Altered lipid raft–associated proximal signaling and translocation of CD45 tyrosine phosphatase in B lymphocytes from patients with systemic lupus erythematosus

Authors


Abstract

Objective

B lymphocytes from patients with systemic lupus erythematosus (SLE) exhibit defective intracellular signaling, hyperactivity, and autoantibody production. Recent evidence indicates a reduced expression of Lyn kinase, a negative regulator of B cell signaling, and reduced translocation of Lyn into membrane signaling domains in SLE. The present study was undertaken to investigate the causes of this altered regulation of Lyn by assessing the expression levels of regulatory molecules and their translocation into the signaling domains of SLE B lymphocytes.

Methods

Blood was obtained from 48 patients with SLE and 28 healthy controls for the assessment of B lymphocytes. Levels and intracellular distribution of Lyn, CD45, COOH-terminal Src kinase (Csk), and c-Cbl were studied by Western blotting, confocal microscopy, and flow cytometry. The kinetics of signaling molecule translocation to the B cell receptor (BCR)–antigen synapse were investigated by confocal microscopy.

Results

A profound alteration in the expression and translocation of regulatory signaling molecules in membrane domains of B cells from patients with SLE was observed. B lymphocytes from SLE patients, but not those from healthy controls, expressed a low molecular weight isoform of CD45 in lipid raft signaling microdomains. Kinetic studies revealed that translocation of Lyn, CD45, Csk, and c-Cbl led to increased recruitment and retention of Lyn and CD45 in the BCR–antigen synapse in SLE B cells.

Conclusion

The results provide evidence of altered expression and translocation/interaction of kinases and phosphatases in membrane signaling microdomains of B cells from patients with SLE. Altered translocation of CD45 correlated with reduced expression of Lyn, indicating that Lyn is a key molecule in the regulation of BCR-mediated signaling.

Systemic lupus erythematosus (SLE) is associated with many immune system abnormalities, including B lymphocyte hyperactivity and the production of autoantibodies to nuclear and cellular components (1, 2). Although the cause of SLE remains unknown, it is believed that a combination of genes, environmental factors, and hormones contributes to the predisposition to lupus autoimmunity (3, 4). Recent findings suggest that defective regulation of intracellular signaling in B or T lymphocytes could directly lead to lupus autoimmunity. For example, mice deficient in genes encoding Lyn, CD22, Fcγ receptor type II, or CD72, all of which function as negative regulators of B cell receptor (BCR) signaling, produce anti–double-stranded DNA (anti-dsDNA) autoantibodies and develop lupus (5–8). In addition, excessive positive stimulation through membrane proteins, such as complement receptor 2 (CD21), promotes B lymphocyte hyperactivity and antinuclear antibody production (9). Studies in human patients with SLE also have provided evidence of reduced levels of regulators of receptor-mediated signaling in lymphocytes (10–12).

Signaling through the BCR is regulated at several levels. Following engagement by antigen, the BCR complex translocates into cholesterol- and glycosphingolipid-enriched plasma membrane microdomains known as lipid rafts (LRs), followed by phosphorylation by Src kinases. One theory suggests that LRs act as platforms for proximal signaling (13–15). In addition to translocated BCR, LRs carry Src protein tyrosine kinases (PTKs), glycosylphosphatidylinositol-anchored proteins, and heterotrimeric G proteins.

The BCR and associated Igα and Igβ proteins (comprising the BCR complex) are phosphorylated in LRs in their immunoreceptor tyrosine–based activation motifs (ITAMs) by Src PTKs (Lyn, Fyn, and Blk) (15). The phosphorylated ITAMs serve as binding sites for Syk, which in turn becomes phosphorylated and activates downstream signaling molecules such as the B cell linker protein, Bruton's tyrosine kinase, and phospholipase Cγ2. This leads to the release of intracellular Ca2+ and the activation of protein kinase C. There is evidence of degeneracy within this system in that many Src PTKs can phosphorylate the BCR. However, Lyn also plays an indispensable negative regulatory role in proximal signaling (16). In this respect, it is notable that alterations in the level of Lyn and/or other regulatory molecules in LRs are related to lupus autoimmunity in human patients (10, 17, 18).

A current model for the role of Lyn infers that COOH-terminal Src kinase (Csk) negatively regulates Lyn by phosphorylating the COOH-terminal negative regulatory tyrosine at position 507 (Y-507) (19). In contrast, a membrane protein tyrosine phosphatase, CD45, positively regulates Lyn by dephosphorylating Y-507, which causes a conformational change that reveals the catalytic domain of Lyn and promotes autophosphorylation of the positive regulatory tyrosine at position 396 (Y-396) (20, 21). This model is supported by studies of CD45−/− B lymphocytes, in which it has been shown that Y-507 in Lyn is hyperphosphorylated and that BCR-mediated signaling is attenuated (22). However, the precise role that CD45 plays remains controversial. CD45 can also dephosphorylate Y-396 in the catalytic domain of Lyn, and thus suppresses Lyn activity (23, 24). This dual effect of CD45 may be dependent on the activation and differentiation state of B lymphocytes (25). The balance between positive and negative regulation of Lyn by CD45 may be explained by its association with signaling microdomains (25). Thus, it has been suggested that regulation of Lyn phosphorylation may be finely tuned by its association with CD45.

The objectives of this investigation were to study the kinetics of Lyn, CD45, Csk, and c-Cbl translocation into LRs in lupus B cells and determine how their differences in comparison with normal B cells could influence the biologic functioning of these cells in SLE. Our results reveal increased translocation of a low molecular weight isoform of CD45 into LRs in B cells from patients with SLE. Following engagement of the BCR, translocation of CD45 and Lyn to the BCR–antigen synapse is prolonged, and this is associated with increased Lyn phosphorylation. These observations may provide new insight into B lymphocyte abnormalities in patients with SLE.

PATIENTS AND METHODS

Patients and controls.

Forty-eight patients with SLE (44 women, 4 men; mean age 38 years, range 19–69 years) attending the rheumatology clinic at University College Hospital in London were recruited. Disease activity was determined according to the British Isles Lupus Assessment Group (BILAG) index (26). Eighteen patients with a global BILAG score of ≥6 had active disease, while 30 patients with a score of ≤5 had inactive disease. The cohort included patients treated with nonsteroidal antiinflammatory drugs, antimalarial drugs, and/or low doses of steroids (<10 mg/day) and patients treated with high doses of steroids (>10 mg/day) with or without immunosuppressive drugs. None of the patients were treated with statins, which could alter LR domains. B cells from 28 matched healthy individuals (23 women, 5 men; mean age 37 years, range 25–50 years) were used as controls. The study was carried out in compliance with the Helsinki Declaration, and patients and controls gave their informed consent. The study was approved by the local ethics committee.

Antibodies.

B cells were isolated by depleting non–B cells with a cocktail of mouse monoclonal antibodies (mAb) against CD3, CD4, CD8, CD14, CD16, and CD56 (10). The following antibodies were used for immunoprecipitation and Western blotting: rabbit anti-Lyn (sc-15), anti–c-Cbl (C-20), and anti-Csk (C-15), mouse anti-CD45 (clones 35-Z6 and T29/33 [both from Santa Cruz Biotechnology, Santa Cruz, CA] and clone DAL1/53 [from Babraham Bioscience Technologies, Cambridge, UK]), mouse anti–Flotillin-2 (clone 29; BD Biosciences, San Jose, CA), mouse anti–β-actin (AC15), horseradish peroxidase (HRP)–conjugated streptavidin (Sigma, St. Louis, MO), HRP-conjugated rabbit anti-mouse, HRP-conjugated goat anti-rabbit (Dako, Carpinteria, CA), rabbit anti–phospho-Src family (anti–Tyr-416), and rabbit anti–phospho-Lyn (anti–Tyr-507) (Cell Signaling Technology, Beverly, MA). Goat F(ab′)2 anti-human IgG, IgA, and IgM antibodies (Cappel-ICN Biomedicals, West Chester, PA), fluorescein isothiocyanate (FITC)– and phycoerythrin (PE)–conjugated mouse anti-CD45 isoform mAb (eBioscience, San Diego, CA), and FITC-conjugated goat F(ab′)2 anti-rabbit IgG (Southern Biotechnology, Birmingham, AL) were used for flow cytometry and intracellular staining.

Isolation and culture of B lymphocytes.

Peripheral blood mononuclear cells (PBMCs) were separated on Ficoll-Hypaque (Pharmacia Biotech, Uppsala, Sweden). B cells were enriched by negative selection using mAb to CD3 (1 μg/106 PBMCs), CD4, CD8, CD14, CD16, and CD56 (0.5 μg/106 PBMCs) followed by magnetic beads coated with anti-mouse IgG (Dynal, New Hyde Park, NY). Flow cytometry consistently showed that >90% of the cells were B lymphocytes. Where indicated, B lymphocytes were cultured in RPMI 1640 medium containing 10% fetal calf serum, antibiotics, L-glutamine, and sodium pyruvate (Sigma).

For experiments involving activation of B lymphocytes, stimulatory beads were prepared by coating epoxy magnetic beads (Dynal) with 10 μg/ml goat F(ab′)2 anti-IgG/IgA/IgM. For some experiments, B cells (1–2 × 105) were rested for 1 hour at 37°C and then activated with the coated beads (cell:bead ratio 1:3) at 37°C. After stimulation, B cells were either fixed and stained for confocal microscopy or lysed in 1% Triton X-100 buffer for Western blotting (10).

For verification of CD45 translocation to LRs, cell membrane proteins were biotin-labeled with EZ-link Sulfo-NHS-LC-Biotin (Pierce, Rockford, IL) and lysed in Triton X-100 buffer. The proteins were then subjected to subcellular fractionation on sucrose gradients and analyzed by Western blotting.

Preparation of LRs.

LRs were purified on a 2.5-ml 5–40% discontinuous sucrose gradient (10). Fractions 2 and 3 (5–30% interphase containing LRs) and fractions 8 and 9 (non-LR fractions) were pooled, and proteins were precipitated with methanol-chloroform.

Immunoprecipitation and Western blotting.

Proteins in LRs of 107 biotin-labeled B cells were separated on sucrose gradients and then immunoprecipitated using 5 μg of anti-CD45 mAb (clone T29/33). The precipitate was isolated with protein G–coated beads and washed with lysis buffer, resuspended in Laemmli's buffer, boiled, and resolved on reducing 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis followed by Western blotting. Bound antibodies were revealed with HRP-conjugated anti-mouse and anti-rabbit antibodies or HRP-conjugated streptavidin using enhanced chemiluminescence (Amersham Pharmacia Biotech, Little Chalfont, UK). Subsequent reprobing was performed after the bound antibodies were stripped off. Semiquantitative densitometry was carried out with the Image-Pro Plus Imaging software (Media Cybernetics, Silver Spring, MD).

Intracellular staining and confocal microscopy.

Ex vivo– and in vitro–activated B cells were fixed with 4% paraformaldehyde, washed with phosphate buffered saline (PBS), and allowed to settle overnight on poly-L-lysine–coated slides. Membrane staining for LRs was achieved with Alexa Fluor 647–conjugated cholera toxin subunit B (CTB) (25 μg/ml; Molecular Probes, Eugene, OR). Membrane staining for CD45 and intracellular staining for Lyn, c-Cbl, and Csk were performed as described previously (10). Cells were mounted with CitiFluor AF1 solution (CitiFluor, Leicester, UK) and analyzed using a Zeiss Axiovert 200M microscope (Wetzlar, Germany). Images were acquired with LSM 510 META software (Zeiss). Red or green fluorescence was recorded using a 63× objective. The extent of CTB staining and protein colocalization with CTB was quantified using the “colocalization” tool in the LSM S10 software, with results expressed as the percent of colocalizing pixels in composite merged RGB images.

Flow cytometry.

PBMCs were washed with PBS and stained with FITC-conjugated CD19 and isoform-specific PE-conjugated CD45 antibodies for 30 minutes on ice. Cells were fixed with paraformaldehyde and assessed using FACSCalibur, with results analyzed using CellQuest software (BD Biosciences).

Statistical analysis.

Differences between healthy controls and patients with SLE were assessed using the Mann-Whitney U test. P values less than 0.05 were considered significant.

RESULTS

Association of CD45 translocation to LRs with reduced Lyn expression.

We recently demonstrated that B cells from patients with SLE exhibit reduced expression levels of Lyn due to increased ubiquitination (10). In the present study, we examined both the translocation to LRs and the kinetics of Lyn, CD45, Csk, and c-Cbl in B cells from SLE patients and controls. CD45, Csk, and c-Cbl play a role in the activity and level of Lyn by regulating the pattern of its phosphorylation.

To trace LRs, B lymphocytes were stained with Alexa Fluor 647–conjugated CTB, which binds to the LR-associated GM1 gangliosides (27, 28). Confocal microscopy revealed intense staining and an irregular morphologic appearance of LRs in B cells from patients with SLE. This pattern was distinct from the pattern seen in B lymphocytes from healthy controls, which showed a light, uniform pattern of staining (Figures 1A–D and 2).

Figure 1.

Altered distribution and association of B cell receptor–associated signaling molecules with lipid rafts (LRs) in B lymphocytes from patients with systemic lupus erythematosus (SLE). A–D, B lymphocytes from SLE patients or healthy controls were fixed and stained for colocalization of LRs (red) with Lyn (A), CD45 (B), COOH-terminal Src kinase (Csk) (C), and c-Cbl (D) (all in green) and analyzed by confocal microscopy for staining patterns and areas of colocalization (yellow). Shown are representative images of cells from 1 SLE patient and 1 healthy control. Bar = 5 μm. Colocalization was also semiquantified by determining the percent of green-staining pixels (for Lyn, CD45, Csk, and c-Cbl intensity) colocalizing with red-staining pixels for the LRs (cholera toxin subunit B [CTB] intensity) (right panels in AD). E, Percent of B cells with areas of colocalization of Lyn, CD45, Csk, and c-Cbl with LR microdomains in individual SLE patients and healthy controls (bar shows the mean). The results are representative of at least 50 B cells scored for each individual studied. F, Staining and colocalization of the signaling molecules to the LRs with Lyn in relation to that with CD45 in B lymphocytes from the same patients as compared with healthy controls. Each line depicts the correlation between levels of Lyn and CD45 colocalization to the LR microdomains.

Figure 2.

A–C, Confocal microscopy showing intense and irregular staining of lipid rafts (LRs) in B lymphocytes from patients with systemic lupus erythematosus (SLE). Using the same laser intensity for each sample (at least 50 cells per sample were analyzed), B lymphocytes were either fixed and stained immediately after isolation (A), rested for 1 hour and then fixed and stained (B), or rested for 1 hour, activated with anti-IgG/IgA/IgM antibodies, and fixed and stained (C). Bar = 5 μm. D, Quantitative data for LR staining, expressed as the mean and SEM intensity (of at least 50 cells for each of the SLE patients [solid bars] and controls [open bars]) of Alexa Fluor 647–conjugated cholera toxin subunit B (CTB) binding, were determined, using Photoshop software, by dividing the intensity of red staining by the area (indicated by the number of pixels) of each individual cell. Color figure can be viewed in the online issue, which is available at http://www.arthritisrheum.org

When the colocalization of Lyn with CTB was examined (visualized as regions of yellow in the merged images in Figures 1A–D), we noticed a significant reduction of Lyn (P = 0.0005) (Figures 1A and E) in SLE B cells compared with control B cells, confirming our previous observations (10). It is unlikely that this reduction in colocalization of Lyn was only due to the reduction in cellular levels of the protein in SLE patients, since decreased Lyn colocalization was observed in both Lyn-sufficient and Lyn-deficient individuals (10).

The pattern of CTB staining in SLE B cells was disordered and revealed that proteins apparently aggregated (Figures 1A–D). In contrast to the results for Lyn, there was a significantly increased percentage of B cells from SLE patients showing intense CTB costaining with CD45, as compared with that from healthy controls (P = 0.0002) (Figures 1B and E). The extent of CD45/CTB colocalization did not correlate with disease activity (P = 0.1807) despite a tendency toward more CD45 colocalization in patients with active disease (results not shown).

Staining for Csk revealed similar, small areas of colocalization to LRs in B cells from SLE patients and in those from healthy controls (Figures 1C and E). When CTB colocalization with c-Cbl was analyzed, we observed an increased percentage of B cells from SLE patients that showed higher levels of colocalization of c-Cbl with CTB than those seen in controls. However, this increase was not statistically significantly higher compared with controls (P > 0.05) (Figures 1D and E).

The extent of CD45 colocalization to LRs correlated with the reduced expression of Lyn in the SLE patients (Figure 1F). B lymphocytes from the healthy controls had normal levels of Lyn, and only a few control cells (5–17%) had detectable CD45 in the LRs. In contrast, SLE patients had increased numbers of B lymphocytes (20–60%) with CD45 in the LRs. Importantly, there was an inverse relationship between CD45 in the LRs and Lyn levels in these microdomains (Figure 1F). This inverse correlation was not significantly associated with disease activity, nor was it influenced by medication use.

To assess the effect of treatment on the B cell phenotype, the patients were divided into 2 groups. The first group comprised patients who received no treatment and those who were treated with nonsteroidal antiinflammatory drugs, antimalarial drugs, and/or low doses (<10 mg/day) of steroids; the second group comprised patients who were treated with high doses (>10 mg/day) of steroids with or without immunosuppressive or cytotoxic drugs. No statistically significant differences in the phenotype of the B cells with regard to Lyn or CD45 expression were seen between the 2 SLE treatment groups. The levels of Csk and c-Cbl in the LRs of the patients in either treatment group and the healthy controls were not significantly different and were independent of the levels of Lyn (results not shown).

Promotion of LR aggregation via BCR engagement.

To determine whether in vivo activation could cause alterations in the structure and protein content of LRs in SLE B cells, we assessed changes in LR staining in ex vivo B cells, after a 1-hour resting period and following BCR engagement. The analysis of ex vivo B cells from 6 SLE patients and 3 healthy controls showed a significantly higher intensity of staining for LRs in B cells from the SLE patients compared with the controls (P = 0.0012) (Figures 2A and D). When the cells were subjected to a resting period for 1 hour before staining, the staining pattern became comparable between the patients and the controls (P = 0.5464) (Figures 2B and D).

To assess whether BCR crosslinking contributes to the changes in LRs, we analyzed staining with CTB after BCR engagement, using anti-IgG/IgA/IgM antibody beads. The results showed that after a resting period followed by activation for 1 hour, the intensity of LR staining increased significantly, compared with that after resting alone, in B cells from SLE patients (P = 0.0002) and in those from controls (P = 0.0003) (Figures 2C and D).

Detection of low molecular weight CD45 in LRs of SLE B lymphocytes.

To verify the confocal microscopy findings with regard to CD45 translocation, we carried out Western blot analyses (Figure 3). The results revealed detectable CD45 in the LRs of B cells from 10 of the 17 SLE patients examined (Figures 3A and F). Furthermore, Western blotting revealed that the CD45 found in the LRs was mostly of low molecular weight. This finding was verified using 2 additional mAb with specificity for CD45 (clones T29-33 and DAL1/53 [29]) (Figures 3B and D). The low molecular weight CD45 isoform was also detectable in non-LRs from B cells of some of the patients (Figures 3A and F). Subsequent experiments examining the effect of repeated separation and freeze-thawing showed that the low molecular weight of CD45 was not attributable to degradation resulting from the isolation and/or storage of the B cells (Flores-Borja F, et al: unpublished observations).

Figure 3.

Level and distribution of B cell receptor–associated signaling molecules in submembrane domains. B lymphocytes (3–4 × 106) from patients with systemic lupus erythematosus (SLE) and healthy controls (N) were lysed and fractionated on sucrose gradients. The protein contents of the lipid raft (LR) and non-LR fractions were determined by Western blotting. A, Representative blots stained for CD45 with monoclonal antibody (mAb) 35-Z6, along with corresponding blots for Lyn. Parity of protein loading was verified by the level of Flotillin-2 (Flot-2) in LRs and actin in non-LRs. B, Representative blots verifying the translocation of a low molecular weight CD45 isoform to LRs in B cells from SLE patients, using a second mAb (T29-33). C, CD45 (top) and Flot-2 (bottom) in LRs from biotin-labeled B lymphocytes, determined by immunoprecipitation (IP) (for CD45, using CD45 mAb clone T29/33), and by Western blotting (IB) for CD45 revealed with horseradish peroxidase–conjugated streptavidin (Strep-HRP). D, The same membrane as in C was stripped of bound Strep-HRP and incubated with a different CD45 mAb (clone DAL1/53). E, Blots for Lyn, COOH-terminal Src kinase (Csk), and c-Cbl in LRs and the corresponding non-LR fractions in B lymphocytes from SLE patients (n = 17) and controls (n = 10). F, Relative (rel.) mean and SEM levels of Lyn, CD45, Csk, and c-Cbl in LR (left) and non-LR (right) fractions as determined by Western blotting of B cells from SLE patients (solid bars) and controls (open bars). Relative levels were determined by comparing the intensity of bands for the signaling molecules to those of Flot-2 in LRs or actin in non-LRs.

To confirm the finding of CD45 in LRs, B cell membrane proteins from 4 SLE patients and 4 controls were labeled with biotin and then lysed and subjected to gradient fractionation. Biotinylated proteins in the LRs were precipitated with anti-CD45 antibody (clone T29-33), analyzed by Western blotting, and revealed with HRP-conjugated streptavidin. The band corresponding to the low molecular weight isoform of CD45 was detectable in B cells from the SLE patients but not in those from the controls (Figure 3C). There were, however, other proteins that coprecipitated with CD45. The finding of CD45 protein in the precipitates from the LRs of SLE B cells but not control B cells was not due to uneven protein loading, as confirmed by the results of staining for Flotillin-2 in total cell lysates before immunoprecipitation with the anti-CD45 antibody (Figure 3C, bottom).

To further confirm the identity of CD45, the membranes were stripped and reprobed with the DAL1/53 anti-CD45 mAb. Probing with this mAb verified the band corresponding to the low molecular weight isoform of CD45 in the LRs of the SLE B cells (Figure 3D). However, we could not detect Csk or c-Cbl in the LRs by Western blotting (Figures 3E and F). This could be due to the low level of these proteins in LRs.

To characterize the isoform of LR-associated CD45, we used Western blotting with mAb that were previously shown, by flow cytometry, to recognize individual CD45 isoforms. The results of these studies were inconclusive, because the mAb lost their predicted specificity on the Western blot. We therefore studied CD45 isoforms on B cells by flow cytometry to assess changes in the distribution of CD45 isoforms in patients compared with controls. The results revealed that ≥90% of the B cells from both patients (n = 11) and controls (n = 6) were CD45RO,RA+,RB+, with ∼10% CD45RB (Figure 4A). However, further analysis of the CD45RB+ population revealed that it could be subdivided into 2 populations based on the level of CD45RB (Figure 4B). By setting an arbitrary limit of fluorescence intensity of 100 to divide the 2 CD45RB+ subpopulations, we were able to identify B cells as being either CD45RBhigh or CD45RBlow (Figure 4B). The frequency of CD45RBhigh B cells in the SLE patients was significantly reduced compared with that in the controls (mean 26% versus 39%; P = 0.03), while the frequency of CD45RBlow B cells was increased in the SLE patients compared with controls (mean 73% versus 60%; P = 0.03) (Figure 4C).

Figure 4.

Flow cytometry and Western blot analyses of CD45 isoform expression in B lymphocytes. Peripheral blood mononuclear cells from 11 patients with systemic lupus erythematosus (SLE) and 6 controls were stained with fluorescein isothiocyanate–conjugated CD19 and phycoerythrin (PE)–conjugated CD45RO, CD45RA, or CD45RB monoclonal antibody. Stained cells were analyzed by flow cytometry. A, Percent of CD19+ cells negative or positive for CD45RO, CD45RA, or CD45RB. Bars show the mean and SEM. B, Representative dot plot analysis of B cells from 1 SLE patient and 1 control for the frequency of CD45RBhigh and CD45RBlow B lymphocytes. An arbitrary limit to classify cells was set at 100 units of PE fluorescence. C, Cumulative percent of CD19+ cells expressing CD45RBhigh and CD45RBlow in 11 SLE patients and 6 normal controls. Bars show the mean.

Altered kinetics of Lyn and CD45 positioning within the BCR–antigen synapse in SLE.

The kinetics of movement and patterning of signaling molecules at the interface between B cells and antigen (using anti-IgG/IgA/IgM–coated beads as surrogate antigen) was investigated to gain insight into the possible effects and extent of altered CD45 translocation after BCR engagement. B cells from 6 SLE patients and 4 controls were rested for 1 hour and then stimulated with anti-IgG/IgA/IgM beads. Following stimulation, B cell–bead complexes were fixed, stained, and analyzed by confocal microscopy.

The results revealed that Lyn was recruited to the BCR–antigen domain within 5 minutes of BCR engagement (Figure 5A). Following 10 minutes of BCR engagement, the mass of Lyn molecules diffused out of the contact domain in B cells from the healthy controls. After 30 minutes and then 60 minutes of BCR engagement, Lyn molecules were visibly excluded from the BCR–antigen contact regions in the control B cells (Figure 5A). In contrast, in B cells from the SLE patients, a delay in the recruitment and subsequent exclusion of Lyn from the engagement domain was evident, and accumulation of Lyn at the BCR–antigen contact domain was not seen until 10 minutes after BCR engagement, and once recruited, Lyn was retained for the duration of the 60-minute stimulation period (Figure 5A).

Figure 5.

Retention of CD45 at the B cell receptor (BCR)–antigen engagement domain during B lymphocyte activation in patients with systemic lupus erythematosus (SLE) compared with normal controls. B lymphocytes were rested for 1 hour at 37°C and then stimulated with goat F(ab′)2 anti-IgG/IgA/IgM antibody–coated beads for 5, 10, 30, or 60 minutes at 37°C. After activation, B lymphocytes were fixed and stained for protein translocation with antibodies to Lyn (A), CD45 (B), COOH-terminal Src kinase (Csk) (C), or c-Cbl (D) and analyzed by confocal microscopy. Shown are representative images of B cells from 6 SLE patients and 4 controls. To show the position of the coated beads in relation to the B cells, the transmission microscopy images are shown below the fluorescence images. Arrows indicate translocation of the signaling proteins to the BCR–surrogate antigen (antibody-coated beads) engagement domain (or synapse). At least 50 B lymphocyte–conjugated-bead complexes were analyzed for each sample at the different activation time points. Bar = 5 μm.

In contrast to Jurkat T cells, in which CD45 has been shown to be excluded from signaling complexes with T cell receptor (TCR) engagement by fixed antibodies (30), these experiments showed that significant numbers of CD45 molecules colocalized to the BCR–antigen contact regions. After activation, CD45 was excluded at 10 minutes after BCR engagement but then retranslocated into the contact region at 30 minutes and was retained during the activation process (Figure 5B).

The kinetics of Csk movement were similar between B cells from the SLE patients and those from the healthy controls. Csk was recruited as early as 5 minutes after BCR engagement and retained thereafter (Figure 5C). Interestingly, c-Cbl was detectable in the SLE B cell–bead contact areas as soon as 5 minutes after activation and retained for the duration of the activation course (Figure 5D). In contrast, in control B cells, c-Cbl was detectable at the engagement domain at 5 minutes and 10 minutes after BCR engagement but excluded thereafter (Figure 5D).

Increased phosphorylation of Lyn in B lymphocytes from patients with SLE.

Since CD45 regulates the phosphorylation of tyrosines in Lyn, we investigated the effect of increased CD45 translocation into the LRs of SLE B cells on the pattern of Lyn phosphorylation. B cells from 6 SLE patients and 4 controls were activated with anti-IgG/IgA/IgM beads for 1–60 minutes. The cells were lysed and phosphorylation of Lyn at Y-396 and Y-507 was determined. B cells from the SLE patients had higher levels of phosphorylated Y-396 (PY-396; active form) compared with controls (Figure 6). This was maintained for the duration of BCR engagement. In contrast, B cells from the controls had some constitutive PY-396, but, as expected, this increased after 1 minute of stimulation and then declined thereafter (Figure 6). Y-507 (PY-507; active form) was also higher in B lymphocytes from the patients compared with the controls, and the difference between SLE and normal B cells was particularly pronounced at 3 minutes (Figure 6B).

Figure 6.

Altered pattern of Lyn phosphorylation in B lymphocytes from patients with systemic lupus erythematosus (SLE) compared with normal controls. Ex vivo B lymphocytes (1–2 × 105) were rested in vitro and activated as indicated in Figure 5. The cells were lysed either before or after activation in 1% Triton X-100 buffer. The proteins were then separated by 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and the pattern of Lyn phosphorylation was analyzed. Membranes were probed for the Lyn phosphorylated tyrosine PY-396, PY-507, total Lyn, and actin (A), and the ratio of phosphorylated Lyn to total Lyn was determined (B).

DISCUSSION

We recently reported that B lymphocytes in patients with SLE exhibit reduced expression and translocation of Lyn to LRs (10). The present investigation was carried out to determine the causes of reduced Lyn expression and the effects on the biologic functioning of B cells in SLE patients. We accomplished this by studying the expression levels and kinetics of translocation into the LRs of molecules that regulate Lyn activity.

The study revealed that CD45, a key regulator of Lyn, abnormally translocates into the LRs in SLE B cells. Previous studies on whether CD45 translocates into LRs have yielded inconsistent results. The results of some studies have suggested that CD45 is excluded, while others have suggested that human leukemia cells and granulocytes transiently express CD45 in LRs (14, 31–34). More recently, we observed that CD45 translocates into LRs in T cells from patients with SLE but not healthy controls (35).

In addition to CD45 translocation into LRs, B lymphocytes from patients with SLE demonstrated altered dynamics of subcellular movement of Lyn and CD45. Previous studies of CD45 translocation into immune system synapses in T cell lines showed that the dynamic of CD45 movement is biphasic (36). Thus, there is an initial colocalization of the TCR with major histocompatibility complex molecules but very little colocalization with CD45. This is followed shortly (after 11 minutes of engagement) by extensive CD45 colocalization to the synapse. Our study reveals that in addition to basal expression in LRs, CD45 colocalized to the BCR–antigen contact regions early during BCR engagement and remained colocalized for the duration of the engagement. Although it is not clear whether the prolonged CD45/BCR colocalization was due to continuous shuttling or due to retention of translocated CD45, the accumulation of CD45 could prolong Src kinase activation and lead, ultimately, to reduced Lyn.

Interestingly, we also observed increased phosphorylation of both regulatory tyrosines in Lyn concurrent with the increased translocation of CD45, as opposed to the occurrence of a change in phosphorylation to favor an open conformation of Lyn (PY-396). Previous results have suggested that CD45 can play a negative as well as positive role in regulating Src family kinase activity (21–25, 37). Overall, the observations suggest that the stage of B lymphocyte development and site of CD45 translocation might have a bearing on the role of CD45 in Lyn regulation (25). In SLE, a disease characterized by responses to self antigens, accumulation of CD45 in LRs and BCR–antigen contact regions could promote Lyn activation and BCR signaling (38).

The study also revealed that B cells from SLE patients had large and intensely stained LRs. Similar observations were reported in an unstimulated mouse cell line in which LR aggregation induced signaling analogous to TCR stimulation (27). Interestingly, our study showed that the intensity of staining for LRs decreased upon in vitro resting. This suggests that in vivo interaction with other cells, cytokines, or other blood proteins may lead to LR aggregation. A propensity for activation might thus be presumed to be associated with LR aggregation in SLE B cells in vivo (Figure 2C). However, despite a tendency for increased LR aggregation in patients with active disease, the correlation was statistically nonsignificant. The reason for this is unknown but could reflect a lack of causal relationship between CD45/CTB colocalization and these pathologic processes.

Thus, whereas disease activity reflects the opposing effects of immunopathologic processes and treatment, increased CD45/CTB colocalization reflects the biologic functions of the cells at the time of sampling. Alternatively, disease activity assessed by the BILAG scoring system involves grading 8 organs/systems for involvement, and a lack of correlation of disease activity with CD45/CTB colocalization could obscure any association with particular disease processes or organs involved. Indeed, there is evidence that some immune system abnormalities in SLE correlate with individual clinical features. For example, anti-dsDNA autoantibodies correlate with nephritis (39).

Experiments to identify the isoform of CD45 found in LRs showed that it has a molecular weight of ∼120–135 kd. It is known that different isoforms of CD45 exist as a result of alternative splicing of exons 4, 5, and 6, which encode the A, B, and C determinants, respectively. Based on reactivity with mAb, isoforms CD45RABC (with all 3 exons), RA, RB, RC, RAB, RBC, and CD45RO (lacking all 3 exons) have been described. These isoforms are subject to posttranslational modification, including Ser/Thr O-glycosylation at the N-terminus and N-glycosylation of the extracellular domain. These modifications generate isoforms with molecular weights ranging from 180 kd to 240 kd (21). The expression of different isoforms is associated with the differentiation and activation status of myeloid cells. Thus, while progenitor/immature B lymphocytes express CD45RABC, activated mature and memory B cells express CD45RA, RB, and RO isoforms (21).

In the present study, using the available mAb that recognize CD45 isoforms in flow cytometry, it was not possible to conclusively identify the LR-associated isoform of CD45 by Western blotting. This could be due to loss of specificity of the mAb or due to lack of reactivity with the isoform. In this regard, there is evidence that in addition to exons 4, 5, and 6, exon 7 is also involved in alternative splicing, and transcripts of an isoform lacking exons 4–7 were previously reported (40). Furthermore, a protein isoform similar to the one described in our study was also reported previously (41). Alternatively, the LR-associated CD45 could be hypoglycosylated CD45RA, RB, or RO with predicted molecular weights of 135 kd, 132 kd, and 124 kd, respectively. In support of this proposition are findings of hyposialated CD45RA in B cell lymphomas and activated human cells (42). Interestingly, a recent study has shown that a mutation in the gene encoding α-mannosidase II, an enzyme involved in the branching of asparagine N-linked oligosaccharide chains (N-glycans), resulted in lupus-like disease in mice (43).

In further efforts to identify the CD45 isoform expressed in the LRs in SLE B cells, we studied the isoform expression by flow cytometry. The data revealed that ≥90% of the B cells were CD45RA+,RB+. Interestingly, however, there was a significant reduction in the proportion of CD45RBhigh B cells in the patients. The physiologic relevance of this observation is not yet known. However, one explanation could be a negative feedback on CD45RB expression by the low molecular weight isoform detected in SLE B cells. In this regard, there is evidence that alternative isoforms exert negative feedback on the expression/function of other isoforms of the protein in question (44–46). Of relevance to our study, this observation could imply that the CD45 found in the LRs is an isoform different from CD45RB. Alternatively, but not mutually exclusive, it is possible that CD45RBhigh and CD45RBlow B cells constitute different functional subsets. For example, there is evidence that rat CD45RChigh and CD45RClow constitute functionally distinct categories of memory CD4 T cells. Thus, whereas CD45RClow are short-lived memory cells that orchestrate rapid kinetics, CD45RChigh are long-lived cells that ensure the endurance of immunologic memory (47). Furthermore, rat CD45RChigh and CD45RClow CD8 cells have also been shown to have different cytokine profiles and functions (48).

In summary, this study provides evidence that a low molecular weight isoform of CD45 translocates into LRs in B cells from SLE patients. This anomalous CD45 translocation is accompanied by the altered phosphorylation and dynamics of Lyn in LRs and BCR–antigen contact regions.

AUTHOR CONTRIBUTIONS

Dr. Mageed 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 design. Drs. Flores-Borja and Mageed.

Acquisition of data. Drs. Flores-Borja, Jury, and Mageed.

Analysis and interpretation of data. Drs. Flores-Borja, Kabouridis, Isenberg, and Mageed.

Manuscript preparation. Drs. Flores-Borja, Kabouridis, Jury, Isenberg, and Mageed.

Statistical analysis. Drs. Flores-Borja and Mageed.

Ancillary