HIV Type 1 Glycoprotein 120 Inhibits Human B Cell Chemotaxis to CXC Chemokine Ligand (CXCL) 12, CC Chemokine Ligand (CCL)20, and CCL211
http://www.100md.com
免疫学杂志 2005年第13期
Abstract
We lyzed the modulation of human B cell chemotaxis by the gp120 proteins of various HIV-1 strains. X4 and X4/R5 gp120 inhibited B cell chemotaxis toward CXCL12, CCL20, and CCL21 by 40–50%, whereas R5 gp120 decreased inhibition by 20%. This gp120-induced inhibition was strictly dependent on CXCR4 or CCR5 and lipid rafts but not on CD4 or VH3-expressing BCR. Inhibition did not impair the expression or ligand-induced internalization of CCR6 and CCR7. Our data suggest that gp120/CXCR4 and gp120/CCR5 interactions lead to the cross-desensitization of CCR6 and CCR7 because gp120 does not bind CCR6 and CCR7. Unlike CXCL12, gp120 did not induce the activation of phospholipase C3 and PI3K downstream from CXCR4, whereas p38 MAPK activation was observed. Similar results were obtained if gp120-treated cells were triggered by CCL21 and CCL20. Our results are consistent with a blockade restricted to signaling pathways using phosphatidylinositol-4,5-bisphosphate as a substrate. X4 and X4/R5 gp120 induced the cleavage of CD62 ligand by a mechanism dependent on matrix metalloproteinase 1 and 3, CD4, CXCR4, Gi, and p38 MAPK, whereas R5 gp120 did not. X4 and X4/R5 gp120 also induced the relocalization of cytoplasmic CD95 to the membrane and a 23% increase in CD95-mediated apoptosis. No such effects were observed with R5 gp120. The gp120-induced decrease in B cell chemotaxis and CD62 ligand expression, and increase in CD95-mediated B cell apoptosis probably have major deleterious effects on B cell responsiveness during HIV infection and in vaccination trials.
Introduction
Human immunodeficiency virus-1 infection is associated with strong polyclonal B cell activation, increasing the percentage of activated B cells in the periphery (1, 2, 3, 4, 5), and with strong, sustained follicular hyperplasia in secondary lymphoid organs during the asymptomatic phase of the disease (6, 7, 8). The B cells of HIV-infected patients spontaneously secrete Ig but are unable to mount a T cell-dependent B cell response (1, 9, 10). Antiretroviral therapy decreases HIV-1-driven B cell hyperactivity, polyclonal B cell activation, and Ig production in patients (11). These findings strongly suggest that the sustained replication of HIV-1 affects differentiation in lymphoid tissue. The mechanisms by which HIV-1 impairs the humoral response may result from intrinsic B cell defects or dysfunctional dialogue between T and B cells. Soluble Tat and gp120, biologically active extracellular proteins released by HIV-1-infected cells, induce intrinsic defects in human B cells (12, 13, 14, 15). We previously showed that Tat selectively inhibits the BCR-mediated proliferation of naive and memory B cells, and the production of cytokines and Ig. In contrast, Tat doubles the germinal center cell proliferation induced by CD40 mAb plus IL-4 (16). The HIV-1 envelope protein gp120 may bind like a superantigen to conserved VH3 framework regions, activating and depleting these cells (17, 18). As VH3 genes are the key determinants of Abs specific for bacterial polysaccharide Ags, the depletion of VH3-expressing B cells may contribute to the observed increase in the incidence of secondary infections in HIV-infected patients (19). Scamurra et al. (10) recently showed in vivo that the depletion of peripheral VH3-expressing memory B cells, but not of naive B cells, in HIV-infected patients involves gp120 binding.
Lymphocyte recirculation, which is critical for effective immunity, is tightly regulated by the expression of adhesion molecules and chemoattractant receptors on lymphocytes, combined with the spatial and temporal expression of ligands for these receptors by a variety of tissue cells (20, 21, 22, 23). Human B cells express several chemokine receptors, including CXCR4, CXCR5, CCR6, and CCR7, and respond to their cognate ligands CXCL12, CXCL13, CCL20, and CCL21 and CCL19, respectively. Triggering of the BCR, CD40 and the IL-4R modulates chemokine receptor expression and chemotaxis in B cells (24, 25, 26). Chemokines can bind to various pertussis toxin (PTX)3-sensitive (Gi) and PTX-insensitive (Gq and G15/16) G proteins, but chemotaxis is only observed upon activation of Gi protein-coupled receptors (22). There is strong evidence that G, rather than Gi subunits initiate the chemotactic response (27). We recently showed that human B cell chemotaxis to CXCL12, CXCL13, CCL21, and CCL20 depends on PI3K, phospholipase C (PLC)3, protein kinase C (PKC), RhoA, and NF-B. Although activated by chemokine receptor-ligand interactions, neither p38 MAPK nor ERK1/2 are involved in B cell chemotaxis (28). G stimulation also activates G protein-coupled receptor kinases, which phosphorylate chemokine receptors, inducing their binding to -arrestin (29) and their rapid internalization (30).
We analyzed the effect of recombinant CXCR4-binding (X4), CCR5-binding (R5), or dual-binding (X4/R5) gp120 on the chemotactic response of human primary B cells. We found that gp120 strongly inhibited the chemotactic response to CCL20 and CCL21, without inducing spontaneous apoptosis or impairing the expression and ligand-induced internalization of CCR6 and CCR7. B cell chemotaxis was inhibited more strongly by X4 and X4/R5 gp120 than by R5 gp120, consistent with CXCR4 being the more strongly expressed of these two receptors. The use of Abs blocking CXCR4, CCR5, or gp120 totally prevented the gp120-induced inhibition of B cell chemotaxis whereas soluble CD4 slightly increased it. In contrast to what was observed with CXCL12, we found that X4 or X4/R5 gp120 did not activate the CXCR4-dependent phosphorylation of PLC3, PI3K/AKT, ERK1/2, and IB, but did induce that of p38 MAPK. The treatment of B cells with X4 or X4/R5 gp120 also blocked the CCL20- and CCL21-induced activation of PLC3, PI3K/AKT, ERK1/2, and IB, but not that of p38 MAPK. X4 and X4/R5 gp120 also modulated CD95 and L-selectin (CD62L) surface expression whereas R5 gp120 did not. CD95 expression increased as a result of the relocalization of a cytoplasmic pool to the membrane, induced by the triggering of CXCR4/Gi complexes by gp120. The decrease in CD62L surface expression was induced by the matrix metalloproteinase (MMP)-mediated cleavage of L-selectin (CD62L) via a mechanism dependent on CD4, CXCR4/Gi complexes and p38 MAPK. Therefore, gp120 strongly inhibits B cell chemotaxis and interactions with endothelial surfaces, and may be involved in decreasing the efficiency of humoral responses in HIV-infected patients.
Materials and Methods
Reagents
Soluble CD4–183 from Pharmacia, recombinant gp120 proteins from Chiron or the Division of AIDS (National Institute of Allergy and Infectious Diseases, Bethesda, MD), and anti-gp120 mAb were provided by the European Union Programme of European vaccine against AIDS (EVA)/Medical Research Council centralized facility for AIDS reagents, National Institute for Biological Standards and Control, United Kingdom (Grant QLK2-CT-1999-00609 and GP828102). Recombinant gp120 proteins were obtained from the R5 (SF162, BaL), X4 (MN, IIIB), and X4/R5 (SF2) HIV-1 strains. Preliminary experiments showed that only gp120SF162 contained endotoxins (data not shown). We therefore assessed the effects of gp120SF162 in the presence of 10 μg/ml polymyxin B throughout this work. The 2G12 mAb (IgG1) is directed against a conformational epitope of gp120 and neutralizes a wide range of R5 isolates, as well as gp120 IIIB, MN, and SF2. The 447-52D (IgG3) mAb recognizes the G protein-coupled receptor sequence at the apex of the V3 loop and neutralizes numerous laboratory strains and clinical isolates. Neutralizing 2F5 mAb (IgG1), directed against gp41 protein from various laboratory strains and clinical isolates, was used as a control. Blocking anti-CCR5 (2D7, IgG1) mAb and mouse IgG1 and IgG2b were purchased from BD Biosciences. Nonblocking anti-CCR5 mAb (45531.111, IgG2b) was obtained from R&D Systems. The CXCR4 antagonist, AMD3100, was purchased from Calbiochem (VWR International) and used at a concentration of 5 μg/ml.
B cell preparation and culture
Total B cells were obtained from palatine tonsils as previously described (31). After one cycle of rosette formation, residual T cells and monocytes were removed with CD2- and CD14-coated magnetic beads (Dynabeads M-450; Dynal Biotech). The total B cell population was then depleted of CD38+ germinal center B cells by Percoll gradient separation, as previously described (32). The resulting B cell population, including only naive and memory B cells, is referred to herein as "B cells" and is comprised of 98 ± 6% CD19+, 93 ± 3% CD44+, 6 ± 3% CD38+, and 2 ± 3% CD3+ (n = 25). The viability of these cells was consistently higher than 98%.
For in vitro culture assays, B cells (2 x 106 cells/ml) were cultured in RPMI 1640 (Invitrogen SARL) supplemented with 10 mM HEPES, 2 mM L-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, 1 mM sodium pyruvate, and 10% heat-inactivated FCS (complete medium) for 2 h, with or without recombinant 10 nM gp120 (unless otherwise indicated). Preliminary experiments showed that B cell chemotaxis was maximally inhibited after 2 h of incubation with 10 nM gp120 from the various HIV-1 strains (data not shown).
Flow cytometry
Chemokine receptor expression was analyzed by flow cytometry, using PE-conjugated anti-CCR6 (clone 53103.111, IgG2b), anti-CCR5 (clone 45531.111, IgG2b), anti-CCR7 (clone 150503, IgG2a), and anti-CXCR4 (clone 44717.111, IgG2b) mAbs from R&D Systems. PE-conjugated CD19 (IgG1) and CD95 (IgG1) mAbs and FITC- and PE-conjugated mouse isotype-matched control IgG were purchased from BD Biosciences. FITC-conjugated CD62L (clone FMC46, IgG2a) was purchased from Diaclone. To detect cytoplasmic CD95 expression, we treated B cells with the Cytofix/Cytoperm kit (BD Biosciences) before staining with PE-conjugated CD95 mAb or mouse IgG1. A FACScan flow cytometer was used for data acquisition, with CellQuest software (BD Biosciences) for data analysis. We gated on viable cells and analyzed 10,000 cells/sample. For each marker, the threshold for positivity was defined based on nonspecific binding in the presence of the relevant control IgG.
Detection of apoptotic cells
B cells were fixed and permeabilized by incubation in 70% ice-cold ethanol for at least 1 h, and washed twice in PBS. DNA was stained by incubating cells at 37°C for 1 h in 40 μg/ml propidium iodide and 100 μg/ml DNase-free RNase in PBS. Samples were analyzed by assessing FL2 red fluorescence on a linear scale. The percentage of cells undergoing apoptosis was determined as the percentage of hypodiploid cells (sub G0/G1 peak).
We evaluated the sensitivity of B cells to CD95-mediated apoptosis, by incubating medium- and gp120-treated B cells (2 x 106 cells/ml) for 24 h at 37°C with 1 μg/ml B-G27 (IgG2a, Diaclone) or mouse IgG2a as a control. B cells were then collected and the percentage of cells undergoing apoptosis was determined as described.
Blocking of CD62L shedding and increase in CD95 expression by various inhibitors
B cells were incubated with 5 μg/ml AMD3100, 1 μM herbimycin A (a tyrosine kinase inhibitor), 10 μM SB 203580 (p38 MAPK inhibitor), 100 ng/ml PTX, 10 μg/ml brefeldin A (specifically blocking protein translocation from endoplasmic reticulum to Golgi apparatus) (all from Calbiochem) or DMSO for 1 h before incubation with medium or 10 nM gp120 for 2 h. We assessed CD62L and CD95 expression by flow cytometry.
For experiments with MMP inhibitors, B cells were cultured with 500 μM EDTA, 10 μM GM-6001 (inhibitor of MMP-1, -2, -3, -8, and -9), MMP-III inhibitor (inhibitor of MMP-1, -2, -3, -7, and -13), MMP-2/9 inhibitor (all from Calbiochem) or DMSO before incubation with medium or 10 nM gp120 for 2 h. We assessed CD62L expression by flow cytometry.
In vitro chemotaxis assay
The chemotaxis assay was performed in 24-well plates (Costar) carrying transwell permeable supports, with a 5-μm polycarbonate membrane. Assays were performed in prewarmed migration buffer (RPMI 1640 plus 10 mM HEPES and 1% FCS). Migration buffer (600 μl) containing no chemokine, 250 ng/ml (CXCL12, CCL21), or 500 ng/ml (CCL20) chemokine (all from R&D Systems) was added to the lower chamber and B cells were loaded onto the inserts at a density of 0.3 x 106 cells/100 μl. Plates were incubated for 3 h at 37°C, and the number of cells migrating into the lower chamber was determined by flow cytometry. Cells from the lower chamber were centrifuged and fixed in 300 μl of 1x PBS, 1% formaldehyde and counted with the FACScan apparatus for 60 s, gating on forward and side light scatter to exclude cell debris. The number of live cells was compared with a 100% migration control in which 100 μl of cell suspension (0.3 x 106 cells) was treated in the same manner. The percentage of cells migrating to medium without chemokine was subtracted from the percentage of cells migrating to the medium with chemokines, to calculate the percentage of specific migration.
In some experiments, B cells were treated with 1 μM herbimycin A, 100 ng/ml PTX, 10 μM SB203580, or medium for 1 h before culturing with medium or 10 nM X4 or X4/R5 gp120 for 2 h. Cells were then washed and assayed for chemotaxis.
F-actin polymerization assay
Intracellular F-actin polymerization was assessed as previously described (31). Medium- and gp120-treated B cells (8 x 106/ml) were incubated in HEPES-buffered RPMI 1640 at 37°C, with or without 250 ng/ml chemokine. At the indicated times, cells (100 μl) were added to 400 μl of assay buffer containing 4 x 10–7 M FITC-labeled phalloidin, 0.5 mg/ml L--lysophosphatidylcholine (both from Sigma-Aldrich) and 4.5% formaldehyde in PBS. Fixed cells were analyzed with a FACS machine and mean fluorescence intensity (MFI) was determined for each sample. The percentage change in MFI was calculated for each sample at each time point as follows: [1 – (MFI before addition of chemokine/MFI after addition of chemokine) x 100].
In some experiments, B cells were pretreated with 2 mM methyl--cyclodextrin (MCD) or medium for 1 h before incubating with medium or 10 nM gp120 for 2 h. Cells were washed and subjected to F-actin polymerization assay. Preliminary experiments showed that 2 mM was the lowest dose of MCD inhibiting BCR- and CD40-induced B cell proliferation by >90% (data not shown).
Ligand-induced chemokine receptor internalization
Medium- and gp120-treated B cells were incubated for 30 and 60 min with medium or 100 ng/ml chemokine at 37°C. Cells were washed in ice-cold medium and stained with PE-conjugated anti-CD19, anti-CXCR4, anti-CCR7, anti-CCR6 mAbs, and PE-conjugated mouse isotype-matched control IgG for 30 min at 4°C. Single-color immunofluorescence analysis was performed on 5000 viable cells.
Western blots
We compared the abilities of CXCL12 and gp120 to trigger CXCR4 by resuspending B cells at a density of 1 x 107 cells/ml in prewarmed RPMI 1640 without FCS and stimulating them for 2 min at 37°C with medium, 100 ng/ml CXCL12 or 10 nM gp120SF2. Alternatively, medium- and gp120-treated B cells were resuspended at a density of 1 x 107 cells/ml in prewarmed RPMI 1640 without FCS and stimulated for 2 min at 37°C with medium, or 100 ng/ml CCL21 or CCL20. Lysates were prepared as previously described (32). Equal amounts of total cellular protein were subjected to SDS-PAGE and analyzed by Western blotting. Abs recognizing phospho-AKT (S473), AKT, phospho-ERK1/2 (T202/Y204), phospho-IB (S32/S36), phospho-PLC3 (S537), PLC3, phospho-p38 MAPK (T180/Y182), p38 MAPK, and pan phospho-PKC (all from New England Biolabs), PKC, PKC, PKC, IB, or ERK1/2 (from Santa Cruz Biotechnology) were used with HRP-conjugated secondary Abs. Protein bands were detected by ECL (Supersignal Westpico Chemiluminescent Substrate; Perbio Sciences). The ECL signal was recorded on ECL Hyperfilm. Films were scanned, saved as TIFF files, and analyzed with NIH Image software.
Results
Recombinant gp120 inhibits human B cell chemotaxis
The chemotactic response of B cells to CCL21 and CXCL12 was investigated after 2 h of incubation with medium or 0.7–40 nM recombinant gp120 from various HIV-1 strains (Fig. 1, A–C). Maximal gp120-induced inhibition was observed at a concentration of 10 nM gp120, regardless of the chemokine and gp120 tested. For gp120SF2, maximum inhibition was 45.2 ± 8.2% for CCL21 (n = 3) and 44 ± 8.7% (n = 3) for CXCL12 (Fig. 1A). For gp120IIIb, maximum inhibition was 34.7 ± 8.3% for CCL21 (n = 3) and 39.3 ± 6.4% (n = 3) for CXCL12 (Fig. 1B). For gp120BAL, maximum inhibition was 19 ± 0.5% for CCL21 and 22 ± 1.4% for CXCL12 (Fig. 1C). Similar results were obtained for CCL20 (data not shown). In the absence of gp120, 19 ± 6.4% (n = 8) of B cells migrated specifically in response to CCL20, 49.6 ± 9.9% (n = 15) in response to CCL21 and 41.4 ± 11.5% (n = 11) in response to CXCL12 (Fig. 1, D–F). The chemotactic response to CCL20 was inhibited by 23 ± 18.4% (n = 8), 39.9 ± 2% (n = 8), and 39.5 ± 12.8% (n = 8) using R5, X4/R5, and X4 gp120, respectively (Fig. 1D), whereas response to CCL21 was inhibited by 20.1 ± 13.7% (n = 14), 44.9 ± 10.2% (n = 14), and 49.6 ± 15.9% (n = 15), respectively (Fig. 1E). The chemotactic response to CXCL12 was decreased by 23.7 ± 11.2% (n = 8), 43.4 ± 8.3% (n = 11), and 40.0 ± 14.8% (n = 11) with R5, X4/R5, and X4 gp120, respectively (Fig. 1F). The addition of 10 μg/ml 2G12 mAb totally abolished R5 gp120-induced inhibition but only partly decreased X4 and X4/R5 gp120-induced inhibition. In contrast, 1.5 μg/ml 447-52D mAb totally abolished the inhibition of B cell chemotaxis induced by the gp120 molecules of all virus strains, whereas 10 μg/ml 2F5 mAb had no effect (Fig. 1G). B cells were cultured for 2 h with medium or 10 nM gp120 (2 h) and then for 3 h in migration buffer containing only 1% FCS (5 h). The percentage of cells undergoing apoptosis was similar in medium- and gp120SF2-treated B cells after 2 h (13.7 ± 6.4% vs 14.7 ± 7.4%, respectively) or 5 h (23.4 ± 7.8% vs 25.2 ± 8.7%, respectively) of culture, suggesting that gp120 does not increase spontaneous apoptosis (Fig. 1H). Similar results were obtained for X4 gp120 (data not shown). Thus, gp120 delivers signals that inhibit B cell chemotaxis through CXCR4 and, to a lesser extent, through CCR5. CCR7 and CCR6 did not bind gp120. Our data therefore suggest that gp120/CXCR4 and gp120/CCR5 interactions lead to cross-desensitization of CCR6 and CCR7.
Recombinant gp120 affects neither the expression nor the ligand-induced internalization of CCR6 and CCR7
B cells were stained with PE-conjugated CCR5 and CXCR4 mAbs to compare the expression of these receptors before and after incubation with recombinant gp120. CXCR4 was strongly expressed on all B cells (MFI, >400), whereas CCR5 was expressed only at low levels (MFI, 40) (Fig. 2A). A 2-h incubation with recombinant gp120SF2 induced CXCR4 internalization but did not affect the spontaneous expression of CCR6 and CCR7 on B cells (Fig. 2B).
We investigated whether ligand-induced internalization, which is important for chemotaxis, was impaired by gp120. We incubated medium- and gp120SF2-treated B cells with medium, 100 ng/ml CCL21, or 100 ng/ml CCL20 for up to 60 min at 37°C. We then placed cells at 4°C and stained them with anti-CCR7 mAb (Fig. 2C), anti-CCR6 mAb (Fig. 2D), or mouse IgG-PE for 30 min at 4°C. In the absence of chemokine, gp120SF2 did not affect the expression of CCR7 or CCR6. The gp120- and medium-treated cells showed similar levels of ligand-induced CCR7 and CCR6 internalization. Similar results were obtained for X4 gp120 (data not shown).
Receptors involved in gp120-induced inhibition of B cell chemotaxis
In addition to CXCR4 and CCR5, gp120 can interact with CD4 and VH3-expressing BCR. We therefore evaluated the effect of blocking CD4 and BCR signaling before adding gp120SF2. As expected, PTX totally abrogated the chemotaxis of medium-treated B cells, whereas the pretreatment of B cells with herbimycin A did not reverse the medium- or gp120-induced inhibition of CCL21-mediated B cell chemotaxis (Fig. 3A). Similar results were obtained for gp120MN and CCL20-mediated B cell chemotaxis (data not shown). Thus, signaling via BCR or CD4 is not required for the gp120-mediated inhibition of B cell chemotaxis.
The CCR5-blocking mAb totally reversed the inhibitory effect of recombinant R5 gp120BAL, but not that of X4/R5 gp120SF2, on CCL21-mediated B cell chemotaxis (Fig. 3B). AMD3100, which blocks CXCR4, abolished the inhibition of CCL21-induced chemotaxis observed with X4 gp120MN (Fig. 3C) or X4/R5 gp120SF2 (data not shown). AMD3100 also totally inhibited CXCL12-induced chemotaxis, independently of the presence of gp120 (Fig. 3C). This gp120-induced effect is thus mediated via CXCR4 or CCR5. The incubation of gp120 with 100 nM soluble CD4 before its addition to B cells slightly increased its potential to inhibit CCL21- (Fig. 3D), CXL12- (Fig. 3E), and CCL20-mediated (Fig. 3F) B cell chemotaxis. Thus, interactions between gp120 and membrane CD4 increase the inhibition of B cell chemotaxis by gp120, without involving CD4 signaling. The integrity of lipid rafts is crucial for efficient target cell infection. We therefore eliminated cholesterol from B cell membranes by treatment with 2 mM MCD before incubation with X4 and X4/R5 gp120. As previously shown (33, 34), this treatment did not change chemokine receptor, CD4, or BCR expression on B cells (data not shown). Levels of CCL21- (Fig. 4, A and B) and CXCL12-induced (Fig. 4C) F-actin polymerization were similar in B cells pretreated with medium and MCD. Treatment with X4/R5 and X4 gp120 strongly decreased CCL21-induced (or CCL20-induced, data not shown) F-actin polymerization at all times tested. This inhibition was abolished in B cells treated with MCD (Fig. 4, A and B). Although CXCL12 and X4 gp120 interacted with CXCR4, gp120 did not induce F-actin polymerization, even in untreated B cells (Fig. 4C). This suggests that CXCR4/gp120 and CXCR4/CXCL12 interactions have different signaling outcomes in B cells, with only the former totally dependent on lipid raft integrity.
Blockade by gp120 of chemokine-induced signaling
We compared the effects of CXCL12, gp120SF2 (Fig. 5, left panels) and gp120MN (Fig. 5, right panels) on the activation of various effectors downstream from CXCR4. Unlike CXCL12, we found that gp120SF2 and gp120MN did not activate ERK1/2, PLC-3, AKT, and PKC (, I, II, , , and n) but both CXCL12 and gp120SF2 increased the phosphorylation of p38 MAPK. Weaker p38 MAPK activation was observed if CXCR4 was triggered by gp120 rather than CXCL12. IB was not phosphorylated upon CXCR4 triggering by gp120SF2 or gp120MN (data not shown). Consistent with it effects on chemotaxis, gp120SF2 totally prevented the CCL21-induced phosphorylation of ERK1/2, AKT, IB, and PLC3 in B cells (Fig. 6). The activation of p38 MAPK in gp120-treated cells was only slightly weaker after the triggering of CCR7 by CCL21 than in medium-treated cells. Similar results were obtained for stimulation by CCL20 (data not shown).
Treatment with gp120 decreases CD62L expression but increases CD95 expression
We compared the expression of various markers on B cells cultured for 2 h with medium or with 10 nM recombinant gp120 (SF162, BAL, SF2, MN). Whatever the gp120 used, the expression of LFA-1 (CD11a/CD18), ICAM1 (CD54), CD5, CD10, CD80, CD86, CD23, CD38, CD44, CD27, SIgD, TACI, BCMA, and BAFF receptor was not significantly modified (data not shown). In contrast, treatment with X4 and X4/R5 gp120 decreased L-selectin (CD62L) expression by 66 ± 12% (n = 10) (Fig. 7A), increased CD95 expression by a factor of 4.1 ± 1.4 (n = 6) (Fig. 7B), and decreased intracytoplasmic CD95 expression (Fig. 7C). These changes were not detected after treatment with R5 gp120 (data not shown). The modulation of CD62L and CD95 expression induced by gp120 was prevented by adding 1.5 μg/ml 447-52D mAb to gp120 protein (Fig. 7D). The incubation of B cells with 5 μg/ml AMD3100 or 100 ng/ml PTX before incubation with gp120SF2 for 2 h totally prevented changes in surface CD62L (Fig. 7E) and CD95 expression (Fig. 7F). No such effect was observed with 1 μM herbimycin. Inhibitors of PI3K (wortmannin), PLC (U73122), PKC (bisindolylmaleimide I), 3-phosphoinositide-dependent kinase-1 (SH5), MEK1/2 (PD98059), and NF-B nuclear translocation (SN50) did not prevent the decrease in CD62L expression (data not shown), whereas an inhibitor of p38 MAPK (SB203580) totally reversed this effect (Fig. 7E). None of these inhibitors blocked the gp120-induced increase in CD95 expression (data not shown), whereas treatment with brefeldin A did (Fig. 7F). Prior incubation of gp120 with soluble CD4 inhibited the cleavage of CD62L (Fig. 7E), but not the increase in CD95 expression (data not shown). As L-selectin is shed from the cell surface, we incubated B cells with MMP inhibitors before adding gp120. The addition of 500 μM EDTA, 10 μM GM6001, or 10 μM MMP-III inhibitor prevented CD62L shedding (Fig. 7G). No such inhibition was observed with 10 μM MMP-2/9 inhibitor or DMSO. These data are consistent with shedding being induced by MMP-1 and MMP-3 downstream from CD4 and CXCR4/Gi complexes. Medium- and gp120-treated B cells were incubated for 24 h in the presence of 1 μg/ml B-G27 mAb, to determine whether CD95 was functional on gp120-treated B cells (Fig. 7H). The percentage of cells undergoing apoptosis increased from 35.3 ± 2% in medium-treated B cells to 58.7 ± 0.6% in gp120-treated B cells. The addition of 1.5 μg/ml 447-52D mAb to gp120 prevented the induction of apoptosis by gp120.
Discussion
Lymphocyte trafficking is crucial for immune surveillance in vivo, and requires a sequence of critical adhesion events enabling naive and memory cells to recirculate rapidly from the blood to the lymphoid organs (35). Circulating naive and memory B cells express L-selectin, which mediates the first step, adhesion to high endothelial venules, and initiates extravasion toward the lymph nodes and spleen (36). The next step is controlled by G protein-linked chemokine receptors that activate integrins (37). CXCR4, a key receptor for B cell chemotaxis, also acts as a coreceptor for HIV-1 infection. Most mature B cells express CXCR4 strongly and CD4 and CCR5 weakly at their surface (38), but are not infected by HIV-1 in vivo. B cell phenotype and functions are nevertheless rapidly impaired after infection, suggesting that virions or viral proteins released by infected cells directly affect B cell response. As gp120 modulates the chemotaxis of monocytes, dendritic cells, CD4, and CD8 T cells (39, 40, 41), we investigated whether recombinant gp120 affected the B cell chemotaxis essential for a high-quality, efficient humoral response. X4 and X4/R5 gp120 decreased, by 40–50%, not only B cell chemotaxis to CXCL12, but also that to CCL20 and CCL21, whereas neither CCR6 nor CCR7 bound gp120. The percentage of cells undergoing apoptosis was similar for medium- and gp120-treated B cells, indicating that the inhibition of B cell chemotaxis was not due to apoptosis. Consistent with the lower levels of CCR5 than CXCR4 on B cells, R5 gp120 decreased B cell chemotaxis only half as much as X4 or X4/R5 gp120. R5 gp120 inhibited B cell chemotaxis although <10% of B cells migrated in response to CCL3, CCL4, and CCL5, the main ligands of CCR5 (data not shown). Consistent with the weak chemotactic response, the triggering of CCR5 by CCL5 activated PLC3, PI3K/AKT, IB, and PKC, I, II, , , and n less strongly than did CXCL12. R5 gp120 did not activate these effectors. In contrast, the triggering of CCR5 by CCL5 or R5 gp120 led to the activation of p38 MAPK, but not ERK1/2 (data not shown). These results suggest that CCR5 is functional. It is possible that the weak chemotaxis and cell signaling are due to the low level of CCR5 expression on the surface of B cells. Blocking CXCR4 or CCR5 totally reversed the X4, X4/R5, and R5 gp120-induced inhibition of B cell chemotaxis. Similarly, blocking gp120 with the 447-52D mAb totally prevented the gp120-mediated inhibition of B cell chemotaxis, whatever the gp120 tested. CD4 signaling was dispensable for the gp120-induced inhibition of B cell chemotaxis, but prior incubation of gp120 with soluble CD4 enhanced its inhibitory effect. Thus, membrane CD4 molecules, despite being present in small numbers on B cells (42), may induce changes in the conformation of gp120, increasing its affinity for CXCR4 (or CCR5) and its inhibition of B cell chemotaxis. In contrast, the gp120-induced inhibition of monocyte chemotaxis is strictly dependent on CD4 signaling (39). No decrease in CCR6 and CCR7 expression or impairment of CCL20- and CCL21-induced chemokine receptor internalization accompanied the gp120-induced inhibition of B cell chemotaxis. Similarly, gp120 inhibited CXCL13-dependent chemotaxis in the absence of CXCR5 down-regulation or ligand-receptor internalization (data not shown). This contrasts with a previous study showing that the gp120-mediated inhibition of chemotactic response was correlated with weaker chemokine receptor expression in monocytes (39). Toth et al. (43) showed that CXCR4 dimerization was involved in CXCL12- and gp120-induced signaling events. Based on CXCR4-associated fluorescence resonance energy transfer analysis, the authors suggested that gp120 might not fully reproduce signaling events normally induced by CXCL12. The dimerization of CXCR4 and CCR5, leading to JAK2/3 activation, has been shown to play a role in chemotaxis in certain cells (44, 45) but, AG490, a potent inhibitor of JAK2/3, did not affect B cell chemotaxis or its inhibition by gp120 (data not shown). Thus, even if gp120 induces CXCR4 or CCR5 aggregation in B cells, JAK activation is not required to inhibit B cell chemotaxis. Consistent with previous data showing that efficient entry of both X4 and R5 HIV-1 requires intact lipid rafts in leukocytes (33, 46), the gp120-induced inhibition of F-actin polymerization was abolished in cholesterol-depleted B cells, whereas the chemokine-induced response was unaffected. The gp120-mediated inhibition of chemokine-induced F-actin polymerization was stronger than the inhibition of chemotaxis (70% vs 40–50% for X4 gp120). This weaker inhibition of chemotaxis is probably due to the re-expression at the membrane of gp120-free CXCR4 during the assay as no gp120 was added to the upper chamber. Several studies have shown that gp120/CXCR4 and CXCL12/CXCR4 interactions lead to different signaling profiles in T cells and monocytes (47, 48, 49). We showed that X4 or X4/R5 gp120, unlike CXCL12, did not activate PLC3, PI3K/AKT, ERK1/2, IB or PKC, I, II, , , and n in B cells. Nevertheless, B cell stimulation with CXCL12 and X4 or X4/R5 gp120 led to p38 MAPK activation by a PTX-sensitive mechanism (data not shown), suggesting that cell signaling downstream from CXCR4/Gi complexes is still partly functional. Similarly, the inhibition of CCL21- and CCL20-mediated B cell chemotaxis by gp120 correlated with a blockade of the CCL20- and CCL21-mediated activation of PLC3, PKC, PI3K/AKT, IB and ERK1/2, but not of p38 MAPK. The downstream chemokine receptor, p38 MAPK may be phosphorylated via activation of the Sos/Ras/Rac pathway by G subunits, as previously described (50), or by PKC. As CXCR4 triggering by gp120 activated p38 MAPK more weakly than CXCL12 and did not initiate PKC activation, p38 MAPK is probably activated exclusively by the Sos/Ras/Rac pathway after exposure to gp120. As the G protein-coupled receptor kinase and p38 MAPK signaling pathways are functional in gp120-treated cells, we suggest that only signaling pathways using phosphatidylinositol-4,5-bisphosphate (PIP2) as a substrate are blocked downstream from chemokine receptors. PIP2 generates inositol-1,4,5-triphosphate and diacylglycerol when hydrolyzed by PLC, and phosphatidyl-inositol-3,4,5-triphosphate when phosphorylated by PI3K. Blocking PIP2 at the plasma membrane might therefore lead to deficiencies in PI3K and PLC activity. Such a blockade could result from the sequestration to the membrane of acidic phospholipids and PIP2 by dephosphorylated myristoylated alanine-rich C kinase substrate (MARCKS) (51). MARCKS also binds cytoskeleton proteins, including actin, and calmodulin, in a Ca2+-dependent manner whereas PKC-mediated phosphorylation leads to its desorption from the membrane (52, 53, 54). Thus, gp120, unlike CXCL12, probably induces strong interactions with MARCKS, reducing the availability of PIP2 for PLC and PI3K. Consistent with this hypothesis, the destruction of lipid organization in the plasma membrane prevents the effects of gp120 on chemotaxis. As treatment with ionomycin, but not phorbol esters, reversed the gp120-induced inhibition of chemotaxis, it is possible that calmodulin protects MARCKS against PKC phosphorylation. The differential recruitment of MARCKS and activation of G proteins downstream from CXCR4 may result from differential conformational changes induced by gp120 and CXCL12. This is consistent with the work of Staudinger et al. (48), showing that gp120 does not induce conformational changes in CXCR4 similar to those induced by CXCL12. The gp120-induced increase in surface CD95 expression was totally dependent on CXCR4 and Gi heterotrimeric proteins and was induced by the relocalization of cytoplasmic CD95 to the membrane by an unknown mechanism. The gp120-induced increase in surface CD95 expression was totally prevented by anti-gp120 mAb. Such a redistribution of CD95 was recently reported in B lymphoma cell lines (55). Increased surface CD95 expression was correlated with a 23% increase in the CD95-mediated apoptosis of gp120-treated B cells, which was totally abolished by adding the neutralizing anti-gp120 447-52D mAb. Luciani et al. (56) recently showed that the gp120-induced susceptibility of human T cells to CD95-mediated apoptosis involves erzin activation and CD95/ezrin binding. Ezrin belongs to a family of proteins connecting transmembrane proteins to the actin cytoskeleton. It may therefore play a role in the gp120-mediated increase in CD95 expression and CD95-mediated apoptosis in B cells. An increase in susceptibility to CD95 apoptosis may be involved in the depletion of peripheral B cells reported by several groups in HIV-infected patients (3, 9, 10, 57). The gp120-induced decrease in membrane CD62L expression was dependent on CD4, CXCR4 triggering, and on the activation of Gi, p38 MAPK, and its cleavage by MMP. As reported by Preece et al. (58), we showed that collagenase (MMP-1) and stromelysin (MMP-3) were involved in this cleavage, whereas gelatinase A (MMP-2) and B (MMP-9) were not. The MMP-dependent cleavage of L-selectin upon cell activation (59), L-selectin cross-linking (60) or CXCR4 triggering by HIV-1 virions (61), has been described before. It results from the dissociation of calmodulin from the serine-phosphorylated cytoplasmic tail of L-selectin, leading to conformational changes in L-selectin exposing its cleavage site to MMP (62, 63, 64). Our data suggest that serine phoshorylation of CD62L by p38 MAPK, but not by PKC, I, II, , , and is involved in its dissociation from calmodulin. The finding that gp120 decreases CD62L expression and B cell chemotaxis reveals a novel mechanism by which HIV-1 can impair the humoral B cell response.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by grants from the Institut National de la Santé et de la Recherche Médicale, Université Paris-Sud, and Sidaction.
2 Address correspondence and reprint requests to Dr. Yolande Richard, Institut National de la Santé et de la Recherche Médicale, Unité 131, Université Paris-Sud, 32 rue des Carnets, 92 140 Clamart, France. E-mail address: yolande.richard@ipsc.u-psud.fr
3 Abbreviations used in this paper: PTX, pertussis toxin; PLC, phospholipase C; PKC, protein kinase C; MFI, mean fluorescence intensity; CD62L, L-selectin; MARCKS, myristoylated alanine-rich C kinase substrate; MCD, methyl--cyclodextrin; MMP, matrix metalloproteinase; PIP2, phosphatidylinositol-4,5-bisphosphate.
Received for publication November 1, 2004. Accepted for publication April 20, 2005.
References
Amadori, A., L. Chieco-Bianchi. 1990. B-cell activation and HIV-1 infection: deeds and misdeeds. Immunol. Today 11: 374-379.
Martinez-Maza, O., E. Crabb, R. T. Mitsuyasu, J. L. Fahey, J. V. Giorgi. 1987. Infection with the human immunodeficiency virus (HIV) is associated with an in vivo increase in B lymphocyte activation and immaturity. J. Immunol. 138: 3720-3724.
amuelsson, A., A. Sonnerborg, N. Heuts, J. Coster, F. Chiodi. 1997. Progressive B cell apoptosis and expression of Fas ligand during human immunodeficiency virus type 1 infection. AIDS Res. Hum. Retroviruses 13: 1031-1038.
F?rster, R., G. Schweigard, S. Johann, T. Emrich, E. Kremmer, C. Nerl, M. Lipp. 1997. Abnormal expression of the B-cell homing chemokine receptor BLR1 during the progression of acquired immunodeficiency syndrome. Blood 90: 520-525.
Beniguel, L., E. Begaud, F. Cognasse, P. Gabrie, C. D. Mbolidi, O. Sabido, M. A. Marovich, C. deFontaine, A. Fresard, F. Lucht, et al 2004. Identification of germinal center B cells in blood from HIV-infected drug-naive individuals in Central Africa. Clin. Dev. Immunol. 11: 23-27.
Racz, P., K. Tenner-Racz, C. Kahl, A. C. Feller, P. Kern, M. Dietrich. 1986. Spectrum of morphologic changes of lymph nodes from patients with AIDS or AIDS-related complexes. Prog. Allergy 37: 81-181.
Pantaleo, G., C. Graziosi, J. F. Demarest, L. Butini, M. Montroni, C. H. Fox, J. M. Orenstein, D. P. Kotler, A. S. Fauci. 1993. HIV infection is active and progressive in lymphoid tissue during the clinically latent stage of disease. Nature 362: 355-358.(Gamal Badr*, Gwenoline Bo)
We lyzed the modulation of human B cell chemotaxis by the gp120 proteins of various HIV-1 strains. X4 and X4/R5 gp120 inhibited B cell chemotaxis toward CXCL12, CCL20, and CCL21 by 40–50%, whereas R5 gp120 decreased inhibition by 20%. This gp120-induced inhibition was strictly dependent on CXCR4 or CCR5 and lipid rafts but not on CD4 or VH3-expressing BCR. Inhibition did not impair the expression or ligand-induced internalization of CCR6 and CCR7. Our data suggest that gp120/CXCR4 and gp120/CCR5 interactions lead to the cross-desensitization of CCR6 and CCR7 because gp120 does not bind CCR6 and CCR7. Unlike CXCL12, gp120 did not induce the activation of phospholipase C3 and PI3K downstream from CXCR4, whereas p38 MAPK activation was observed. Similar results were obtained if gp120-treated cells were triggered by CCL21 and CCL20. Our results are consistent with a blockade restricted to signaling pathways using phosphatidylinositol-4,5-bisphosphate as a substrate. X4 and X4/R5 gp120 induced the cleavage of CD62 ligand by a mechanism dependent on matrix metalloproteinase 1 and 3, CD4, CXCR4, Gi, and p38 MAPK, whereas R5 gp120 did not. X4 and X4/R5 gp120 also induced the relocalization of cytoplasmic CD95 to the membrane and a 23% increase in CD95-mediated apoptosis. No such effects were observed with R5 gp120. The gp120-induced decrease in B cell chemotaxis and CD62 ligand expression, and increase in CD95-mediated B cell apoptosis probably have major deleterious effects on B cell responsiveness during HIV infection and in vaccination trials.
Introduction
Human immunodeficiency virus-1 infection is associated with strong polyclonal B cell activation, increasing the percentage of activated B cells in the periphery (1, 2, 3, 4, 5), and with strong, sustained follicular hyperplasia in secondary lymphoid organs during the asymptomatic phase of the disease (6, 7, 8). The B cells of HIV-infected patients spontaneously secrete Ig but are unable to mount a T cell-dependent B cell response (1, 9, 10). Antiretroviral therapy decreases HIV-1-driven B cell hyperactivity, polyclonal B cell activation, and Ig production in patients (11). These findings strongly suggest that the sustained replication of HIV-1 affects differentiation in lymphoid tissue. The mechanisms by which HIV-1 impairs the humoral response may result from intrinsic B cell defects or dysfunctional dialogue between T and B cells. Soluble Tat and gp120, biologically active extracellular proteins released by HIV-1-infected cells, induce intrinsic defects in human B cells (12, 13, 14, 15). We previously showed that Tat selectively inhibits the BCR-mediated proliferation of naive and memory B cells, and the production of cytokines and Ig. In contrast, Tat doubles the germinal center cell proliferation induced by CD40 mAb plus IL-4 (16). The HIV-1 envelope protein gp120 may bind like a superantigen to conserved VH3 framework regions, activating and depleting these cells (17, 18). As VH3 genes are the key determinants of Abs specific for bacterial polysaccharide Ags, the depletion of VH3-expressing B cells may contribute to the observed increase in the incidence of secondary infections in HIV-infected patients (19). Scamurra et al. (10) recently showed in vivo that the depletion of peripheral VH3-expressing memory B cells, but not of naive B cells, in HIV-infected patients involves gp120 binding.
Lymphocyte recirculation, which is critical for effective immunity, is tightly regulated by the expression of adhesion molecules and chemoattractant receptors on lymphocytes, combined with the spatial and temporal expression of ligands for these receptors by a variety of tissue cells (20, 21, 22, 23). Human B cells express several chemokine receptors, including CXCR4, CXCR5, CCR6, and CCR7, and respond to their cognate ligands CXCL12, CXCL13, CCL20, and CCL21 and CCL19, respectively. Triggering of the BCR, CD40 and the IL-4R modulates chemokine receptor expression and chemotaxis in B cells (24, 25, 26). Chemokines can bind to various pertussis toxin (PTX)3-sensitive (Gi) and PTX-insensitive (Gq and G15/16) G proteins, but chemotaxis is only observed upon activation of Gi protein-coupled receptors (22). There is strong evidence that G, rather than Gi subunits initiate the chemotactic response (27). We recently showed that human B cell chemotaxis to CXCL12, CXCL13, CCL21, and CCL20 depends on PI3K, phospholipase C (PLC)3, protein kinase C (PKC), RhoA, and NF-B. Although activated by chemokine receptor-ligand interactions, neither p38 MAPK nor ERK1/2 are involved in B cell chemotaxis (28). G stimulation also activates G protein-coupled receptor kinases, which phosphorylate chemokine receptors, inducing their binding to -arrestin (29) and their rapid internalization (30).
We analyzed the effect of recombinant CXCR4-binding (X4), CCR5-binding (R5), or dual-binding (X4/R5) gp120 on the chemotactic response of human primary B cells. We found that gp120 strongly inhibited the chemotactic response to CCL20 and CCL21, without inducing spontaneous apoptosis or impairing the expression and ligand-induced internalization of CCR6 and CCR7. B cell chemotaxis was inhibited more strongly by X4 and X4/R5 gp120 than by R5 gp120, consistent with CXCR4 being the more strongly expressed of these two receptors. The use of Abs blocking CXCR4, CCR5, or gp120 totally prevented the gp120-induced inhibition of B cell chemotaxis whereas soluble CD4 slightly increased it. In contrast to what was observed with CXCL12, we found that X4 or X4/R5 gp120 did not activate the CXCR4-dependent phosphorylation of PLC3, PI3K/AKT, ERK1/2, and IB, but did induce that of p38 MAPK. The treatment of B cells with X4 or X4/R5 gp120 also blocked the CCL20- and CCL21-induced activation of PLC3, PI3K/AKT, ERK1/2, and IB, but not that of p38 MAPK. X4 and X4/R5 gp120 also modulated CD95 and L-selectin (CD62L) surface expression whereas R5 gp120 did not. CD95 expression increased as a result of the relocalization of a cytoplasmic pool to the membrane, induced by the triggering of CXCR4/Gi complexes by gp120. The decrease in CD62L surface expression was induced by the matrix metalloproteinase (MMP)-mediated cleavage of L-selectin (CD62L) via a mechanism dependent on CD4, CXCR4/Gi complexes and p38 MAPK. Therefore, gp120 strongly inhibits B cell chemotaxis and interactions with endothelial surfaces, and may be involved in decreasing the efficiency of humoral responses in HIV-infected patients.
Materials and Methods
Reagents
Soluble CD4–183 from Pharmacia, recombinant gp120 proteins from Chiron or the Division of AIDS (National Institute of Allergy and Infectious Diseases, Bethesda, MD), and anti-gp120 mAb were provided by the European Union Programme of European vaccine against AIDS (EVA)/Medical Research Council centralized facility for AIDS reagents, National Institute for Biological Standards and Control, United Kingdom (Grant QLK2-CT-1999-00609 and GP828102). Recombinant gp120 proteins were obtained from the R5 (SF162, BaL), X4 (MN, IIIB), and X4/R5 (SF2) HIV-1 strains. Preliminary experiments showed that only gp120SF162 contained endotoxins (data not shown). We therefore assessed the effects of gp120SF162 in the presence of 10 μg/ml polymyxin B throughout this work. The 2G12 mAb (IgG1) is directed against a conformational epitope of gp120 and neutralizes a wide range of R5 isolates, as well as gp120 IIIB, MN, and SF2. The 447-52D (IgG3) mAb recognizes the G protein-coupled receptor sequence at the apex of the V3 loop and neutralizes numerous laboratory strains and clinical isolates. Neutralizing 2F5 mAb (IgG1), directed against gp41 protein from various laboratory strains and clinical isolates, was used as a control. Blocking anti-CCR5 (2D7, IgG1) mAb and mouse IgG1 and IgG2b were purchased from BD Biosciences. Nonblocking anti-CCR5 mAb (45531.111, IgG2b) was obtained from R&D Systems. The CXCR4 antagonist, AMD3100, was purchased from Calbiochem (VWR International) and used at a concentration of 5 μg/ml.
B cell preparation and culture
Total B cells were obtained from palatine tonsils as previously described (31). After one cycle of rosette formation, residual T cells and monocytes were removed with CD2- and CD14-coated magnetic beads (Dynabeads M-450; Dynal Biotech). The total B cell population was then depleted of CD38+ germinal center B cells by Percoll gradient separation, as previously described (32). The resulting B cell population, including only naive and memory B cells, is referred to herein as "B cells" and is comprised of 98 ± 6% CD19+, 93 ± 3% CD44+, 6 ± 3% CD38+, and 2 ± 3% CD3+ (n = 25). The viability of these cells was consistently higher than 98%.
For in vitro culture assays, B cells (2 x 106 cells/ml) were cultured in RPMI 1640 (Invitrogen SARL) supplemented with 10 mM HEPES, 2 mM L-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin, 1 mM sodium pyruvate, and 10% heat-inactivated FCS (complete medium) for 2 h, with or without recombinant 10 nM gp120 (unless otherwise indicated). Preliminary experiments showed that B cell chemotaxis was maximally inhibited after 2 h of incubation with 10 nM gp120 from the various HIV-1 strains (data not shown).
Flow cytometry
Chemokine receptor expression was analyzed by flow cytometry, using PE-conjugated anti-CCR6 (clone 53103.111, IgG2b), anti-CCR5 (clone 45531.111, IgG2b), anti-CCR7 (clone 150503, IgG2a), and anti-CXCR4 (clone 44717.111, IgG2b) mAbs from R&D Systems. PE-conjugated CD19 (IgG1) and CD95 (IgG1) mAbs and FITC- and PE-conjugated mouse isotype-matched control IgG were purchased from BD Biosciences. FITC-conjugated CD62L (clone FMC46, IgG2a) was purchased from Diaclone. To detect cytoplasmic CD95 expression, we treated B cells with the Cytofix/Cytoperm kit (BD Biosciences) before staining with PE-conjugated CD95 mAb or mouse IgG1. A FACScan flow cytometer was used for data acquisition, with CellQuest software (BD Biosciences) for data analysis. We gated on viable cells and analyzed 10,000 cells/sample. For each marker, the threshold for positivity was defined based on nonspecific binding in the presence of the relevant control IgG.
Detection of apoptotic cells
B cells were fixed and permeabilized by incubation in 70% ice-cold ethanol for at least 1 h, and washed twice in PBS. DNA was stained by incubating cells at 37°C for 1 h in 40 μg/ml propidium iodide and 100 μg/ml DNase-free RNase in PBS. Samples were analyzed by assessing FL2 red fluorescence on a linear scale. The percentage of cells undergoing apoptosis was determined as the percentage of hypodiploid cells (sub G0/G1 peak).
We evaluated the sensitivity of B cells to CD95-mediated apoptosis, by incubating medium- and gp120-treated B cells (2 x 106 cells/ml) for 24 h at 37°C with 1 μg/ml B-G27 (IgG2a, Diaclone) or mouse IgG2a as a control. B cells were then collected and the percentage of cells undergoing apoptosis was determined as described.
Blocking of CD62L shedding and increase in CD95 expression by various inhibitors
B cells were incubated with 5 μg/ml AMD3100, 1 μM herbimycin A (a tyrosine kinase inhibitor), 10 μM SB 203580 (p38 MAPK inhibitor), 100 ng/ml PTX, 10 μg/ml brefeldin A (specifically blocking protein translocation from endoplasmic reticulum to Golgi apparatus) (all from Calbiochem) or DMSO for 1 h before incubation with medium or 10 nM gp120 for 2 h. We assessed CD62L and CD95 expression by flow cytometry.
For experiments with MMP inhibitors, B cells were cultured with 500 μM EDTA, 10 μM GM-6001 (inhibitor of MMP-1, -2, -3, -8, and -9), MMP-III inhibitor (inhibitor of MMP-1, -2, -3, -7, and -13), MMP-2/9 inhibitor (all from Calbiochem) or DMSO before incubation with medium or 10 nM gp120 for 2 h. We assessed CD62L expression by flow cytometry.
In vitro chemotaxis assay
The chemotaxis assay was performed in 24-well plates (Costar) carrying transwell permeable supports, with a 5-μm polycarbonate membrane. Assays were performed in prewarmed migration buffer (RPMI 1640 plus 10 mM HEPES and 1% FCS). Migration buffer (600 μl) containing no chemokine, 250 ng/ml (CXCL12, CCL21), or 500 ng/ml (CCL20) chemokine (all from R&D Systems) was added to the lower chamber and B cells were loaded onto the inserts at a density of 0.3 x 106 cells/100 μl. Plates were incubated for 3 h at 37°C, and the number of cells migrating into the lower chamber was determined by flow cytometry. Cells from the lower chamber were centrifuged and fixed in 300 μl of 1x PBS, 1% formaldehyde and counted with the FACScan apparatus for 60 s, gating on forward and side light scatter to exclude cell debris. The number of live cells was compared with a 100% migration control in which 100 μl of cell suspension (0.3 x 106 cells) was treated in the same manner. The percentage of cells migrating to medium without chemokine was subtracted from the percentage of cells migrating to the medium with chemokines, to calculate the percentage of specific migration.
In some experiments, B cells were treated with 1 μM herbimycin A, 100 ng/ml PTX, 10 μM SB203580, or medium for 1 h before culturing with medium or 10 nM X4 or X4/R5 gp120 for 2 h. Cells were then washed and assayed for chemotaxis.
F-actin polymerization assay
Intracellular F-actin polymerization was assessed as previously described (31). Medium- and gp120-treated B cells (8 x 106/ml) were incubated in HEPES-buffered RPMI 1640 at 37°C, with or without 250 ng/ml chemokine. At the indicated times, cells (100 μl) were added to 400 μl of assay buffer containing 4 x 10–7 M FITC-labeled phalloidin, 0.5 mg/ml L--lysophosphatidylcholine (both from Sigma-Aldrich) and 4.5% formaldehyde in PBS. Fixed cells were analyzed with a FACS machine and mean fluorescence intensity (MFI) was determined for each sample. The percentage change in MFI was calculated for each sample at each time point as follows: [1 – (MFI before addition of chemokine/MFI after addition of chemokine) x 100].
In some experiments, B cells were pretreated with 2 mM methyl--cyclodextrin (MCD) or medium for 1 h before incubating with medium or 10 nM gp120 for 2 h. Cells were washed and subjected to F-actin polymerization assay. Preliminary experiments showed that 2 mM was the lowest dose of MCD inhibiting BCR- and CD40-induced B cell proliferation by >90% (data not shown).
Ligand-induced chemokine receptor internalization
Medium- and gp120-treated B cells were incubated for 30 and 60 min with medium or 100 ng/ml chemokine at 37°C. Cells were washed in ice-cold medium and stained with PE-conjugated anti-CD19, anti-CXCR4, anti-CCR7, anti-CCR6 mAbs, and PE-conjugated mouse isotype-matched control IgG for 30 min at 4°C. Single-color immunofluorescence analysis was performed on 5000 viable cells.
Western blots
We compared the abilities of CXCL12 and gp120 to trigger CXCR4 by resuspending B cells at a density of 1 x 107 cells/ml in prewarmed RPMI 1640 without FCS and stimulating them for 2 min at 37°C with medium, 100 ng/ml CXCL12 or 10 nM gp120SF2. Alternatively, medium- and gp120-treated B cells were resuspended at a density of 1 x 107 cells/ml in prewarmed RPMI 1640 without FCS and stimulated for 2 min at 37°C with medium, or 100 ng/ml CCL21 or CCL20. Lysates were prepared as previously described (32). Equal amounts of total cellular protein were subjected to SDS-PAGE and analyzed by Western blotting. Abs recognizing phospho-AKT (S473), AKT, phospho-ERK1/2 (T202/Y204), phospho-IB (S32/S36), phospho-PLC3 (S537), PLC3, phospho-p38 MAPK (T180/Y182), p38 MAPK, and pan phospho-PKC (all from New England Biolabs), PKC, PKC, PKC, IB, or ERK1/2 (from Santa Cruz Biotechnology) were used with HRP-conjugated secondary Abs. Protein bands were detected by ECL (Supersignal Westpico Chemiluminescent Substrate; Perbio Sciences). The ECL signal was recorded on ECL Hyperfilm. Films were scanned, saved as TIFF files, and analyzed with NIH Image software.
Results
Recombinant gp120 inhibits human B cell chemotaxis
The chemotactic response of B cells to CCL21 and CXCL12 was investigated after 2 h of incubation with medium or 0.7–40 nM recombinant gp120 from various HIV-1 strains (Fig. 1, A–C). Maximal gp120-induced inhibition was observed at a concentration of 10 nM gp120, regardless of the chemokine and gp120 tested. For gp120SF2, maximum inhibition was 45.2 ± 8.2% for CCL21 (n = 3) and 44 ± 8.7% (n = 3) for CXCL12 (Fig. 1A). For gp120IIIb, maximum inhibition was 34.7 ± 8.3% for CCL21 (n = 3) and 39.3 ± 6.4% (n = 3) for CXCL12 (Fig. 1B). For gp120BAL, maximum inhibition was 19 ± 0.5% for CCL21 and 22 ± 1.4% for CXCL12 (Fig. 1C). Similar results were obtained for CCL20 (data not shown). In the absence of gp120, 19 ± 6.4% (n = 8) of B cells migrated specifically in response to CCL20, 49.6 ± 9.9% (n = 15) in response to CCL21 and 41.4 ± 11.5% (n = 11) in response to CXCL12 (Fig. 1, D–F). The chemotactic response to CCL20 was inhibited by 23 ± 18.4% (n = 8), 39.9 ± 2% (n = 8), and 39.5 ± 12.8% (n = 8) using R5, X4/R5, and X4 gp120, respectively (Fig. 1D), whereas response to CCL21 was inhibited by 20.1 ± 13.7% (n = 14), 44.9 ± 10.2% (n = 14), and 49.6 ± 15.9% (n = 15), respectively (Fig. 1E). The chemotactic response to CXCL12 was decreased by 23.7 ± 11.2% (n = 8), 43.4 ± 8.3% (n = 11), and 40.0 ± 14.8% (n = 11) with R5, X4/R5, and X4 gp120, respectively (Fig. 1F). The addition of 10 μg/ml 2G12 mAb totally abolished R5 gp120-induced inhibition but only partly decreased X4 and X4/R5 gp120-induced inhibition. In contrast, 1.5 μg/ml 447-52D mAb totally abolished the inhibition of B cell chemotaxis induced by the gp120 molecules of all virus strains, whereas 10 μg/ml 2F5 mAb had no effect (Fig. 1G). B cells were cultured for 2 h with medium or 10 nM gp120 (2 h) and then for 3 h in migration buffer containing only 1% FCS (5 h). The percentage of cells undergoing apoptosis was similar in medium- and gp120SF2-treated B cells after 2 h (13.7 ± 6.4% vs 14.7 ± 7.4%, respectively) or 5 h (23.4 ± 7.8% vs 25.2 ± 8.7%, respectively) of culture, suggesting that gp120 does not increase spontaneous apoptosis (Fig. 1H). Similar results were obtained for X4 gp120 (data not shown). Thus, gp120 delivers signals that inhibit B cell chemotaxis through CXCR4 and, to a lesser extent, through CCR5. CCR7 and CCR6 did not bind gp120. Our data therefore suggest that gp120/CXCR4 and gp120/CCR5 interactions lead to cross-desensitization of CCR6 and CCR7.
Recombinant gp120 affects neither the expression nor the ligand-induced internalization of CCR6 and CCR7
B cells were stained with PE-conjugated CCR5 and CXCR4 mAbs to compare the expression of these receptors before and after incubation with recombinant gp120. CXCR4 was strongly expressed on all B cells (MFI, >400), whereas CCR5 was expressed only at low levels (MFI, 40) (Fig. 2A). A 2-h incubation with recombinant gp120SF2 induced CXCR4 internalization but did not affect the spontaneous expression of CCR6 and CCR7 on B cells (Fig. 2B).
We investigated whether ligand-induced internalization, which is important for chemotaxis, was impaired by gp120. We incubated medium- and gp120SF2-treated B cells with medium, 100 ng/ml CCL21, or 100 ng/ml CCL20 for up to 60 min at 37°C. We then placed cells at 4°C and stained them with anti-CCR7 mAb (Fig. 2C), anti-CCR6 mAb (Fig. 2D), or mouse IgG-PE for 30 min at 4°C. In the absence of chemokine, gp120SF2 did not affect the expression of CCR7 or CCR6. The gp120- and medium-treated cells showed similar levels of ligand-induced CCR7 and CCR6 internalization. Similar results were obtained for X4 gp120 (data not shown).
Receptors involved in gp120-induced inhibition of B cell chemotaxis
In addition to CXCR4 and CCR5, gp120 can interact with CD4 and VH3-expressing BCR. We therefore evaluated the effect of blocking CD4 and BCR signaling before adding gp120SF2. As expected, PTX totally abrogated the chemotaxis of medium-treated B cells, whereas the pretreatment of B cells with herbimycin A did not reverse the medium- or gp120-induced inhibition of CCL21-mediated B cell chemotaxis (Fig. 3A). Similar results were obtained for gp120MN and CCL20-mediated B cell chemotaxis (data not shown). Thus, signaling via BCR or CD4 is not required for the gp120-mediated inhibition of B cell chemotaxis.
The CCR5-blocking mAb totally reversed the inhibitory effect of recombinant R5 gp120BAL, but not that of X4/R5 gp120SF2, on CCL21-mediated B cell chemotaxis (Fig. 3B). AMD3100, which blocks CXCR4, abolished the inhibition of CCL21-induced chemotaxis observed with X4 gp120MN (Fig. 3C) or X4/R5 gp120SF2 (data not shown). AMD3100 also totally inhibited CXCL12-induced chemotaxis, independently of the presence of gp120 (Fig. 3C). This gp120-induced effect is thus mediated via CXCR4 or CCR5. The incubation of gp120 with 100 nM soluble CD4 before its addition to B cells slightly increased its potential to inhibit CCL21- (Fig. 3D), CXL12- (Fig. 3E), and CCL20-mediated (Fig. 3F) B cell chemotaxis. Thus, interactions between gp120 and membrane CD4 increase the inhibition of B cell chemotaxis by gp120, without involving CD4 signaling. The integrity of lipid rafts is crucial for efficient target cell infection. We therefore eliminated cholesterol from B cell membranes by treatment with 2 mM MCD before incubation with X4 and X4/R5 gp120. As previously shown (33, 34), this treatment did not change chemokine receptor, CD4, or BCR expression on B cells (data not shown). Levels of CCL21- (Fig. 4, A and B) and CXCL12-induced (Fig. 4C) F-actin polymerization were similar in B cells pretreated with medium and MCD. Treatment with X4/R5 and X4 gp120 strongly decreased CCL21-induced (or CCL20-induced, data not shown) F-actin polymerization at all times tested. This inhibition was abolished in B cells treated with MCD (Fig. 4, A and B). Although CXCL12 and X4 gp120 interacted with CXCR4, gp120 did not induce F-actin polymerization, even in untreated B cells (Fig. 4C). This suggests that CXCR4/gp120 and CXCR4/CXCL12 interactions have different signaling outcomes in B cells, with only the former totally dependent on lipid raft integrity.
Blockade by gp120 of chemokine-induced signaling
We compared the effects of CXCL12, gp120SF2 (Fig. 5, left panels) and gp120MN (Fig. 5, right panels) on the activation of various effectors downstream from CXCR4. Unlike CXCL12, we found that gp120SF2 and gp120MN did not activate ERK1/2, PLC-3, AKT, and PKC (, I, II, , , and n) but both CXCL12 and gp120SF2 increased the phosphorylation of p38 MAPK. Weaker p38 MAPK activation was observed if CXCR4 was triggered by gp120 rather than CXCL12. IB was not phosphorylated upon CXCR4 triggering by gp120SF2 or gp120MN (data not shown). Consistent with it effects on chemotaxis, gp120SF2 totally prevented the CCL21-induced phosphorylation of ERK1/2, AKT, IB, and PLC3 in B cells (Fig. 6). The activation of p38 MAPK in gp120-treated cells was only slightly weaker after the triggering of CCR7 by CCL21 than in medium-treated cells. Similar results were obtained for stimulation by CCL20 (data not shown).
Treatment with gp120 decreases CD62L expression but increases CD95 expression
We compared the expression of various markers on B cells cultured for 2 h with medium or with 10 nM recombinant gp120 (SF162, BAL, SF2, MN). Whatever the gp120 used, the expression of LFA-1 (CD11a/CD18), ICAM1 (CD54), CD5, CD10, CD80, CD86, CD23, CD38, CD44, CD27, SIgD, TACI, BCMA, and BAFF receptor was not significantly modified (data not shown). In contrast, treatment with X4 and X4/R5 gp120 decreased L-selectin (CD62L) expression by 66 ± 12% (n = 10) (Fig. 7A), increased CD95 expression by a factor of 4.1 ± 1.4 (n = 6) (Fig. 7B), and decreased intracytoplasmic CD95 expression (Fig. 7C). These changes were not detected after treatment with R5 gp120 (data not shown). The modulation of CD62L and CD95 expression induced by gp120 was prevented by adding 1.5 μg/ml 447-52D mAb to gp120 protein (Fig. 7D). The incubation of B cells with 5 μg/ml AMD3100 or 100 ng/ml PTX before incubation with gp120SF2 for 2 h totally prevented changes in surface CD62L (Fig. 7E) and CD95 expression (Fig. 7F). No such effect was observed with 1 μM herbimycin. Inhibitors of PI3K (wortmannin), PLC (U73122), PKC (bisindolylmaleimide I), 3-phosphoinositide-dependent kinase-1 (SH5), MEK1/2 (PD98059), and NF-B nuclear translocation (SN50) did not prevent the decrease in CD62L expression (data not shown), whereas an inhibitor of p38 MAPK (SB203580) totally reversed this effect (Fig. 7E). None of these inhibitors blocked the gp120-induced increase in CD95 expression (data not shown), whereas treatment with brefeldin A did (Fig. 7F). Prior incubation of gp120 with soluble CD4 inhibited the cleavage of CD62L (Fig. 7E), but not the increase in CD95 expression (data not shown). As L-selectin is shed from the cell surface, we incubated B cells with MMP inhibitors before adding gp120. The addition of 500 μM EDTA, 10 μM GM6001, or 10 μM MMP-III inhibitor prevented CD62L shedding (Fig. 7G). No such inhibition was observed with 10 μM MMP-2/9 inhibitor or DMSO. These data are consistent with shedding being induced by MMP-1 and MMP-3 downstream from CD4 and CXCR4/Gi complexes. Medium- and gp120-treated B cells were incubated for 24 h in the presence of 1 μg/ml B-G27 mAb, to determine whether CD95 was functional on gp120-treated B cells (Fig. 7H). The percentage of cells undergoing apoptosis increased from 35.3 ± 2% in medium-treated B cells to 58.7 ± 0.6% in gp120-treated B cells. The addition of 1.5 μg/ml 447-52D mAb to gp120 prevented the induction of apoptosis by gp120.
Discussion
Lymphocyte trafficking is crucial for immune surveillance in vivo, and requires a sequence of critical adhesion events enabling naive and memory cells to recirculate rapidly from the blood to the lymphoid organs (35). Circulating naive and memory B cells express L-selectin, which mediates the first step, adhesion to high endothelial venules, and initiates extravasion toward the lymph nodes and spleen (36). The next step is controlled by G protein-linked chemokine receptors that activate integrins (37). CXCR4, a key receptor for B cell chemotaxis, also acts as a coreceptor for HIV-1 infection. Most mature B cells express CXCR4 strongly and CD4 and CCR5 weakly at their surface (38), but are not infected by HIV-1 in vivo. B cell phenotype and functions are nevertheless rapidly impaired after infection, suggesting that virions or viral proteins released by infected cells directly affect B cell response. As gp120 modulates the chemotaxis of monocytes, dendritic cells, CD4, and CD8 T cells (39, 40, 41), we investigated whether recombinant gp120 affected the B cell chemotaxis essential for a high-quality, efficient humoral response. X4 and X4/R5 gp120 decreased, by 40–50%, not only B cell chemotaxis to CXCL12, but also that to CCL20 and CCL21, whereas neither CCR6 nor CCR7 bound gp120. The percentage of cells undergoing apoptosis was similar for medium- and gp120-treated B cells, indicating that the inhibition of B cell chemotaxis was not due to apoptosis. Consistent with the lower levels of CCR5 than CXCR4 on B cells, R5 gp120 decreased B cell chemotaxis only half as much as X4 or X4/R5 gp120. R5 gp120 inhibited B cell chemotaxis although <10% of B cells migrated in response to CCL3, CCL4, and CCL5, the main ligands of CCR5 (data not shown). Consistent with the weak chemotactic response, the triggering of CCR5 by CCL5 activated PLC3, PI3K/AKT, IB, and PKC, I, II, , , and n less strongly than did CXCL12. R5 gp120 did not activate these effectors. In contrast, the triggering of CCR5 by CCL5 or R5 gp120 led to the activation of p38 MAPK, but not ERK1/2 (data not shown). These results suggest that CCR5 is functional. It is possible that the weak chemotaxis and cell signaling are due to the low level of CCR5 expression on the surface of B cells. Blocking CXCR4 or CCR5 totally reversed the X4, X4/R5, and R5 gp120-induced inhibition of B cell chemotaxis. Similarly, blocking gp120 with the 447-52D mAb totally prevented the gp120-mediated inhibition of B cell chemotaxis, whatever the gp120 tested. CD4 signaling was dispensable for the gp120-induced inhibition of B cell chemotaxis, but prior incubation of gp120 with soluble CD4 enhanced its inhibitory effect. Thus, membrane CD4 molecules, despite being present in small numbers on B cells (42), may induce changes in the conformation of gp120, increasing its affinity for CXCR4 (or CCR5) and its inhibition of B cell chemotaxis. In contrast, the gp120-induced inhibition of monocyte chemotaxis is strictly dependent on CD4 signaling (39). No decrease in CCR6 and CCR7 expression or impairment of CCL20- and CCL21-induced chemokine receptor internalization accompanied the gp120-induced inhibition of B cell chemotaxis. Similarly, gp120 inhibited CXCL13-dependent chemotaxis in the absence of CXCR5 down-regulation or ligand-receptor internalization (data not shown). This contrasts with a previous study showing that the gp120-mediated inhibition of chemotactic response was correlated with weaker chemokine receptor expression in monocytes (39). Toth et al. (43) showed that CXCR4 dimerization was involved in CXCL12- and gp120-induced signaling events. Based on CXCR4-associated fluorescence resonance energy transfer analysis, the authors suggested that gp120 might not fully reproduce signaling events normally induced by CXCL12. The dimerization of CXCR4 and CCR5, leading to JAK2/3 activation, has been shown to play a role in chemotaxis in certain cells (44, 45) but, AG490, a potent inhibitor of JAK2/3, did not affect B cell chemotaxis or its inhibition by gp120 (data not shown). Thus, even if gp120 induces CXCR4 or CCR5 aggregation in B cells, JAK activation is not required to inhibit B cell chemotaxis. Consistent with previous data showing that efficient entry of both X4 and R5 HIV-1 requires intact lipid rafts in leukocytes (33, 46), the gp120-induced inhibition of F-actin polymerization was abolished in cholesterol-depleted B cells, whereas the chemokine-induced response was unaffected. The gp120-mediated inhibition of chemokine-induced F-actin polymerization was stronger than the inhibition of chemotaxis (70% vs 40–50% for X4 gp120). This weaker inhibition of chemotaxis is probably due to the re-expression at the membrane of gp120-free CXCR4 during the assay as no gp120 was added to the upper chamber. Several studies have shown that gp120/CXCR4 and CXCL12/CXCR4 interactions lead to different signaling profiles in T cells and monocytes (47, 48, 49). We showed that X4 or X4/R5 gp120, unlike CXCL12, did not activate PLC3, PI3K/AKT, ERK1/2, IB or PKC, I, II, , , and n in B cells. Nevertheless, B cell stimulation with CXCL12 and X4 or X4/R5 gp120 led to p38 MAPK activation by a PTX-sensitive mechanism (data not shown), suggesting that cell signaling downstream from CXCR4/Gi complexes is still partly functional. Similarly, the inhibition of CCL21- and CCL20-mediated B cell chemotaxis by gp120 correlated with a blockade of the CCL20- and CCL21-mediated activation of PLC3, PKC, PI3K/AKT, IB and ERK1/2, but not of p38 MAPK. The downstream chemokine receptor, p38 MAPK may be phosphorylated via activation of the Sos/Ras/Rac pathway by G subunits, as previously described (50), or by PKC. As CXCR4 triggering by gp120 activated p38 MAPK more weakly than CXCL12 and did not initiate PKC activation, p38 MAPK is probably activated exclusively by the Sos/Ras/Rac pathway after exposure to gp120. As the G protein-coupled receptor kinase and p38 MAPK signaling pathways are functional in gp120-treated cells, we suggest that only signaling pathways using phosphatidylinositol-4,5-bisphosphate (PIP2) as a substrate are blocked downstream from chemokine receptors. PIP2 generates inositol-1,4,5-triphosphate and diacylglycerol when hydrolyzed by PLC, and phosphatidyl-inositol-3,4,5-triphosphate when phosphorylated by PI3K. Blocking PIP2 at the plasma membrane might therefore lead to deficiencies in PI3K and PLC activity. Such a blockade could result from the sequestration to the membrane of acidic phospholipids and PIP2 by dephosphorylated myristoylated alanine-rich C kinase substrate (MARCKS) (51). MARCKS also binds cytoskeleton proteins, including actin, and calmodulin, in a Ca2+-dependent manner whereas PKC-mediated phosphorylation leads to its desorption from the membrane (52, 53, 54). Thus, gp120, unlike CXCL12, probably induces strong interactions with MARCKS, reducing the availability of PIP2 for PLC and PI3K. Consistent with this hypothesis, the destruction of lipid organization in the plasma membrane prevents the effects of gp120 on chemotaxis. As treatment with ionomycin, but not phorbol esters, reversed the gp120-induced inhibition of chemotaxis, it is possible that calmodulin protects MARCKS against PKC phosphorylation. The differential recruitment of MARCKS and activation of G proteins downstream from CXCR4 may result from differential conformational changes induced by gp120 and CXCL12. This is consistent with the work of Staudinger et al. (48), showing that gp120 does not induce conformational changes in CXCR4 similar to those induced by CXCL12. The gp120-induced increase in surface CD95 expression was totally dependent on CXCR4 and Gi heterotrimeric proteins and was induced by the relocalization of cytoplasmic CD95 to the membrane by an unknown mechanism. The gp120-induced increase in surface CD95 expression was totally prevented by anti-gp120 mAb. Such a redistribution of CD95 was recently reported in B lymphoma cell lines (55). Increased surface CD95 expression was correlated with a 23% increase in the CD95-mediated apoptosis of gp120-treated B cells, which was totally abolished by adding the neutralizing anti-gp120 447-52D mAb. Luciani et al. (56) recently showed that the gp120-induced susceptibility of human T cells to CD95-mediated apoptosis involves erzin activation and CD95/ezrin binding. Ezrin belongs to a family of proteins connecting transmembrane proteins to the actin cytoskeleton. It may therefore play a role in the gp120-mediated increase in CD95 expression and CD95-mediated apoptosis in B cells. An increase in susceptibility to CD95 apoptosis may be involved in the depletion of peripheral B cells reported by several groups in HIV-infected patients (3, 9, 10, 57). The gp120-induced decrease in membrane CD62L expression was dependent on CD4, CXCR4 triggering, and on the activation of Gi, p38 MAPK, and its cleavage by MMP. As reported by Preece et al. (58), we showed that collagenase (MMP-1) and stromelysin (MMP-3) were involved in this cleavage, whereas gelatinase A (MMP-2) and B (MMP-9) were not. The MMP-dependent cleavage of L-selectin upon cell activation (59), L-selectin cross-linking (60) or CXCR4 triggering by HIV-1 virions (61), has been described before. It results from the dissociation of calmodulin from the serine-phosphorylated cytoplasmic tail of L-selectin, leading to conformational changes in L-selectin exposing its cleavage site to MMP (62, 63, 64). Our data suggest that serine phoshorylation of CD62L by p38 MAPK, but not by PKC, I, II, , , and is involved in its dissociation from calmodulin. The finding that gp120 decreases CD62L expression and B cell chemotaxis reveals a novel mechanism by which HIV-1 can impair the humoral B cell response.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This work was supported by grants from the Institut National de la Santé et de la Recherche Médicale, Université Paris-Sud, and Sidaction.
2 Address correspondence and reprint requests to Dr. Yolande Richard, Institut National de la Santé et de la Recherche Médicale, Unité 131, Université Paris-Sud, 32 rue des Carnets, 92 140 Clamart, France. E-mail address: yolande.richard@ipsc.u-psud.fr
3 Abbreviations used in this paper: PTX, pertussis toxin; PLC, phospholipase C; PKC, protein kinase C; MFI, mean fluorescence intensity; CD62L, L-selectin; MARCKS, myristoylated alanine-rich C kinase substrate; MCD, methyl--cyclodextrin; MMP, matrix metalloproteinase; PIP2, phosphatidylinositol-4,5-bisphosphate.
Received for publication November 1, 2004. Accepted for publication April 20, 2005.
References
Amadori, A., L. Chieco-Bianchi. 1990. B-cell activation and HIV-1 infection: deeds and misdeeds. Immunol. Today 11: 374-379.
Martinez-Maza, O., E. Crabb, R. T. Mitsuyasu, J. L. Fahey, J. V. Giorgi. 1987. Infection with the human immunodeficiency virus (HIV) is associated with an in vivo increase in B lymphocyte activation and immaturity. J. Immunol. 138: 3720-3724.
amuelsson, A., A. Sonnerborg, N. Heuts, J. Coster, F. Chiodi. 1997. Progressive B cell apoptosis and expression of Fas ligand during human immunodeficiency virus type 1 infection. AIDS Res. Hum. Retroviruses 13: 1031-1038.
F?rster, R., G. Schweigard, S. Johann, T. Emrich, E. Kremmer, C. Nerl, M. Lipp. 1997. Abnormal expression of the B-cell homing chemokine receptor BLR1 during the progression of acquired immunodeficiency syndrome. Blood 90: 520-525.
Beniguel, L., E. Begaud, F. Cognasse, P. Gabrie, C. D. Mbolidi, O. Sabido, M. A. Marovich, C. deFontaine, A. Fresard, F. Lucht, et al 2004. Identification of germinal center B cells in blood from HIV-infected drug-naive individuals in Central Africa. Clin. Dev. Immunol. 11: 23-27.
Racz, P., K. Tenner-Racz, C. Kahl, A. C. Feller, P. Kern, M. Dietrich. 1986. Spectrum of morphologic changes of lymph nodes from patients with AIDS or AIDS-related complexes. Prog. Allergy 37: 81-181.
Pantaleo, G., C. Graziosi, J. F. Demarest, L. Butini, M. Montroni, C. H. Fox, J. M. Orenstein, D. P. Kotler, A. S. Fauci. 1993. HIV infection is active and progressive in lymphoid tissue during the clinically latent stage of disease. Nature 362: 355-358.(Gamal Badr*, Gwenoline Bo)