CD22 ligation inhibits downstream B cell receptor signaling and Ca2+ flux upon activation

Authors

  • N. Sieger,

    1. Charité University Medicine Berlin and the German Rheumatism Research Center, Berlin, Germany
    Search for more papers by this author
  • S. J. Fleischer,

    1. Charité University Medicine Berlin and the German Rheumatism Research Center, Berlin, Germany
    Search for more papers by this author
  • H. E. Mei,

    1. Charité University Medicine Berlin and the German Rheumatism Research Center, Berlin, Germany
    Search for more papers by this author
  • K. Reiter,

    1. Charité University Medicine Berlin and the German Rheumatism Research Center, Berlin, Germany
    Search for more papers by this author
  • A. Shock,

    1. UCB Pharma, Slough, UK
    Search for more papers by this author
  • G. R. Burmester,

    1. Charité University Medicine Berlin and the German Rheumatism Research Center, Berlin, Germany
    Search for more papers by this author
    • Dr. Burmester has received consulting fees, speaking fees, and/or honoraria from UCB (consulting, scientific grants, and honoraria for lectures) and Immunomedics (clinical trial) (less than $10,000 each).

  • C. Daridon,

    1. Charité University Medicine Berlin and the German Rheumatism Research Center, Berlin, Germany
    Search for more papers by this author
    • Drs. Daridon and Dörner contributed equally to this work.

  • T. Dörner

    Corresponding author
    1. Charité University Medicine Berlin and the German Rheumatism Research Center, Berlin, Germany
    • Department of Medicine/Rheumatology and Clinical Immunology, Charité Berlin, Charitéplatz 1, 10098 Berlin, Germany
    Search for more papers by this author
    • Drs. Daridon and Dörner contributed equally to this work.

    • Dr. Dörner has received consulting fees, speaking fees, and/or honoraria from UCB Pharma, Roche, and Sanofi-Aventis (less than $10,000 each) and research support from Immunomedics.


Abstract

Objective

CD22 is a surface molecule exclusively expressed on B cells that regulates adhesion and B cell receptor (BCR) signaling as an inhibitory coreceptor of the BCR. Central downstream signaling molecules that are activated upon BCR engagement include spleen tyrosine kinase (Syk) and, subsequently, phospholipase Cγ2 (PLCγ2), which results in calcium (Ca2+) mobilization. The humanized anti-CD22 monoclonal antibody epratuzumab is currently being tested in clinical trials. This study was undertaken to determine the potential mechanism by which this drug regulates B cell activation.

Methods

Purified B cells were preincubated with epratuzumab, and the colocalization of CD22 and CD79α, without BCR engagement, was assessed by confocal microscopy. The phosphorylation of Syk (Y348, Y352) and PLCγ2 (Y759) as well as the Ca2+ flux in the cells were analyzed by flow cytometry upon stimulation of the BCR and/or Toll-like receptor 9 (TLR-9). The influence of CD22 ligation on BCR signaling was assessed by pretreating the cells with epratuzumab or F(ab′)2 fragment of epratuzumab, in comparison with control cells (medium alone or isotype-matched IgG1).

Results

Epratuzumab induced colocalization of CD22 and components of the BCR independent of BCR engagement, and also reduced intracellular Ca2+ mobilization and diminished the phosphorylation of Syk and PLCγ2 after BCR stimulation in vitro. Inhibition of kinase phosphorylation was demonstrated in both CD27− and CD27+ B cells, and this appeared to be independent of Fc receptor signaling. Preactivation of the cells via the stimulation of TLR-9 did not circumvent the inhibitory effect of epratuzumab on BCR signaling.

Conclusion

These findings are consistent with the concept of targeting CD22 to raise the threshold of BCR activation, which could offer therapeutic benefit in patients with autoimmune diseases.

Antigen binding to the cognate B cell receptor (BCR) induces intracellular activation signals, usually resulting in differentiation, proliferation, or, alternatively, apoptosis of B cells. An intact BCR is a requirement for differentiation of naive and, in part, memory B cells into plasma cells, based on BCR engagement together with costimulatory activation. This pathway also results in immunoglobulin class switching and somatic hypermutation, characteristics known to establish B cell memory. Furthermore, an intact antigen receptor signal is an essential precondition for B cell maturation from pre–B cells to the immature stage, and for the survival of mature B cells (1, 2).

CD22 is an inhibitory coreceptor of the BCR that is exclusively expressed on B cells (3). It plays a key role in setting the threshold of the BCR response (3–5). Engagement of CD22 initiates its phosphorylation, which subsequently activates SH2 domain–containing protein tyrosine phosphatase 1 (SHP-1) and leads to attenuation of BCR signaling via dephosphorylation of substrates of the BCR intracellular signaling pathway, including Igα, Igβ, spleen tyrosine kinase (Syk), and B cell linker protein (6, 7). One of the first events that occurs after BCR engagement is the phosphorylation of Syk, which induces a change to an active conformation.

Data from gene-knockout mice and studies in Syk-deficient cell lines have provided evidence for the importance of this signaling molecule in B cell development and for mouse viability (8). In particular, phosphorylation of tyrosine residues Y348 and Y352 is known to be essential for downstream signaling events, e.g., mobilization of calcium (Ca2+) and activation of MAP kinases (9), since both act as binding sites for further signaling molecules and also contribute to kinase function of CD22 (10). Syk, together with Bruton's tyrosine kinase, activates phospholipase Cγ2 (PLCγ2), which is responsible for the release of Ca2+ from intracellular stores—in particular, the endoplasmic reticulum (11). Subsequently, Ca2+ ion channels of the plasma membrane are triggered and regulate intracellular Ca2+ concentrations by influx of extracellular Ca2+ (12, 13).

In turn, Ca2+ flux has been reported to be negatively regulated by CD22, based on results obtained from mice lacking CD22, in which Ca2+ responses were exaggerated after BCR activation (14, 15). How CD22 reduces the BCR-induced Ca2+ flux and whether such a mechanism also applies to human B cells remain unclear. However, the ability of CD22 to regulate the intracellular Ca2+ concentration is mediated, at least in part, by triggering plasma membrane calcium ATPase 4 (PMCA-4), which enhances Ca2+ extrusion from the cytoplasm to the extracellular environment (16).

Additional important pathways of B cell activation that mainly operate on antigen-experienced memory B cells and appear to be dependent on their microenvironment include signaling via the Toll-like receptors (TLRs) and certain cytokines, such as BAFF and interleukin-21 (IL-21). Recent findings indicate that CD22 has a regulatory function in innate immune responses, via negative regulation of TLR signaling in B cells. This is based on observations that CD22-deficient B cells display increased activation and proliferation after TLR-3/TLR-4 and TLR-9 stimulation (17). TLR-9 activation is mediated via myeloid differentiation factor 88, IL-1 receptor–associated kinase 4, and tumor necrosis factor receptor–associated factor 6, and finally results in signaling through the NF-κB pathway.

Epratuzumab is being evaluated in clinical studies for the treatment of systemic lupus erythematosus (SLE), and also has been tested in patients with Sjögren's syndrome and in those with non-Hodgkin's lymphoma (18–20). Recent clinical studies have shown that epratuzumab leads to improvements in SLE patients known to have B cell hyperreactivity (18, 21). Epratuzumab is a humanized IgG1 monoclonal antibody that specifically targets CD22. In vitro studies have shown that receptor binding leads to rapid internalization of the antibody–receptor complex into the cytoplasm (22). Such studies have also shown that proliferation of isolated B cells from SLE patients is diminished in the presence of epratuzumab (23). Of note, treatment of SLE patients with the epratuzumab nondepleting antibody leads, nevertheless, to an ∼35% reduction in the number of circulating B cells, primarily CD27− B cells (23). Further data demonstrated enhanced migration toward the chemokine CXCL12 and changes in the expression and function of adhesion molecules (24). The same study showed that the expression of CD22 was substantially higher on CD27− B cells compared to CD27+ memory B cells, consistent with the more pronounced effects observed on CD27− B cells.

Since the mechanism of action of epratuzumab has not been fully delineated, the present study addressed whether epratuzumab has an effect on intracellular signaling pathways, including Ca2+ mobilization.

MATERIALS AND METHODS

Patients.

The study was approved by the ethics committee of the Charité University Hospitals, Berlin. Blood was obtained from 21 healthy donors (18 female/3 male) with a mean age of 30.4 years (range 21–56 years) and from 5 individuals undergoing orthopedic surgery.

Isolation of peripheral blood mononuclear cells (PBMCs) and B cells.

PBMCs were prepared as previously described (24). B cells were enriched from the PBMCs of healthy donors by negative sorting using the B Cell Isolation Kit II (Miltenyi Biotec).

Confocal microscopy.

Purified B cells (1 × 105) or PBMCs (1 × 106) were incubated with either Alexa Fluor 488 (A488)–conjugated epratuzumab, A488-conjugated F(ab′)2 fragment of epratuzumab, or human IgG1 (Sigma-Aldrich) at 10 μg/ml or were incubated in Glutamax RPMI 1640 medium alone (Invitrogen) for 1 hour at 37°C. The epratuzumab and F(ab′)2-epratuzumab were provided by UCB Pharma (Slough, UK). Subsequently, cells were fixed with Lyse/Fix Buffer (Becton Dickinson) for 10 minutes at 37°C. After 2 washing steps with phosphate buffered saline (PBS), cells were permeabilized by adding cold Perm Buffer II (Becton Dickinson) and kept overnight at −20°C. Cells were washed with PBS/1% fetal calf serum (FCS).

Cells that had been incubated with epratuzumab–A488 or F(ab′)2-epratuzumab–A488 were stained at room temperature (RT) for 30 minutes with anti-CD79α antibody (clone HM47; Becton Dickinson) in PBS/1% FCS. Cells that had been incubated with or without IgG1 were stained at RT for 30 minutes with the anti-CD79α antibody and with epratuzumab–A488 to visualize the distribution of CD22 on B cells. The mouse anti-CD79α antibody was further revealed by staining with rhodamine red X (RRX)–conjugated donkey anti-mouse IgG antibody (Jackson ImmunoResearch/Dianova).

The cells were cytocentrifuged at 250g for 3 minutes and covered with Vectashield Hard Set mounting medium with DAPI (Vector). Subsequently, the cells were analyzed using a Zeiss LSM 710 confocal microscope (1,000× magnification) and Zen 2009 light edition software. Colocalization was evaluated using the overlap coefficient described by Manders et al (25) (range 0–1, where 0 = no colocalization and 1 = all pixels colocalized).

Stimulation of the BCR and/or TLR-9 for intracellular phosphorylation analysis.

Epratuzumab dose.

The optimal concentration of epratuzumab (10 μg/ml) was established in previous experiments, in which a comparable dose had been found to be effective in vivo (23). Titration experiments (results not shown) provided sufficient evidence to confirm that the dose used provided good separation of the effects of epratuzumab from the effects of controls.

Stimulation of the BCR.

PBMCs (1 × 106) were incubated with either epratuzumab or F(ab′)2-epratuzumab at 10 μg/ml for 1 hour at 37°C. Treatment of PBMCs with PBS/RPMI or a nonbinding IgG1 at 10 μg/ml served as controls. Cells were stimulated with 12 μg/ml F(ab′)2 goat anti-IgM/IgG (Jackson ImmunoResearch) at 37°C for 5–8 minutes.

Stimulation of the BCR and TLR-9.

PBMCs (1 × 106) were incubated with 10 μg/ml epratuzumab for 1 hour at 37°C. CpG 2006 (2.5 μg/ml; TIB MolBiol) was added to the epratuzumab-treated or untreated PBMCs after 30 minutes. Cells were stimulated with 12 μg/ml anti-IgM/IgG (Jackson ImmunoResearch) at 37°C for 5 minutes.

Intracellular phosphospecific flow cytometry.

After stimulation, PBMCs were immediately mixed with Lyse/Fix Buffer 1× and permeabilized with Perm Buffer II according to the manufacturer's protocol (Becton Dickinson). Cells were stained at RT for 1 hour with the following targeting antibodies: phycoerythrin (PE)–conjugated p-Syk (Y348) (clone I120-722), PE-conjugated p-Syk (Y352) (clone 17a/p-ZAP70), Alexa Fluor 647–conjugated CD20 (clone H1 [FB1]), fluorescein isothiocyanate–conjugated CD27 (clone L128), and Pacific Blue (PB)–conjugated CD3 (clone UCHT1). In order to simultaneously analyze Syk and PLCγ2 phosphorylation, cells were stained with the following targeting antibodies: PE-conjugated p-Syk (Y348), A488-conjugated p-PLCγ2 (Y759), PerCP–Cy5.5–conjugated CD20, allophycocyanin-conjugated CD27, and PB-conjugated CD3 (all from Becton Dickinson). Flow cytometric analyses were performed using a Becton Dickinson Canto II flow cytometer. Data were analyzed using FlowJo software (TreeStar).

Measurement of intracellular calcium.

B cells were washed twice in RPMI medium (Invitrogen) supplemented with 10% FCS and loaded with cell permeant Indo 1-AM dye (Invitrogen) at 0.2 μM for 30–45 minutes at 37°C. B cells were washed and subsequently incubated in RPMI medium (Invitrogen) with or without F(ab′)2-epratuzumab at 10 μg/ml, and then stored on ice. Measurements were performed using a Becton Dickinson LSRII flow cytometer. Readings were recorded for 40 seconds to establish a baseline measurement, and then 12 μg/ml anti-IgM/IgG was added for BCR stimulation. Intracellular Ca2+ concentrations were recorded for 10 minutes.

To evaluate the origin of the Ca2+ flux, extracellular Ca2+ was chelated with 1 mM EGTA, and Ca2+ was added to these cells after 4.5 minutes. Changes in the intracellular Ca2+ concentration were detected as a shift from an Indo 1-AM emission peak at 475 nm for unbound dye to an emission peak at 405 nm when the Indo 1-AM molecule was bound to Ca2+. The intracellular Ca2+ concentration was quantified by calculating the ratio of Indo 1-AM emission at 405 nm to that at 475 nm (bound:unbound ratio). The intracellular Ca2+ concentration was estimated by using the measurement of the area under the curve (AUC), while Ca2+ release from the cells was evaluated from the slope of the curve after the Ca2+ peak (26).

Statistical analysis.

Paired data sets were compared using the Wilcoxon signed rank test, with results analyzed using the GraphPad Prism 4 software package. P values less than 0.05 were considered statistically significant.

RESULTS

Induction of CD22 recruitment to the BCR complex by epratuzumab.

To study the influence of CD22 on BCR signaling, an initial confocal microscopy analysis addressed whether epratuzumab induces recruitment of CD22 to the BCR. This is regarded as a functional prerequisite of CD22-dependent modulation of BCR signaling (3). A characteristic clustering and internalization of the CD22–epratuzumab complex was observed (Figure 1A). Moreover, confocal microscopy examination of the cells preincubated with epratuzumab–A488 revealed that CD22 and CD79α, an intracellular component of the BCR, were in close proximity, consistent with their colocalization and formation of a functional complex (Figure 1A). There was no nonspecific binding of the control antibody (RRX-conjugated donkey anti-mouse IgG antibody) (results not shown). When the cells were preincubated with medium alone (Figure 1B) or IgG1 (Figure1C), no clear colocalization of CD22 and CD79α was observed.

Figure 1.

Association of CD22 with CD79α after incubation of B cells with epratuzumab, independent of B cell receptor stimulation. A–C, B cells were preincubated with Alexa Fluor 488–conjugated epratuzumab (epratuzumab–A488) (A), medium alone (B), or isotype-matched control IgG1 (C) and then stained with anti-CD79α with rhodamine red X–conjugated donkey anti-mouse IgG antibody (red) and with epratuzumab–A488 (green) to visualize the distribution of CD22, using confocal microscopy. DAPI (blue) was used as counterstain. Clear colocalization (yellow) could be identified only after incubation with epratuzumab. Original magnification × 1,000. D, The colocalization of CD22 and CD79α was quantified by determining the overlap coefficient of the green staining (CD22) and red staining (CD79α) under the different treatment conditions (≥37 cells/condition) as described in A–C. Circles represent individual cell samples; horizontal bars indicate the median.

Using Zen 2009 light edition software to quantify the colocalization, single-cell analysis of the cells from 4 different healthy donors showed a significantly higher overlap coefficient (25) for the colocalization of CD22 with CD79α in B cells preincubated with epratuzumab (mean ± SD overlap coefficient 0.8 ± 0.08; n = 56) than in B cells preincubated with IgG1 (overlap coefficient 0.55 ± 0.2; n = 39) or medium alone (overlap coefficient 0.56 ± 0.1; n = 37) (P < 0.0001) (Figure 1D). Similar results were observed when the cells were preincubated with F(ab′)2-epratuzumab (results not shown). These data provide evidence that CD22 and the BCR colocalize in close proximity after incubation with epratuzumab, without direct activation of the BCR.

Effects of CD22 ligation by epratuzumab in reducing intracellular kinase phosphorylation upon BCR activation.

Since an earlier study demonstrated a reduction in B cell proliferation after various stimulation treatments (23), a central hypothesis to be tested was whether binding of epratuzumab to the inhibitory BCR coreceptor CD22 could influence downstream BCR kinase signaling. Therefore, initial studies evaluated the effects of BCR activation of CD20+ B cells on the phosphorylation status of Syk at sites Y348 and Y352. In B cells, constitutive BCR stimulation over a period of 5–8 minutes resulted in an ∼5.5-fold mean increase in the phosphorylation of Syk at site Y348 (n = 8) and site Y352 (n = 6) (Figures 2A and B). Specifically, in unstimulated cells, the mean fluorescence intensity (MFI) for p-Syk was 285.5 ± 59.8 (mean ± SD) at Y348 and 367 ± 129 at Y352, and upon activation of BCR signaling, these values increased to 1,706 ± 497.7 and 1,899 ± 658, respectively (Figures 2A and B). In comparison, as a control, T cells did not show any response.

Figure 2.

Epratuzumab inhibits B cell receptor (BCR)–induced spleen tyrosine kinase (Syk) phosphorylation in both CD27− and CD27+ B cells. Peripheral blood mononuclear cells were preincubated with medium alone, control IgG1, epratuzumab, or F(ab′)2-epratuzumab (n = 8 per group), and then stimulated with F(ab′)2–anti-IgM/IgG to achieve BCR stimulation. A, Phosphospecific flow cytometry was used to assess the phosphorylation of Syk (Y348 and Y352) in CD20+ B cells in the absence of BCR crosslinking (black) or pretreated with phosphate buffered saline (red) or IgG1 (gray) or epratuzumab (blue) and BCR stimulation. CD3+ T cells were used as a control. B, The geometric mean fluorescence intensity (MFI) for phosphorylation of Syk (Y348) was determined under the different conditions in the CD20+CD27− and CD20+CD27+ B cell subsets. Data are shown as box plots. Each box represents the 25th to 75th percentiles. Lines inside the boxes represent the median. Lines outside the boxes represent the minimum and maximum values. ∗ = P < 0.05; ∗∗ = P < 0.01 by Wilcoxon signed rank test. NS = not significant. C, The kinetics of the changes in MFI for phosphorylation of Syk (Y348) (left) and phospholipase Cγ2 (PLCγ2) (Y759) (right) were assessed in CD20+ B cells pretreated with phosphate buffered saline (red) or F(ab′)2-epratuzumab (blue) and then stimulated with F(ab′)2–anti-IgM/IgG at the indicated time points (0, 10, and 30 seconds, and 1, 2, 5, 8, and 30 minutes). Data are the mean values of at least 3 independent experiments.

To evaluate the effect of CD22 on BCR signaling in B cells (Figure 2A), PBMCs were preincubated with epratuzumab or isotype-matched control IgG1, or PBS as control, before BCR stimulation. As shown in Figure 2A, the MFI of p-Syk (Y348 and Y352) was reduced in B cells pretreated with epratuzumab compared to control B cells (PBS and IgG1-treated). Notably, there was no effect on T cells after preincubation with epratuzumab (Figure 2A).

Moreover, preincubation of the cells with epratuzumab resulted in a significant reduction in the phosphorylation of Syk (Y348) in both the CD27− and CD27+ B cell subsets (each P = 0.008 compared to controls) (Figure 2B). CD27+ memory B cells displayed a substantially higher MFI for p-Syk at baseline than did CD27− B cells (mean ± SD) MFI at baseline 339.6 ± 118.8 versus 263.9 ± 47.8). BCR stimulation induced an increase in the MFI for p-Syk of 5.8 ± 1.8–fold (mean ± SD) on CD27− B cells and 6.8 ± 2.7–fold on CD27+ B cells.

In the B cell subsets after treatment with epratuzumab, we observed a reduction in the phosphorylation of Syk at Y348 (mean ± SD 29.8 ± 13% decrease in MFI in CD27− B cells and 34 ± 19.3% decrease in MFI in CD27+ B cells). The effects seen with epratuzumab were significantly higher (P = 0.008) than those observed after exposure of the cells to IgG1 control, for both subsets of B cells (Figure 2B). In comparison to untreated cells, CD27+ B cells pretreated with IgG1 showed only a very modest inhibition of p-Syk after BCR stimulation (12.5 ± 12.4% decrease in MFI; P = 0.02); there was no significant difference in CD27− B cells (P = 0.05).

Phosphorylation of Syk at site Y352 (n = 6) showed a comparable effect after preincubation of the cells with epratuzumab. Syk phosphorylation at site Y352 was down-modulated by 33 ± 14% (mean ± SD) after treatment with epratuzumab in the CD20+ B cell population (results not shown). In CD27− B cells, the reduction in Syk phosphorylation after preincubation with epratuzumab was 30 ± 13%, while in CD27+ B cells, it was 37 ± 16%. When PBMCs were incubated with epratuzumab (10 μg/ml) in combination with IgG1 (10 μg/ml), the resulting inhibitory effect on the phosphorylation of Syk (Y352) or PLCγ2 (Y759) was similar to the effect observed with epratuzumab alone. Thus, no synergistic effect between IgG1 and epratuzumab was identified (results not shown).

To address the possibility that Fc receptor binding may have an influence on Syk phosphorylation, further experiments were carried out using the F(ab′)2-epratuzumab fragment, thereby excluding the possibility of an influence of binding to the Fcγ receptor type IIB (FcγRIIB), which is also an inhibitory coreceptor of the BCR. In such experiments, preincubation with F(ab′)2-epratuzumab, as compared to control treatments, triggered a reduction in the MFI for p-Syk (Y348) of 34.4 ± 8.5% (mean ± SD) in CD27− B cells (P = 0.008) and 31.5 ± 10.9% in CD27+ B cells (Figure 2B). The reduction in the MFI for p-Syk (Y348 and Y352) was similar between F(ab′)2-epratuzumab–treated cells and epratuzumab-treated cells (P not significant). These data are consistent with the notion that Fc-dependent pathways are not responsible for the inhibitory effect of epratuzumab, but rather that specific CD22 binding accounts for the signaling effects observed.

To evaluate whether epratuzumab treatment also affects the kinetics of the BCR response, further studies analyzed the effect of epratuzumab on the kinetics of Syk and PLCγ2 phosphorylation after BCR stimulation. In keeping with our above results, pretreatment of the cells with F(ab′)2-epratuzumab led to a substantial reduction in the phosphorylation of both Syk and PLCγ2 when compared to that in controls pretreated with PBS (Figure 2C).

The MFIs for Syk and PLCγ2 phosphorylation were reduced at all time points measured after 30 seconds of BCR engagement. The maximum MFI for p-Syk (Y348) and p-PLCγ2 (Y759) occurred between 2 and 5 minutes after BCR stimulation in untreated cells and between 5 and 8 minutes in epratuzumab-treated cells (Figure 2C), suggesting that both an overall reduction and a delay in the phosphorylation of these components will occur. It should be noted that F(ab′)2-epratuzumab treatment reduced the MFI for p-Syk by 35–47% and the MFI for p-PLCγ2 by 33–43% during the time period from 30 seconds to 30 minutes. The effects on the phosphorylation of Syk and PLCγ2 were observed in both CD27− B cells and CD27+ B cells (results not shown). Thus, preincubation with epratuzumab prevents full activation of both CD27− and CD27+ memory B cells and consistently inhibits signaling after BCR stimulation.

No influence of TLR-9 preactivation on the inhibitory effects of epratuzumab on BCR-induced Syk and PLCγ2 phosphorylation.

TLR-dependent activation of B cells represents an important pathway of B cell hyperactivity in autoimmunity (27). In our next set of studies, we addressed whether epratuzumab might be able to modulate BCR signaling pathways in TLR-9–preactivated B cells or whether the activation of TLR-9 may abrogate the inhibitory effect of epratuzumab.

Pretreatment of the cells with CpG 2006 did not have an effect on the phosphorylation of Syk (Y352) or PLCγ2 (Y759) (Figure 3A). Interestingly, only very modest increases in Syk phosphorylation (mean ± SD 8 ± 5% increase in MFI) and PLCγ2 phosphorylation (5 ± 12% increase in MFI) were observed after combined stimulation with CpG and anti-BCR antibody, as compared to stimulation with anti-BCR alone (Figure 3B). Preincubation of BCR/TLR-9–activated B cells with epratuzumab led to a reduction in p-Syk (Y352) and p-PLCγ2 (Y759) to an extent similar to that in B cells after BCR activation alone (Figure 3C). These results provide evidence that prestimulation of B cells with CpG does not prevent the reduction in phosphorylation of Syk and PLCγ2 that is observed after epratuzumab treatment.

Figure 3.

Lack of effects of Toll-like receptor 9 (TLR-9) prestimulation on B cell receptor (BCR) signaling and on the epratuzumab-induced reduction in phosphorylation after BCR stimulation. A and B, Representative histograms (A) and ratio of the mean fluorescence intensity (MFI) in stimulated cells to that in unstimulated cells (B) are shown for the phosphorylation of spleen tyrosine kinase (Syk) (Y352) (left) and phospholipase Cγ2 (PLCγ2) (Y759) (right) in CD20+ B cells under conditions of no stimulation, TLR-9 stimulation, BCR stimulation, and both TLR-9 and BCR stimulation. PE = phycoerythrin; A488 = Alexa Fluor 488. C, Comparison of the ratio of the MFI for p-Syk (Y352) (left) or p-PLCγ2 (Y759) (right) in epratuzumab-treated B cells to untreated B cells under conditions of BCR stimulation or both TLR-9 and BCR stimulation. Data in B and C are shown as box plots. Each box represents the 25th to 75th percentiles. Lines inside the boxes represent the median. Lines outside the boxes represent the minimum and maximum values.

Reduced BCR-induced Ca2+ mobilization through the binding of F(ab′)2-epratuzumab.

A further analysis assessed whether binding of epratuzumab, following the changes in Syk and PLCγ2 phosphorylation, could induce changes in the intracellular Ca2+ concentration. After the 40-second baseline measurement, the subsequent activation of the BCR induced an initially rapid increase in the intracellular Ca2+ concentration (Figure 4A). When B cells were preincubated with F(ab′)2-epratuzumab, we observed an increase in the peak intracellular Ca2+ concentration during early activation, as shown by the AUC (mean ± SD 12.1 ± 8.5% increase in bound:unbound ratio [n = 10 independent experiments]; P = 0.0002 versus BCR stimulation alone). After this initial Ca2+ peak, a long-term decrease in the intracellular Ca2+ concentration was observed in antibody-treated B cells, starting at ∼90 seconds after activation and lasting until the end of monitoring at 10 minutes (Figure 4A). Preincubation of the cells with F(ab′)2-epratuzumab also strongly reduced the overall Ca2+ concentration in the B cells, after the initial peak. Analysis of the AUC showed that there was a significant reduction in intracellular Ca2+ in the samples treated with F(ab′)2-epratuzumab (P < 0.0001 versus BCR stimulation alone) (Figure 4A).

Figure 4.

Reduced intracellular Ca2+ concentrations in B cells after incubation with F(ab′)2-epratuzumab. A, Left, Distinct Ca2+ mobilization was evident after stimulation with anti-IgM/IgG in Indo 1-AM–loaded B cells that were pretreated with F(ab′)2-epratuzumab (blue) or left untreated (red). Pretreatment of the cells with F(ab′)2-epratuzumab produced a slight increase in Ca2+ in the early phase but, most notably, a robust and longstanding reduction in Ca2+ in the late phase. Under conditions with EGTA (chelating extracellular Ca2+) added, untreated B cells (black) or F(ab′)2-epratuzumab–treated B cells (grey), which were stimulated with anti-IgM/IgG (anti–B cell receptor [anti-BCR]) at the indicated time point (arrow), did not show a significant difference in intracellular Ca2+ mobilization. Middle, Under conditions without EGTA, the intracellular Ca2+ concentration during the second phase (green areas) was significantly reduced with F(ab′)2-epratuzumab treatment. Right, This change was also evaluated by the area under the curve (AUC) (mean of 22 measurements in 9 donors). B, Similar experiments were performed starting at 40 seconds of BCR stimulation and under conditions with EGTA (to block extracellular Ca2+) in cells pretreated with F(ab′)2-epratuzumab or left untreated. Left, A reduction in the initial Ca2+ mobilization was evident in both untreated (red) and treated (blue) cells, but, most importantly, the second phase was unchanged by F(ab′)2-epratuzumab treatment. Middle, When CaCl2 (2 mM) was added at the indicated time point (arrow), there was a clear effect of F(ab′)2-epratuzumab on Ca2+ release from cells in the second phase (purple areas). Right, This effect was also evaluated by the slope of the curve within the indicated area (n = 4 donors).

Further analyses sought to better define the influence of epratuzumab on the kinetics of Ca2+ flux. Thus, cells resuspended in RPMI containing 1 mM of EGTA, which blocks extracellular Ca2+, were analyzed after BCR activation (Figure 4A). Interestingly, EGTA substantially diminished the first Ca2+ peak after BCR activation but, significantly, the second increment in Ca2+ levels was completely inhibited. These results clearly show that the regulation of the second Ca2+ phase is caused by the influx of extracellular Ca2+ (Figure 4A).

As shown in Figure 4B, adding Ca2+ to the EGTA-treated cells at 4.5 minutes led to an immediate influx of Ca2+, but a significantly faster reduction in intracellular Ca2+ was seen when cells were pretreated with F(ab′)2-epratuzumab (Figure 4B). When Ca2+ release from the cells was analyzed using the slope of the curve, a significantly lower intracellular Ca2+ concentration was observed in B cells treated with F(ab′)2-epratuzumab (P < 0.05 versus BCR stimulation alone) (Figure 4B). These results show that F(ab′)2-epratuzumab can not only inhibit kinase phosphorylation events downstream of the BCR, but also modulate Ca2+ efflux pump activity, and thus regulate Ca2+ homeostasis.

DISCUSSION

Although the precise mechanism of action of the anti-CD22 monoclonal antibody epratuzumab in autoimmunity remains to be further delineated, the current study aimed to investigate the effects of epratuzumab on BCR signaling in vitro, since previous data from CD22-knockout mice suggested a role for CD22 as an inhibitory coreceptor of the BCR (28, 29). The first observation was that epratuzumab is able to trigger colocalization of CD22 with CD79α in the absence of any BCR engagement that appeared to be specific. This finding indicates that the inhibitory effect of CD22 on BCR signaling probably requires the formation of this functional complex. Other authors have reported that synthetic sialosides are able to impair the CD22–BCR interaction, resulting in a reduced inhibitory effect of CD22 on BCR signaling (4). Thus, CD22 can influence the BCR threshold response only when it is in close proximity to, and builds a complex with, the BCR (3, 6).

When cells were pretreated with an isotype-matched control IgG1, no increased recruitment of CD22 to the BCR, relative to that in untreated cells, was observed. This finding also excludes the possibility that other binding characteristics (i.e., binding to the FcγR) can compensate for the effects observed. Thus, the current study shows that epratuzumab has a specific effect on the recruitment of CD22 to the BCR, and excludes the possibility that epratuzumab interferes with the establishment of a functional complex.

Further experiments addressed whether epratuzumab can exert an inhibitory effect on BCR signaling, as revealed by analyzing the phosphorylation of Syk (Y352 and Y348) and PLCγ2 (Y759), key molecules of this signaling cascade. In this context, the concept of tonic signaling of the BCR suggests a homeostatic interaction between continuous de novo phosphorylation of protein tyrosine kinases and termination of the signal by tyrosine phosphatases such as SHP-1 (30, 31). Epratuzumab did not change tonic Syk or PLCγ2 phosphorylation in unstimulated B cells, supporting the idea that CD22-dependent inhibition of BCR signaling is uniquely related to BCR-activated B cells.

To further analyze this concept, we compared the phosphorylation status of activated B cells after stimulation of the BCR alone with that after stimulation of the BCR and TLR-9. A key result of the present study was that the BCR-dependent increases in phosphorylation of Syk (Y348 and Y352) and PLCγ2 (Y759) were substantially reduced by epratuzumab. These data are consistent with earlier studies analyzing the effect of epratuzumab on B cells from patients with SLE, which demonstrated an inhibition of the proliferation of B cells cultured with IL-2, IL-10, CpG, and anti-BCR (23). Thus, targeting CD22 with epratuzumab inhibits downstream BCR signaling, which probably accounts for the previously observed inhibition of proliferation.

Analyses of the kinase phosphorylation of B cell subsets according to their expression of CD27 confirmed that CD27+ memory B cells exhibited a higher response to BCR stimulation than did CD27− B cells, consistent with the notion that memory B cells have a lower activation threshold than naive B cells (32, 33). In addition, a recent study elegantly showed that memory B cells (IgG–BCR) demonstrate more robust responses immediately after BCR activation than do naive B cells, including mobilization of lipid rafts, leading to the formation of submicroscopic BCR oligomers (IgM–BCR) and an increased recruitment of phosphorylated Syk to the BCR (33).

While memory B cells are more prone to respond to BCR stimuli and have a lower expression of CD22 when compared to naive B cells (24, 33), it should be emphasized that epratuzumab significantly reduced the phosphorylation of Syk and PLCγ2 in both CD27− and CD27+ populations. Thus, epratuzumab modulates BCR signaling in CD27+ memory B cells, which are found at increased frequency in SLE patients compared to healthy donors (34). Extending published results from in vitro (24) and in vivo (18, 23) studies, mainly on CD27− naive B cells, the current study is the first to show the substantial effects of epratuzumab on intracellular signaling in memory B cells.

The very modest inhibitory effects observed with the control IgG1 antibody on CD27+ B cells may be related to binding to the FcγRIIB, another inhibitory receptor that is phosphorylated upon BCR activation on its inhibitory motifs (35). However, there was no synergistic effect of IgG1 and epratuzumab on Syk or PLCγ2 phosphorylation, indicating that downstream signaling effects are specific to binding of CD22 by epratuzumab. To further exclude the possibility of immune complex binding to FcγRIIB, which also signals via Syk (36), experiments using a F(ab′)2-epratuzumab fragment were performed. These results demonstrated an equivalent reduced phosphorylation of Syk and PLCγ2. Thus, epratuzumab mediates its effects specifically through CD22 ligation and not through FcγRIIb.

In addition to reducing the amplitude of BCR stimulation, epratuzumab was able to delay the BCR response. In fact, the peak of phosphorylation of p-Syk and p-PLCγ2 was between 2 and 5 minutes after BCR stimulation, which was reproducible across different donors. Using F(ab′)2-epratuzumab, this peak of phosphorylation was delayed to 5–8 minutes after BCR stimulation. Thus, B cells treated with the anti-CD22 antibody exhibited not only an overall reduction in BCR-dependent intracellular signaling, but also a delayed response, although it needs to be established whether this delay will impair B cell differentiation in patients (34).

Given that epratuzumab clearly affected BCR-activated B cells, the influence of concomitant stimulation with CpG on the phosphorylation status was evaluated as an additional mechanism to activate B cells. There are some data from CD22-deficient mice suggesting that CD22 may modulate TLR-9 signaling, since B cells from these mice showed striking hyperproliferation after TLR-9 stimulation (17). In the current study, however, activation of TLR-9 alone had no effect on the phosphorylation of Syk and PLCγ2, consistent with previous data (17), and thus could not circumvent the inhibitory effect observed after epratuzumab treatment on BCR-induced p-Syk and p-PLCγ2. These results support the notion that epratuzumab binding to CD22 primarily modulates BCR-specific signaling events. Interestingly, we have demonstrated that another CD22-targeting antibody, clone HD39 (37), did not have any effect on the phosphorylation of BCR-induced signaling events, which would suggest that epratuzumab has a specific effect on BCR signaling (results not shown).

Finally, a detailed analysis of Ca2+ flux was performed, since it is known that activation of B cells by crosslinking of the BCR leads to mobilization of Ca2+. It is well established that the increase in cytoplasmic Ca2+ occurs in 2 phases after BCR activation (12). The first phase is characterized by an immediate and rapid increase of cytoplasmic Ca2+, caused by Ca2+ release from intracellular stores, which is activated by binding of inositol trisphosphate to its receptor. The second phase reflects the persistent influx of extracellular calcium through ion channels across the cell membrane (known as Icrac) (12, 13). Experiments using the Ca2+ chelator EGTA allowed us to differentiate these 2 phases. The data obtained using EGTA clearly demonstrated that epratuzumab had no effect on the release of intracellular Ca2+, suggesting that the increase in intracellular Ca2+ is due to an influx from outside the cells via channels, and one could speculate that epratuzumab influences the Icrac channels, thus supporting the idea that the signaling cascade downstream of Syk primarily influences Icrac channel activation, as reported by Chung and colleagues (13). However, further analyses are underway to clearly delineate which mechanism underlies the enhancement of Ca2+ in the initial phase.

A long-lasting, stable increase in intracellular Ca2+ concentration was characteristically observed after BCR stimulation. However, when B cells were treated with epratuzumab, a significant reduction was seen in intracellular Ca2+ concentration. Furthermore, analysis of the slope of the curve and the AUC suggested that epratuzumab increased the release of Ca2+ from cells, thereby reducing the intracellular Ca2+ concentration. This finding is consistent with the observation that CD22 can activate the Ca2+ pump (PMCA-4) and increases the release of Ca2+ from the cells (16). Given that the data show that experiments with the F(ab′)2-epratuzumab fragment were able to exclude the possible effects of FcR binding, the changes in intracellular Ca2+ appear to be uniquely dependent on CD22 ligation.

In conclusion, this study provides fresh insight into the complexities of BCR stimulation in B cells and the effects of epratuzumab, including induction of CD22 colocalization with the BCR, reduction of B cell activation as evidenced by diminished Syk and PLCγ2 phosphorylation, and reduction of the intracellular Ca2+ concentration. Preactivation with TLR-9 agonists did not circumvent the inhibitory effect of epratuzumab on BCR signaling. Although such intracellular signals do not appear to be inhibited fully in in vitro systems such as the one tested herein, our findings nevertheless add further insights into the mechanisms of action of CD22 targeting.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Dörner had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Sieger, Fleischer, Reiter, Shock, Burmester, Daridon, Dörner.

Acquisition of data. Sieger, Fleischer, Reiter, Daridon, Dörner.

Analysis and interpretation of data. Sieger, Fleischer, Mei, Reiter, Shock, Burmester, Daridon, Dörner.

ROLE OF THE STUDY SPONSOR

Dr. Shock is an employee of UCB Pharma. UCB Pharma provided the epratuzumab and F(ab′)2-epratuzumab, but had no role in the study design or in the collection, analysis, or interpretation of the data, the writing of the manuscript, or the decision to submit the manuscript for publication. Publication of this article was not contingent upon approval by UCB Pharma.

Acknowledgements

Final editorial assistance in the preparation of the manuscript was provided, in part, by Darwin Healthcare Communications.

Ancillary