺ store release in T cells, matching our findings in B cells from patients with CVID Ia (46). It is noteworthy that WAS protein mutations also lead to the expansion of CD21^low B cells (47), consistent with our finding that Ca²⁺ signaling defects are associated with the expansion of this population.

The decreased calcium response plays a role in the functional impairment of B cells from patients with CVID Ia. In this study, we could demonstrate an association between defective calcium signaling and impaired gene expression of MIP-1α, B cell activation, and proliferation, which we had reported previously (25). Therefore, impaired calcium signaling very likely contributes to the B cell dysfunction, an anergy-like phenotype of CD21^low B cells (48) and possibly the hypogammaglobulinemia in patients with CVID Ia, even though none of these responses are purely calcium dependent, and other factors are certainly involved.

Noteworthy, low Ca²⁺ flux in B cells of patients with CVID was associated with the clinical manifestation of autoimmune cytopenia and lymphadenopathy, thus suggesting that altered Ca²⁺ signals may affect immune tolerance, which is also seen in STIM1-deficient patients who may also present with autoimmune cytopenia and lymphadenopathy (49).

In conclusion, patients with CVID Ia present with significantly reduced BCR-mediated Ca²⁺ influx in all mature, posttransitional B cell populations. This reduced signaling capacity very likely contributes to the humoral immunodeficiency, altered B cell homeostasis, and immune dysregulation in these patients. The defect resides not in proximal BCR signaling events resulting in normal phosphorylation of PLCγ2 and Ca²⁺ release from internal Ca²⁺ stores, but most likely in mechanisms regulating the function of plasma membrane Ca²⁺ channels or the homeostatis of intracellular Ca²⁺ levels. A possible explanation for decreased Ca²⁺ responses is the increased expression of SIGLEC CD22 found in B cells from patients with CVID Ia. It is important to note that the negative regulatory mechanisms impairing the Ca²⁺ response of CD21^low B cells are not restricted to patients with CVID but appear to be a general physiological property of B cells at this differentiation stage. The precise characterization of the molecular defect in patients with CVID class Ia will facilitate the diagnosis of this clinically relevant subgroup of patients with CVID.

Disclosures The authors have no financial conflicts of interest.