Cutting Edge: B-1 Cells Are Deficient in Lck: Defective B Cell Receptor Signal Transduction in B-1 Cells Occurs in the Absence of Elevated L
http://www.100md.com
免疫学杂志 2005年第13期
Abstract
B-1 cells constitute a unique B cell subset that is primarily responsible for producing nonimmune Ig. This natural Ig acts as a principal line of defense against infection. A key feature of B-1 cells is the failure of BCR-triggered signal transduction. Recently, defective BCR signaling in B-1 cells has been attributed to elevated expression of the canonical T cell src kinase, Lck. In the present study, we re-examined Lck expression in normal B-1 cells. We found that B-1 cells expressed less Lck at both the protein and RNA levels than did B-2 cells. The same B-1 cells manifested defective BCR-mediated induction of IKK phosphorylation, IB degradation, and intracellular Ca2+ mobilization. Thus, the failure of BCR signaling in B-1 cells does not relate to subset-specific elevation of Lck.
Introduction
B-1 cells constitute a unique B cell subset that is primarily responsible for producing natural (nonimmune) Ig (reviewed in Refs. 1, 2, 3, 4). This natural Ig plays a critical and nonredundant role in host immune defense against infection (5, 6, 7, 8). B-1 cells are characterized by a number of unusual phenotypic, ontogenetic, and functional characteristics which have contributed to an ongoing controversy regarding their developmental origin. A key feature of B-1 cells is deficient BCR signaling. Whereas BCR cross-linking produces Ca2+ mobilization, NF-B activation, and proliferation in conventional B (B-2) cells, none of these things takes place in anti-Ig-stimulated B-1 cells (9, 10, 11, 12, 13, 14). This indolence may result from prior BCR triggering, which has been proposed to play a role in B-1 cell development, and so could reflect Ag experience and may represent a form of tolerance (15, 16, 17, 18).
The mechanism underlying defective BCR signaling in B-1 cells remains uncertain. Because BCR-triggered Ca2+ mobilization is affected, but PMA-induced NF-B activation is not (10), it has been suggested that BCR coupling to phospholipase C is impaired in B-1 cells, and direct measurements of BCR-induced phospholipase C activity suggest that this is so (11, 17). This in turn is thought to result from the inhibitory effect of CD5, particularly the tyrosine phosphorylated form of CD5, due to its association with SHP-1 phosphatase (12, 13, 19). Recently, the origin of phosphorylated CD5 and impaired BCR signaling has been reported to lie in an increased level of the canonical T cell src kinase Lck. Increased levels of Lck have been reported in human CLL cells, human B-1 cells, and murine peritoneal B-1 cells (but not murine splenic B-1 cells) (20, 21, 22). In the murine studies, peritoneal B-1 cells expressed much more Lck than did splenic B-2 cells (21, 22). In view of these reports, it is now well accepted that B-1 cells express an unusually high level of Lck in relation to B-2 cells. However, in earlier work examining multiple src kinases in unpurified peritoneal and splenic B cells, we did not observe increased Lck in the B-1 cell-rich peritoneal population (11). We have now re-examined the Lck content of B-1 cells using highly pure B cell populations and improved techniques for isolation and identification.
Materials and Methods
Animals
Male BALB/cByJ and C57BL/6 mice at 8–14 wk of age were obtained from The Jackson Laboratory. Mice were cared for and handled in accordance with National Institutes of Health and institutional guidelines.
B cell purification and culture
Unseparated cells were obtained by peritoneal washout and splenic disruption and were purified in one of three ways: 1) Cells were stained with immunofluorescent Abs directed against B220 and CD5 and were subjected to FACS at 4°C using a Mo-Flo flow cytometer (DakoCytomation) to yield purified peritoneal B-1 (B220+CD5+) and splenic B-2 (B220+CD5–) cells (sort-purified B-1 and B-2 cells) as previously described, including the use of an anti-CD8 "dump" channel for B cell purification (23). B-1 and B-2 cells were >97% pure (i.e., B220+CD5+ and B220+CD5–, respectively). T cells (B220–CD5+) were also isolated by cell sorting. For experiments in which CD5 was immunoprecipitated, B-1 cells were sort-purified using anti-B220 and anti-Mac-1 Abs. 2) Cells were depleted of T cells by treatment with anti-Thy 1.2 plus complement and of macrophages by two rounds of plate adherence to yield B-1 and B-2 cell-enriched populations (adherence/Thy1.2-purified B cells) as previously described (24). 3) Cells were depleted of macrophages by plate adherence, were depleted of non-B cells by biotinylated Thy 1.2-, CD4-, and CD8-specific Abs plus streptavidin magnetic beads, and, for peritoneal washouts, were positively selected for CD5 (magnetic bead-purified B cells) as described by Cambier and colleagues (22). All isolated B cell populations contained <0.5% T cells (B220–CD3+). Sort-purified and magnetic bead-purified B-1 cells contained 0.71% ± 0.39 (mean ± SEM) macrophages (B220–CD14+), and adherence/Thy1.2-purified B-1 cells contained 4.8% ± 1.3 macrophages. Sort-purified and magnetic bead-purified B-2 cells contained <0.5% macrophages and adherence/Thy1.2-purified B-2 cells contained 1.0% ± 0.27 macrophages. B cells were cultured in RPMI 1640 medium as previously described.
Protein expression
B cells were extracted with radioimmunoprecipitation assay (RIPA)3 buffer (25), after which proteins were separated by SDS-PAGE, transferred to polyvinylidene fluoride membranes, and immunoblotted as previously described (26). In some cases B cells were extracted with CHAPS buffer (27). CD5 was immunoprecipitated from RIPA buffer-extracted B cell lysates with Sepharose beads coupled to anti-CD5 Ab and immunoblotted as described above with a different CD5-specific Ab and with a phospho-CD5-specific Ab.
Intracellular Ca2+
Peritoneal or splenic cells were loaded with Indo-1AM (Molecular Probes) at 1 μM for 30 min at 37°C, then washed and stained at 4°C with B220- and CD5-specific fluorescent Abs. Cells were washed again and then resuspended at 2 x 106 cells/ml in dyeless RPMI 1640 with 5% FCS. Intracellular Ca2+ was evaluated by measuring fluorescence at 405 and 485 nm after excitation at 355 nm with a Mo-Flo flow cytometer (DakoCytomation). The various lymphocyte populations were evaluated through gating and data analysis performed using FlowJo software (Tree Star).
Reagents
Fluorescent-labeled anti-B220, anti-CD5, anti-Mac-1, anti-CD3, anti-CD14, and anti-CD8 Abs were obtained from BD Pharmingen. Two different monoclonal Lck-specific Abs were obtained from BD Biosciences (clone 28) and Santa Cruz Biotechnology (clone 3A5). Polyclonal Ab specific for the (Tyr505) phosphorylated form of Lck and anti-phospho-IKK (Ser180/181, respectively) were obtained from Cell Signaling Technology. Anti-IB and two different monoclonal anti-CD5 Abs were obtained from Santa Cruz Biotechnology. Monoclonal anti-phosphotyrosine Ab was obtained from Cell Signaling Technology. Anti--actin Ab was obtained from Sigma-Aldrich. F(ab')2 of anti-IgM were obtained from Jackson ImmunoResearch Laboratories.
Results and Discussion
B-1 cells express less Lck than B-2 cells
We determined the Lck content of sort-purified BALB/c lymphocyte populations by Western blotting. Immunofluorescent staining and FACS were conducted at 4°C to avoid any room temperature or at 37°C manipulations during which protein expression might change. Isolated B-1, B-2, and T cells were then immediately extracted with RIPA buffer, after which proteins were size separated by SDS-PAGE and immunoblotted with a monoclonal Lck-specific Ab. As expected, T cells expressed substantial amounts of Lck; in direct contrast and B cells expressed very little (Fig. 1A). Among B cells, no Lck was detected in B-1 cell extracts, whereas B-2 cell extracts contained very small amounts. Although in most experiments 15 μg protein was analyzed, no Lck was detected in B-1 cell extracts even after scaling up to 35 μg protein (data not shown). Normalized to -actin as a control, B-2 cells expressed less than 1/20 the amount of Lck expressed by T cells, and B-1 cells expressed less than that. A similar population distribution was obtained when blots were probed with a phospho-Lck (pLck)-specific Ab: T cells expressed substantial amounts of pLck, whereas B cells expressed very little. Thus, sort-purified B-1 cells do not express any more Lck or pLck than sort-purified B-2 cells.
We obtained lymphocyte populations from C57BL/6 mice to rule out the possibility that strain differences in Lck expression might account for previous reports of elevated Lck in B-1 cells. Again, sort-purified B-1 cells failed to express detectable levels of Lck and pLck, in contrast to sort-purified T cells which expressed much of both (Fig. 1A). In particular, as before, B-1 cell extracts contained less Lck than did extracts obtained from B-2 cells. We separately extracted BALB/c lymphocyte populations with CHAPS to rule out the possibility that differences in solubilization conditions might account for our inability to detect Lck in B-1 cells. We obtained the same results with CHAPS extraction as we did with RIPA buffer, and again B-1 cells failed to express Lck and pLck (Fig. 1B). We further tested all of these samples, involving different mouse strains and extraction conditions, with a different monoclonal Lck-specific Ab and obtained the same results; again B-1 cells contained no detectable Lck, whereas B-2 cells contained low levels and T cells expressed abundant amounts (data not shown).
Previous reports indicating that B-1 cells express Lck used plate adherence to remove macrophages (20, 21, 22). Because this represented a potential explanation for the disparity in B-1 cell Lck expression between this report and others, we compared adherence/Thy 1.2-purified and sort-purified B cell populations from BALB/c mice, but found little difference in the results obtained. Adherence/Thy 1.2-purified B-1-enriched cells contained no detectable Lck or pLck just like sort-purified B-1 cells, whereas both adherence/Thy 1.2-purified B-2-enriched cells and sort-purified B-2 cells expressed small amounts of Lck (Fig. 1C). As an assay control, we verified in the same experiment that sort-purified T cells contained substantial amounts of Lck and pLck. We further purified B-1 and B-2 cells by plate adherence combined with negative and positive magnetic bead separation, as outlined in Materials and Methods (22), yet still found that B-1 cells expressed less Lck than B-2 cells (Fig. 1D). Inasmuch as CD5 phosphorylation has been reported to be induced by BCR triggering and to correlate with up-regulated Lck expression (22), we stimulated B-1 and B-2 cells with anti-Ig for 24 h to determine whether Lck or pLck levels respond to B cell activation (Fig. 1C). No increase in Lck, or pLck, was identified in either B cell population as a result of BCR engagement, consistent with the view that B-1 cells do not express or up-regulate Lck beyond the level contained in B-2 cells.
We further examined Lck gene expression of sort-purified BALB/c lymphocyte populations by quantitative PCR and normalized results to 2-microglobulin. In these, as in other, experiments, all steps in lymphocyte purification were conducted at 4°C to avoid any room temperature or at 37°C manipulations during which gene expression might change. As expected, T cells expressed substantial amounts of Lck mRNA; in direct contrast, B-1 and B-2 cells expressed very little, with B-1 cells expressing less than the level expressed by B-2 cells (Fig. 1E). These results are consistent with the results obtained by Western blotting, which together indicate that B-1 cells fail to express increased amounts of Lck at both the mRNA and protein levels in relation to B-2 cells.
BCR signaling is defective in B-1 cells
We analyzed several events downstream of BCR engagement to determine whether B-1 cells that lack Lck also fail to signal after BCR triggering. In a previous report, we showed that BCR signaling for NF-B activation is impaired in B-1 cells (10). In the present study, we sought to extend that work by examining IKK phosphorylation and IB degradation, two key steps in the signaling pathway from BCR to NF-B. To do this, B-1 and B-2 cells were sort-purified and then stimulated with anti-Ig; at various times after stimulation, aliquots of B cells were extracted with RIPA buffer as described above. Western blotting showed a marked increase in IKK phosphorylation after anti-Ig stimulation of B-2 cells that was completely lacking in similarly treated B-1 cells (Fig. 2A). Along the same lines, Western blotting showed clear-cut IB degradation within 1 h of BCR triggering in B-2 cells that also did not occur in B-1 cells. Thus, BCR signaling is blocked at an early point in B-1 cells such that IKK phosphorylation and IB degradation do not occur.
We also examined BCR-induced intracellular Ca2+. Peritoneal and splenic cells were immunofluorescently stained and loaded with the Ca2+-sensitive dye Indo-1AM, after which BCR-triggered Ca2+ mobilization of selected B cell subsets was measured by flow cytometry. Basal Ca2+ levels were similar in B-1 and B-2 cells, as previously described in several, but not all, studies (12, 19, 22, 28). In keeping with published results (13, 14, 22), we found that both B-1 cells and B-2 cells experienced a rapid increase in intracellular Ca2+ following anti-Ig stimulation, but that the increase in B-1 cells was much less than the increase in B-2 cells. Thus, in B-1 cells, BCR signaling is impaired at a point before Ca2+ elevation. Taken together, the results on IKK, IB, and Ca2+ demonstrate that the same B-1 cells that lack Lck are deficient in BCR-triggered downstream signaling events.
As noted above, it has been suggested that the phosphorylated form of CD5 is responsible for blocked BCR signaling, and that high levels of Lck are responsible for constitutive CD5 phosphorylation, in naive B-1 cells (13, 22). Because B-1 cells express little, if any, Lck, it was of interest to determine whether B-1 cells contain phosphorylated CD5. We immunoprecipitated CD5 from sort-purified B-1, B-2, and T cells and, following separation of proteins by SDS-PAGE, Western blotted for CD5 (with a different monoclonal CD5-specific Ab). B-1 cells contained immunoprecipitable CD5, as expected, although this was much less than that of T cells; B-2 cells contained no detectable immunoprecipitable CD5 (Fig. 3). This pattern matches that obtained by directly Western blotting sort-purified lymphocyte extracts after SDS-PAGE. A substantial amount of the B-1 cell CD5 was recognized by a phosphotyrosine-specific Ab, whereas no phosphotyrosine was detected in the much more abundant CD5 precipitated from T cells. Thus, abundant Lck (and pLck) failed to produce constitutive CD5 phosphorylation in T cells, whereas phosphorylated CD5 was found in B-1 cells despite undetectable levels of Lck and pLck. This lack of correlation among Lck, pLck, and pCD5 strongly suggests that Lck is not responsible for constitutive phosphorylation of CD5 in B-1 cells.
Our results on defective BCR signaling in B-1 cells are consistent with previous studies from this and other laboratories (9, 10, 11, 12, 14, 17, 19, 22, 29). However, our finding that B-1 cells lack Lck contrasts with several recent reports (20, 21, 22), although it is consistent with earlier work in which increased expression of Lck was not detected in peritoneal as compared with splenic B cells by immune complex kinase assay (11). The reason for the different results obtained in these different studies remains unclear, although in this respect it should be noted that a relatively small degree of T cell contamination would significantly raise the level of Lck detected in isolated B-1 cells. Regardless of how the disparate results on B-1 cell Lck content were obtained, the rapid, rigorous, and temperature-controlled purification procedure used here, along with the use of two different Lck-specific Ab reagents for Western blotting supplemented with quantitative PCR to assess gene expression, strongly suggests that the relative absence of Lck in B-1 cells most accurately reflects the normal situation in vivo, as does our finding of relatively deficient Lck expression in B-1 cells purified in the manner described by Cambier and colleagues. However, previous work indicating that BCR signaling is restored in Lck-deficient B-1 cells raises the possibility that a threshold level of Lck is sufficient to modulate signal transduction, although it should be pointed out that Baldari and colleagues (20, 21) reported diminished, not enhanced, BCR-induced activation of Raf, MEK, and ERK in Lck-deficient B-1 cells, and that secondary, non-cell autonomous effects have not been ruled out in Lck-deficient mice (21, 22).
The finding that abnormal BCR signaling in B-1 cells is not related to aberrantly increased expression of Lck raises the question of just what might be responsible for the generally acknowledged signaling defect in this B cell population. Although we found that B-1 cells lack Lck, we also found that B-1 cells express phosphorylated CD5, and so the latter may, as reported by others (12, 13, 19), play a key role in blocking BCR signaling in B-1 cells, presumably after phosphorylation by another src kinase. Recently, however, Bondada and colleagues reported that BCR signaling is defective in peritoneal B-1b cells that lack CD5 (30), indicating that abnormal BCR signaling takes place in the absence of CD5 in this B-1 cell subset. Thus, CD5 may be responsible for abnormal BCR signaling in some, but not all, B-1 cells. Alternatively, CD5 phosphorylation may only be a sign of, rather than a motive force for, increased kinase activity or diminished phosphatase activity that is ultimately responsible for defective BCR signaling. Although it seems clear that CD5 phosphorylation in B-1 cells is not driven by unusually high levels of Lck, the kinase responsible remains unidentified.
Acknowledgments
We thank Sean Gurdak for assistance with B cell purification.
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 Public Health Service Grants AI29690 and AI60896 awarded by the National Institutes of Health. R.F. is the recipient of a fellowship from the Spanish Liver Society and Agencia Valenciana de Ciencia y Tecnologia, Generalitat Valenciana.
2 Address correspondence and reprint requests to Dr. Thomas L. Rothstein, Immunobiology Unit, Evans Biomedical Research Center, Room 437, Boston Medical Center, 650 Albany Street, Boston, MA 02118. E-mail address: tr@bumc.bu.edu
3 Abbreviations used in this paper: RIPA, radioimmunoprecipitation assay; pLck, phosphor-Lck.
Received for publication February 24, 2005. Accepted for publication April 19, 2005.
References
Herzenberg, L. A.. 2000. B-1 cells: the lineage question revisited. Immunol. Rev. 175: 9-22.
Hardy, R. R., K. Hayakawa. 2001. B cell development pathways. Annu. Rev. Immunol. 19: 595-621.
Berland, R., H. H. Wortis. 2002. Origins and functions of B-1 cells with notes on the role of CD5. Annu. Rev. Immunol. 20: 253-300.
Rothstein, T. L.. 2002. Cutting edge: two B-1 or not to be one. J. Immunol. 168: 4257-4261.
Briles, D. E., M. Nahm, K. Schroer, J. Davie, P. Baker, J. Kearney, R. Barletta. 1981. Antiphosphocholine antibodies found in normal mouse serum are protective against intravenous infection with type 3 streptococcus pneumoniae. J. Exp. Med. 153: 694-705.
Boes, M., A. P. Prodeus, T. Schmidt, M. C. Carroll, J. Chen. 1998. A critical role of natural immunoglobulin M in immediate defense against systemic bacterial infection. J. Exp. Med. 188: 2381-2386.
Ochsenbein, A. F., T. Fehr, C. Lutz, M. Suter, F. Brombacher, H. Hengartner, R. M. Zinkernagel. 1999. Control of early viral and bacterial distribution and disease by natural antibodies. Science 286: 2156-2159.(Rubén Francés*, Joseph R.)
B-1 cells constitute a unique B cell subset that is primarily responsible for producing nonimmune Ig. This natural Ig acts as a principal line of defense against infection. A key feature of B-1 cells is the failure of BCR-triggered signal transduction. Recently, defective BCR signaling in B-1 cells has been attributed to elevated expression of the canonical T cell src kinase, Lck. In the present study, we re-examined Lck expression in normal B-1 cells. We found that B-1 cells expressed less Lck at both the protein and RNA levels than did B-2 cells. The same B-1 cells manifested defective BCR-mediated induction of IKK phosphorylation, IB degradation, and intracellular Ca2+ mobilization. Thus, the failure of BCR signaling in B-1 cells does not relate to subset-specific elevation of Lck.
Introduction
B-1 cells constitute a unique B cell subset that is primarily responsible for producing natural (nonimmune) Ig (reviewed in Refs. 1, 2, 3, 4). This natural Ig plays a critical and nonredundant role in host immune defense against infection (5, 6, 7, 8). B-1 cells are characterized by a number of unusual phenotypic, ontogenetic, and functional characteristics which have contributed to an ongoing controversy regarding their developmental origin. A key feature of B-1 cells is deficient BCR signaling. Whereas BCR cross-linking produces Ca2+ mobilization, NF-B activation, and proliferation in conventional B (B-2) cells, none of these things takes place in anti-Ig-stimulated B-1 cells (9, 10, 11, 12, 13, 14). This indolence may result from prior BCR triggering, which has been proposed to play a role in B-1 cell development, and so could reflect Ag experience and may represent a form of tolerance (15, 16, 17, 18).
The mechanism underlying defective BCR signaling in B-1 cells remains uncertain. Because BCR-triggered Ca2+ mobilization is affected, but PMA-induced NF-B activation is not (10), it has been suggested that BCR coupling to phospholipase C is impaired in B-1 cells, and direct measurements of BCR-induced phospholipase C activity suggest that this is so (11, 17). This in turn is thought to result from the inhibitory effect of CD5, particularly the tyrosine phosphorylated form of CD5, due to its association with SHP-1 phosphatase (12, 13, 19). Recently, the origin of phosphorylated CD5 and impaired BCR signaling has been reported to lie in an increased level of the canonical T cell src kinase Lck. Increased levels of Lck have been reported in human CLL cells, human B-1 cells, and murine peritoneal B-1 cells (but not murine splenic B-1 cells) (20, 21, 22). In the murine studies, peritoneal B-1 cells expressed much more Lck than did splenic B-2 cells (21, 22). In view of these reports, it is now well accepted that B-1 cells express an unusually high level of Lck in relation to B-2 cells. However, in earlier work examining multiple src kinases in unpurified peritoneal and splenic B cells, we did not observe increased Lck in the B-1 cell-rich peritoneal population (11). We have now re-examined the Lck content of B-1 cells using highly pure B cell populations and improved techniques for isolation and identification.
Materials and Methods
Animals
Male BALB/cByJ and C57BL/6 mice at 8–14 wk of age were obtained from The Jackson Laboratory. Mice were cared for and handled in accordance with National Institutes of Health and institutional guidelines.
B cell purification and culture
Unseparated cells were obtained by peritoneal washout and splenic disruption and were purified in one of three ways: 1) Cells were stained with immunofluorescent Abs directed against B220 and CD5 and were subjected to FACS at 4°C using a Mo-Flo flow cytometer (DakoCytomation) to yield purified peritoneal B-1 (B220+CD5+) and splenic B-2 (B220+CD5–) cells (sort-purified B-1 and B-2 cells) as previously described, including the use of an anti-CD8 "dump" channel for B cell purification (23). B-1 and B-2 cells were >97% pure (i.e., B220+CD5+ and B220+CD5–, respectively). T cells (B220–CD5+) were also isolated by cell sorting. For experiments in which CD5 was immunoprecipitated, B-1 cells were sort-purified using anti-B220 and anti-Mac-1 Abs. 2) Cells were depleted of T cells by treatment with anti-Thy 1.2 plus complement and of macrophages by two rounds of plate adherence to yield B-1 and B-2 cell-enriched populations (adherence/Thy1.2-purified B cells) as previously described (24). 3) Cells were depleted of macrophages by plate adherence, were depleted of non-B cells by biotinylated Thy 1.2-, CD4-, and CD8-specific Abs plus streptavidin magnetic beads, and, for peritoneal washouts, were positively selected for CD5 (magnetic bead-purified B cells) as described by Cambier and colleagues (22). All isolated B cell populations contained <0.5% T cells (B220–CD3+). Sort-purified and magnetic bead-purified B-1 cells contained 0.71% ± 0.39 (mean ± SEM) macrophages (B220–CD14+), and adherence/Thy1.2-purified B-1 cells contained 4.8% ± 1.3 macrophages. Sort-purified and magnetic bead-purified B-2 cells contained <0.5% macrophages and adherence/Thy1.2-purified B-2 cells contained 1.0% ± 0.27 macrophages. B cells were cultured in RPMI 1640 medium as previously described.
Protein expression
B cells were extracted with radioimmunoprecipitation assay (RIPA)3 buffer (25), after which proteins were separated by SDS-PAGE, transferred to polyvinylidene fluoride membranes, and immunoblotted as previously described (26). In some cases B cells were extracted with CHAPS buffer (27). CD5 was immunoprecipitated from RIPA buffer-extracted B cell lysates with Sepharose beads coupled to anti-CD5 Ab and immunoblotted as described above with a different CD5-specific Ab and with a phospho-CD5-specific Ab.
Intracellular Ca2+
Peritoneal or splenic cells were loaded with Indo-1AM (Molecular Probes) at 1 μM for 30 min at 37°C, then washed and stained at 4°C with B220- and CD5-specific fluorescent Abs. Cells were washed again and then resuspended at 2 x 106 cells/ml in dyeless RPMI 1640 with 5% FCS. Intracellular Ca2+ was evaluated by measuring fluorescence at 405 and 485 nm after excitation at 355 nm with a Mo-Flo flow cytometer (DakoCytomation). The various lymphocyte populations were evaluated through gating and data analysis performed using FlowJo software (Tree Star).
Reagents
Fluorescent-labeled anti-B220, anti-CD5, anti-Mac-1, anti-CD3, anti-CD14, and anti-CD8 Abs were obtained from BD Pharmingen. Two different monoclonal Lck-specific Abs were obtained from BD Biosciences (clone 28) and Santa Cruz Biotechnology (clone 3A5). Polyclonal Ab specific for the (Tyr505) phosphorylated form of Lck and anti-phospho-IKK (Ser180/181, respectively) were obtained from Cell Signaling Technology. Anti-IB and two different monoclonal anti-CD5 Abs were obtained from Santa Cruz Biotechnology. Monoclonal anti-phosphotyrosine Ab was obtained from Cell Signaling Technology. Anti--actin Ab was obtained from Sigma-Aldrich. F(ab')2 of anti-IgM were obtained from Jackson ImmunoResearch Laboratories.
Results and Discussion
B-1 cells express less Lck than B-2 cells
We determined the Lck content of sort-purified BALB/c lymphocyte populations by Western blotting. Immunofluorescent staining and FACS were conducted at 4°C to avoid any room temperature or at 37°C manipulations during which protein expression might change. Isolated B-1, B-2, and T cells were then immediately extracted with RIPA buffer, after which proteins were size separated by SDS-PAGE and immunoblotted with a monoclonal Lck-specific Ab. As expected, T cells expressed substantial amounts of Lck; in direct contrast and B cells expressed very little (Fig. 1A). Among B cells, no Lck was detected in B-1 cell extracts, whereas B-2 cell extracts contained very small amounts. Although in most experiments 15 μg protein was analyzed, no Lck was detected in B-1 cell extracts even after scaling up to 35 μg protein (data not shown). Normalized to -actin as a control, B-2 cells expressed less than 1/20 the amount of Lck expressed by T cells, and B-1 cells expressed less than that. A similar population distribution was obtained when blots were probed with a phospho-Lck (pLck)-specific Ab: T cells expressed substantial amounts of pLck, whereas B cells expressed very little. Thus, sort-purified B-1 cells do not express any more Lck or pLck than sort-purified B-2 cells.
We obtained lymphocyte populations from C57BL/6 mice to rule out the possibility that strain differences in Lck expression might account for previous reports of elevated Lck in B-1 cells. Again, sort-purified B-1 cells failed to express detectable levels of Lck and pLck, in contrast to sort-purified T cells which expressed much of both (Fig. 1A). In particular, as before, B-1 cell extracts contained less Lck than did extracts obtained from B-2 cells. We separately extracted BALB/c lymphocyte populations with CHAPS to rule out the possibility that differences in solubilization conditions might account for our inability to detect Lck in B-1 cells. We obtained the same results with CHAPS extraction as we did with RIPA buffer, and again B-1 cells failed to express Lck and pLck (Fig. 1B). We further tested all of these samples, involving different mouse strains and extraction conditions, with a different monoclonal Lck-specific Ab and obtained the same results; again B-1 cells contained no detectable Lck, whereas B-2 cells contained low levels and T cells expressed abundant amounts (data not shown).
Previous reports indicating that B-1 cells express Lck used plate adherence to remove macrophages (20, 21, 22). Because this represented a potential explanation for the disparity in B-1 cell Lck expression between this report and others, we compared adherence/Thy 1.2-purified and sort-purified B cell populations from BALB/c mice, but found little difference in the results obtained. Adherence/Thy 1.2-purified B-1-enriched cells contained no detectable Lck or pLck just like sort-purified B-1 cells, whereas both adherence/Thy 1.2-purified B-2-enriched cells and sort-purified B-2 cells expressed small amounts of Lck (Fig. 1C). As an assay control, we verified in the same experiment that sort-purified T cells contained substantial amounts of Lck and pLck. We further purified B-1 and B-2 cells by plate adherence combined with negative and positive magnetic bead separation, as outlined in Materials and Methods (22), yet still found that B-1 cells expressed less Lck than B-2 cells (Fig. 1D). Inasmuch as CD5 phosphorylation has been reported to be induced by BCR triggering and to correlate with up-regulated Lck expression (22), we stimulated B-1 and B-2 cells with anti-Ig for 24 h to determine whether Lck or pLck levels respond to B cell activation (Fig. 1C). No increase in Lck, or pLck, was identified in either B cell population as a result of BCR engagement, consistent with the view that B-1 cells do not express or up-regulate Lck beyond the level contained in B-2 cells.
We further examined Lck gene expression of sort-purified BALB/c lymphocyte populations by quantitative PCR and normalized results to 2-microglobulin. In these, as in other, experiments, all steps in lymphocyte purification were conducted at 4°C to avoid any room temperature or at 37°C manipulations during which gene expression might change. As expected, T cells expressed substantial amounts of Lck mRNA; in direct contrast, B-1 and B-2 cells expressed very little, with B-1 cells expressing less than the level expressed by B-2 cells (Fig. 1E). These results are consistent with the results obtained by Western blotting, which together indicate that B-1 cells fail to express increased amounts of Lck at both the mRNA and protein levels in relation to B-2 cells.
BCR signaling is defective in B-1 cells
We analyzed several events downstream of BCR engagement to determine whether B-1 cells that lack Lck also fail to signal after BCR triggering. In a previous report, we showed that BCR signaling for NF-B activation is impaired in B-1 cells (10). In the present study, we sought to extend that work by examining IKK phosphorylation and IB degradation, two key steps in the signaling pathway from BCR to NF-B. To do this, B-1 and B-2 cells were sort-purified and then stimulated with anti-Ig; at various times after stimulation, aliquots of B cells were extracted with RIPA buffer as described above. Western blotting showed a marked increase in IKK phosphorylation after anti-Ig stimulation of B-2 cells that was completely lacking in similarly treated B-1 cells (Fig. 2A). Along the same lines, Western blotting showed clear-cut IB degradation within 1 h of BCR triggering in B-2 cells that also did not occur in B-1 cells. Thus, BCR signaling is blocked at an early point in B-1 cells such that IKK phosphorylation and IB degradation do not occur.
We also examined BCR-induced intracellular Ca2+. Peritoneal and splenic cells were immunofluorescently stained and loaded with the Ca2+-sensitive dye Indo-1AM, after which BCR-triggered Ca2+ mobilization of selected B cell subsets was measured by flow cytometry. Basal Ca2+ levels were similar in B-1 and B-2 cells, as previously described in several, but not all, studies (12, 19, 22, 28). In keeping with published results (13, 14, 22), we found that both B-1 cells and B-2 cells experienced a rapid increase in intracellular Ca2+ following anti-Ig stimulation, but that the increase in B-1 cells was much less than the increase in B-2 cells. Thus, in B-1 cells, BCR signaling is impaired at a point before Ca2+ elevation. Taken together, the results on IKK, IB, and Ca2+ demonstrate that the same B-1 cells that lack Lck are deficient in BCR-triggered downstream signaling events.
As noted above, it has been suggested that the phosphorylated form of CD5 is responsible for blocked BCR signaling, and that high levels of Lck are responsible for constitutive CD5 phosphorylation, in naive B-1 cells (13, 22). Because B-1 cells express little, if any, Lck, it was of interest to determine whether B-1 cells contain phosphorylated CD5. We immunoprecipitated CD5 from sort-purified B-1, B-2, and T cells and, following separation of proteins by SDS-PAGE, Western blotted for CD5 (with a different monoclonal CD5-specific Ab). B-1 cells contained immunoprecipitable CD5, as expected, although this was much less than that of T cells; B-2 cells contained no detectable immunoprecipitable CD5 (Fig. 3). This pattern matches that obtained by directly Western blotting sort-purified lymphocyte extracts after SDS-PAGE. A substantial amount of the B-1 cell CD5 was recognized by a phosphotyrosine-specific Ab, whereas no phosphotyrosine was detected in the much more abundant CD5 precipitated from T cells. Thus, abundant Lck (and pLck) failed to produce constitutive CD5 phosphorylation in T cells, whereas phosphorylated CD5 was found in B-1 cells despite undetectable levels of Lck and pLck. This lack of correlation among Lck, pLck, and pCD5 strongly suggests that Lck is not responsible for constitutive phosphorylation of CD5 in B-1 cells.
Our results on defective BCR signaling in B-1 cells are consistent with previous studies from this and other laboratories (9, 10, 11, 12, 14, 17, 19, 22, 29). However, our finding that B-1 cells lack Lck contrasts with several recent reports (20, 21, 22), although it is consistent with earlier work in which increased expression of Lck was not detected in peritoneal as compared with splenic B cells by immune complex kinase assay (11). The reason for the different results obtained in these different studies remains unclear, although in this respect it should be noted that a relatively small degree of T cell contamination would significantly raise the level of Lck detected in isolated B-1 cells. Regardless of how the disparate results on B-1 cell Lck content were obtained, the rapid, rigorous, and temperature-controlled purification procedure used here, along with the use of two different Lck-specific Ab reagents for Western blotting supplemented with quantitative PCR to assess gene expression, strongly suggests that the relative absence of Lck in B-1 cells most accurately reflects the normal situation in vivo, as does our finding of relatively deficient Lck expression in B-1 cells purified in the manner described by Cambier and colleagues. However, previous work indicating that BCR signaling is restored in Lck-deficient B-1 cells raises the possibility that a threshold level of Lck is sufficient to modulate signal transduction, although it should be pointed out that Baldari and colleagues (20, 21) reported diminished, not enhanced, BCR-induced activation of Raf, MEK, and ERK in Lck-deficient B-1 cells, and that secondary, non-cell autonomous effects have not been ruled out in Lck-deficient mice (21, 22).
The finding that abnormal BCR signaling in B-1 cells is not related to aberrantly increased expression of Lck raises the question of just what might be responsible for the generally acknowledged signaling defect in this B cell population. Although we found that B-1 cells lack Lck, we also found that B-1 cells express phosphorylated CD5, and so the latter may, as reported by others (12, 13, 19), play a key role in blocking BCR signaling in B-1 cells, presumably after phosphorylation by another src kinase. Recently, however, Bondada and colleagues reported that BCR signaling is defective in peritoneal B-1b cells that lack CD5 (30), indicating that abnormal BCR signaling takes place in the absence of CD5 in this B-1 cell subset. Thus, CD5 may be responsible for abnormal BCR signaling in some, but not all, B-1 cells. Alternatively, CD5 phosphorylation may only be a sign of, rather than a motive force for, increased kinase activity or diminished phosphatase activity that is ultimately responsible for defective BCR signaling. Although it seems clear that CD5 phosphorylation in B-1 cells is not driven by unusually high levels of Lck, the kinase responsible remains unidentified.
Acknowledgments
We thank Sean Gurdak for assistance with B cell purification.
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 Public Health Service Grants AI29690 and AI60896 awarded by the National Institutes of Health. R.F. is the recipient of a fellowship from the Spanish Liver Society and Agencia Valenciana de Ciencia y Tecnologia, Generalitat Valenciana.
2 Address correspondence and reprint requests to Dr. Thomas L. Rothstein, Immunobiology Unit, Evans Biomedical Research Center, Room 437, Boston Medical Center, 650 Albany Street, Boston, MA 02118. E-mail address: tr@bumc.bu.edu
3 Abbreviations used in this paper: RIPA, radioimmunoprecipitation assay; pLck, phosphor-Lck.
Received for publication February 24, 2005. Accepted for publication April 19, 2005.
References
Herzenberg, L. A.. 2000. B-1 cells: the lineage question revisited. Immunol. Rev. 175: 9-22.
Hardy, R. R., K. Hayakawa. 2001. B cell development pathways. Annu. Rev. Immunol. 19: 595-621.
Berland, R., H. H. Wortis. 2002. Origins and functions of B-1 cells with notes on the role of CD5. Annu. Rev. Immunol. 20: 253-300.
Rothstein, T. L.. 2002. Cutting edge: two B-1 or not to be one. J. Immunol. 168: 4257-4261.
Briles, D. E., M. Nahm, K. Schroer, J. Davie, P. Baker, J. Kearney, R. Barletta. 1981. Antiphosphocholine antibodies found in normal mouse serum are protective against intravenous infection with type 3 streptococcus pneumoniae. J. Exp. Med. 153: 694-705.
Boes, M., A. P. Prodeus, T. Schmidt, M. C. Carroll, J. Chen. 1998. A critical role of natural immunoglobulin M in immediate defense against systemic bacterial infection. J. Exp. Med. 188: 2381-2386.
Ochsenbein, A. F., T. Fehr, C. Lutz, M. Suter, F. Brombacher, H. Hengartner, R. M. Zinkernagel. 1999. Control of early viral and bacterial distribution and disease by natural antibodies. Science 286: 2156-2159.(Rubén Francés*, Joseph R.)