Although the Src homology 2 domain–containing 5′ inositol phosphatase (SHIP) is a well-known mediator of inhibitory signals after B cell antigen receptor (BCR) coaggregation with the low affinity Fc receptor, it is not known whether SHIP functions to inhibit signals after stimulation through the BCR alone. Here, we show using gene-ablated mice that SHIP is a crucial regulator of BCR-mediated signaling, B cell activation, and B cell development. We demonstrate a critical role for SHIP in termination of phosphatidylinositol 3,4,5-triphosphate (PI[3,4,5]P3) signals that follow BCR aggregation. Consistent with enhanced PI(3,4,5)P3 signaling, we find that splenic B cells from SHIP-deficient mice display enhanced sensitivity to BCR-mediated induction of the activation markers CD86 and CD69. We further demonstrate that SHIP regulates the rate of B cell development in the bone marrow and spleen, as B cell precursors from SHIP-deficient mice progress more rapidly through the immature and transitional developmental stages. Finally, we observe that SHIP-deficient B cells have increased resistance to BCR-mediated cell death. These results demonstrate a central role for SHIP in regulation of BCR signaling and B cell biology, from signal driven development in the bone marrow and spleen, to activation and death in the periphery.
The B cell antigen receptor (BCR) plays a central role in B cell biology, transducing signals that determine cell fate 1. BCR-mediated signals are required not only for initiation and regulation of the immune response, but also for B cell development and survival. Interestingly, a growing body of evidence indicates that the B cell continuously receives signals through the BCR, apparently independent of antigen binding 2. Evidence for the existence and importance of these signals comes from findings that genetic ablation of components of the BCR complex 3,4, or certain BCR signal transduction intermediaries 5,6,7,8,9,10,11, results in developmental arrest and lack of mature B cells in the periphery. Furthermore, induced genetic ablation of the BCR, using conditional knockout mice, results in rapid depletion of the mature B cell pool 2. Thus, it seems likely that developmental progression and survival in both the bone marrow and the periphery require constitutive or “tonic” signaling through the BCR. These data support a hypothesis wherein both antigen-independent and -dependent signals emanating from the BCR provide critical stimuli for the selection, development, activation, and survival of B cells.
Transduction of signals through the BCR occurs via multiple parallel, yet cross-regulating, biochemical pathways. Aggregation of the BCR triggers a rapid increase in tyrosyl phosphorylation of the Igα/β signaling components of the receptor, Src- and Syk-family kinases, and many downstream effector proteins 12,13. One of the most rapidly phosphorylated substrates is CD19, which functions to recruit and activate phosphatidylinositol 3-kinase (PI3-K) 14. PI3-K generates phosphatidylinositol 3,4,5-triphosphate (PI[3,4,5]P3), which activates a wide range of signaling molecules through recruitment to the plasma membrane. PI(3,4,5)P3-mediated activation of the enzymes Bruton's tyrosine kinase (Btk) and phospholipase C (PLC)γ results in generation of inositol polyphosphates (inositol 1,4,5-triphosphate [IP3]) leading to intracellular calcium release and extracellular calcium influx 15,16,17,18,19. PI(3,4,5)P3 also regulates activation of additional signaling pathways including those involving Akt 20,21,22 and mitogen-activated protein (MAP) kinases 23, and regulates activation of transcription factors including nuclear factor (NF)-κB 24,25,26. Furthermore, loss of PI(3,4,5)P3 generation, through genetic ablation of the regulatory subunit of PI3-K, p85α, results in an absence of mature B cells in the periphery 9,10. Thus, signaling cascades that mediate production of PI(3,4,5)P3 are critical not only for active BCR-mediated responses, but also for B cell development and survival.
PI(3,4,5)P3 signals are attenuated upon BCR coaggregation with the low affinity Fc receptor, FcγRIIB1 27,28,29. This inhibition occurs via recruitment of the Src homology 2 domain–containing 5′ inositol phosphatase (SHIP), which hydrolyzes the 5′ phosphate from PI(3,4,5)P3 to produce phosphatidylinositol 3,4-biphosphate (PI[3,4]P2) 30,31,32. This hydrolysis disrupts BCR-mediated activation of Btk and PLCγ, and consequently, the calcium mobilization response 32,33. In both the chicken DT40 B cell lymphoma line and in ex vivo B cells, genetic ablation of SHIP abrogates FcγRIIB1-mediated inhibition of the biochemical processes described above 17,21,34.
Although these results establish SHIP as a crucial negative regulator of B cell signaling during BCR–FcγRIIB1 coaggregation, the role of SHIP during autonomous BCR signaling is not defined. However, recent data have demonstrated increased calcium mobilization and Akt activation upon aggregation of the BCR in SHIP-deficient DT40 cells 21,35. Also important in this regard is the fact that BCR aggregation leads to tyrosine phosphorylation of SHIP and its association with linker molecules including Shc 36, suggesting a possible role for SHIP in regulation of BCR signaling. Recently, it has been reported that mice deficient in SHIP display elevated levels of serum antibody, splenomegaly, and increased mortality 34,37.
To define the role of SHIP in BCR-mediated signal transduction and in B cell biology, we examined ex vivo B cells from SHIP-deficient mice. We found that BCR ligation in SHIP-deficient B cells resulted in levels of PI(3,4,5)P3 that were elevated over fivefold compared with wild-type littermates. Additionally, we found that SHIP regulates the biological outcome of signaling through the BCR. SHIP-deficient B cells were more sensitive to BCR-mediated induction of the activation markers CD86 and CD69. Furthermore, SHIP-deficient B cells displayed accelerated temporal development in the bone marrow, suggesting that PI(3,4,5)P3 signals affect the rate of positive selection of immature B cells. SHIP ablation also resulted in accelerated maturation of transitional B cells in the spleen. Thus, SHIP normally limits the rate of B cell development from the immature stage onward through maturation in the periphery. Finally, we determined that SHIP functions as a positive regulator of BCR-induced cell death. These results demonstrate a central role for SHIP in regulation of BCR signaling and B cell biology, from tonic signal–driven development in the bone marrow and spleen, to activation and death in the periphery.
Materials And Methods
Reagents and Antibodies.
Purified rabbit anti–mouse IgG antibody (H+L) F(ab′)2 fragments were purchased from Zymed, and used for cell stimulation. Antibodies directed against the following molecules were used for flow cytometry: MHC class II (I-Ab/d, D3.137; clone was provided by S. Tonkonogy, North Carolina State University, Raleigh, NC), CD21 (7G6; PharMingen), CD23 (B3B4; PharMingen), CD24 (M1/69; PharMingen), CD45R (anti-B220, RA3-6B2; PharMingen), CD43 (Ly 48; PharMingen), CD69 (H1.2F3; PharMingen), CD86 (GL-1; PharMingen), IgD (JA12.5), IgM (polyclonal; Caltag), and mAb 493 (a gift from A. Rolink, Basel Institute for Immunology, Basel, Switzerland). 7-amino-actinomycin D (7AAD; via probe) was purchased from PharMingen, and propidium iodide and LPS (055:B5) were purchased from Sigma Chemical Co.
Animals and Cells.
All animals used in this study were age-matched 6–10-wk-old SHIP−/− mice and control SHIP+/+ littermates generated as the F1 progeny of SHIP+/− mice 37. Usually mice between 6 and 8 wk old were used to avoid the pathology associated with aged SHIP−/− mice 37. Splenic B cells were prepared as described previously 16. In brief, spleens were excised from mice, cells were dispersed, and RBCs were lysed using Gey's solution. For studies of induction of activation marker expression and cell survival, splenocytes were depleted of T cells by complement lysis using HO13.4 and T24 antibodies, and B cells were further purified by discontinuous Percoll density gradient centrifugation (ρ > 1.07). Purified B cells were cultured with 0–12 μg/ml F(ab′)2 anti-IgG in 20% fetal bovine serum, with 100 IU/ml penicillin and 100 μg/ml streptomycin for 6–24 h. Bone marrow cells were prepared by flushing femurs with IMDM to dislodge cells, followed by gentle deaggregation using a 5-ml syringe. Bone marrow was depleted of RBCs using Gey's solution and washed twice in IMDM.
Cells were washed, resuspended in PBS containing 1% BSA and 0.1% sodium azide, and incubated with optimal concentrations of directly conjugated antibody. Cells were incubated for 30 min at 4°C and washed twice in PBS/BSA/azide. After washing, cells were analyzed by flow cytometry. Cell viability was assessed by exclusion of 7AAD or propidium iodide.
For measurements of intracellular free calcium ([Ca2+]i), 106 cells/ml in IMDM cells were loaded with Indo-AM (Molecular Probes, Inc.), and stimulated with F(ab′)2 anti-IgG antibody. Mean [Ca2+]i was evaluated over time using a flow cytometer (model 50H; Ortho Diagnostic Systems, Inc.) with appended data acquisition system and MultiTime software (Phoenix Flow Systems) as described previously 16.
Measurement of B Cell Survival.
Freshly purified splenic B cells were either left unstimulated, or were stimulated with 1–12 μg/ml of F(ab′)2 anti-IgG antibody. After 24 h of incubation, cell viability was assessed by exclusion of propidium iodide. Experiments were performed in triplicate wells.
Measurement of PI(3,4,5)P3 and PI(3,4)P2 Generation.
Splenic B cells from SHIP−/− or SHIP+/+ control littermates were depleted of RBCs and T cells and further purified by discontinuous Percoll density gradient centrifugation as described above. Cells were cultured for 48 h in 20% fetal bovine serum as described above with the addition of 25 μg/ml LPS (incorporation of 32P label into the cellular ATP pool requires actively metabolizing cells). After 48 h in culture, cells were harvested, washed three times, and incubated for 90 min at 107 cells/ml in low phosphate medium with 0.5 mCi/ml 32P-orthophosphate. 32P-labeled cells were stimulated with F(ab′)2 fragments of anti-IgG for the indicated time, and immediately lysed in 2.4 N HCl/methanol/chloroform (1:0.9:1.4 vol/vol/vol). Lipids were extracted, deacylated with methanol/25% methylamine/n-butanol, and analyzed by HPLC on a SAX ion exchange column (Phenomenex) as described previously 38.
Measurement of IP3 Generation.
IP3 generation was measured using a 3H radioreceptor inhibition assay kit (DuPont-NEN) according to the manufacturer's instructions. Splenic B cells used in this study were first cultured for 48 h as described above.
Sublethal Irradiation and Autoreconstitution.
Age-matched SHIP−/− and SHIP+/+ littermates were treated with 500 rads of whole body irradiation 0–14 d before killing 39. For studies of autoreconstitution, single cell suspensions of spleen or bone marrow were depleted of RBCs, stained, and analyzed by flow cytometry.
SHIP−/− B Cells Display Elevated PI(3,4,5)P3 Levels upon BCR Aggregation.
To define the role of SHIP in modulation of BCR signaling, levels of PI(3,4,5)P3 and its breakdown product, PI(3,4)P2, were measured by HPLC 31,38 after stimulation of splenic B cells derived from SHIP−/− mice and normal littermates. Aggregation of the BCR with F(ab′)2 fragments of anti–mouse Ig resulted in an increase in levels of radiolabeled PI(3,4,5)P3 in both SHIP−/− and SHIP+/+ cells (Fig. 1 A); however, the induced increase in PI(3,4,5)P3 was significantly greater in B cells from SHIP−/− mice. After stimulation of SHIP−/− B cells, PI(3,4,5)P3 levels increased at least fivefold at all measured time points compared with wild-type cells. Particularly striking was the more sustained nature of the PI(3,4,5)P3 signal observed upon BCR aggregation. PI(3,4,5)P3 levels remained highly elevated in B cells from SHIP−/− mice throughout the duration of the experiment. Loss of PI(3,4,5)P3 after stimulation of wild-type cells was primarily attributable to degradation by SHIP, as levels of the SHIP breakdown product, PI(3,4)P2, were significantly elevated in the control cells compared with SHIP−/− cells (Fig. 1 B). These results show that most of the PI(3,4,5)P3 produced upon BCR ligation is rapidly hydrolyzed by SHIP, causing a reduction both in the peak levels of PI(3,4,5)P3 and in the duration of the PI(3,4,5)P3 response. Interestingly, a small but significant amount of PI(3,4,5)P3 degradation and conversion to PI(3,4)P2 was observed in SHIP−/− B cells, suggesting the presence of a second 5′ inositol phosphatase. This may be the recently described SHIP-2 40. From our results, it is apparent that degradation of PI(3,4,5)P3 that follows stimulation through the BCR is primarily mediated through SHIP, and therefore SHIP plays a significant role in regulation of BCR signaling pathways that are PI(3,4,5)P3 dependent.
Although the rate of PI(3,4,5)P3 degradation was increased in the SHIP+/+ cells, the additive counts from PI(3,4)P2 plus PI(3,4,5)P3 (Fig. 1A plus B) were approximately equivalent in the two cell populations, suggesting that PI3-K was equally active. BCR-mediated activation of PI3-K is primarily dependent on phosphorylation of CD19, and CD19 phosphorylation was found to be comparable between the two cell populations (data not shown). In conclusion, analysis of phosphatidylinositol levels reveals that SHIP is a primary regulator of PI(3,4,5)P3 levels. After aggregation of the BCR, SHIP significantly attenuates both the peak and the duration of the PI(3,4,5)P3 response.
One of the effects of BCR aggregation downstream of PI(3,4,5)P3 is activation of PLCγ and consequent mobilization of calcium. To determine whether PI(3,4,5)P3 generation limits BCR-mediated PLCγ activation, the BCR was aggregated and subsequent IP3 generation and calcium mobilization were measured in SHIP−/− and SHIP+/+ cells. SHIP−/− B cells display equivalent IP3 production and slightly increased calcium mobilization after F(ab′)2 anti-Ig stimulation (Fig. 1C and Fig. D). These results indicate that PI(3,4,5)P3 levels achieved during BCR signaling, in either SHIP−/− or wild-type B cells, are sufficient to support PLCγ activation. However, other PI(3,4,5)P3-dependent signaling pathways, and their downstream biologic sequelae, may be affected by excessive PI(3,4,5)P3 accumulation in SHIP−/− B cells.
Reduced Numbers of Immature and/or Transitional B Cells and Increased Numbers of Mature B Cells in the Periphery of SHIP-Deficient Mice.
The biological consequences of SHIP modulation of BCR signaling are poorly defined. It has been reported that mice deficient in SHIP have elevated levels of serum antibody, splenomegaly, and increased mortality 34,37. B cells from SHIP−/− mice are reduced in number by 50% in the bone marrow, and yet surprisingly, increased by ≤50% in the spleen 34,37. As these observations imply a role for SHIP in regulation of B cell development, we compared splenic B cell development in SHIP−/− and wild-type mice. The majority of IgMhiIgDlow cells are transitional (T1) cells 39,41,42, which are recent immigrants from the bone marrow. These cells require constitutive BCR-derived signals to progress to the IgMhiIgDhi (T2) stage 42. Additional BCR-mediated signals then allow the cell to progress to the IgMloIgDhi (M) mature stage 42. In spleens of SHIP−/− mice, there is a two- to threefold reduction in the total number of the most immature (T1) cells, as well as a 1.5-fold reduction in the T2 population (Fig. 2 A, and Table). This conclusion is based on analyses in which contaminating marginal zone cells, distinguished as IgMhiIgDlowCD21hi, were excluded from the T1 population (Fig. 2 B, and Table). Curiously, marginal zone cells are also underrepresented in the spleens of SHIP-deficient mice. Finally, the mature IgMloIgDhi population of cells appears to be overrepresented. It is also of interest that surface Ig (sIg)M levels on mature, SHIP−/− B cells are reduced by two- to fivefold compared with control B cells. The reduction in numbers of transitional B cells in the spleen of SHIP−/− mice, as well as the downregulation of sIgM, suggests a role for SHIP in regulation of maturation, possibly due to overamplification of BCR-generated signals.
To confirm that the unusual splenic B cell phenotype represents increased numbers of mature and reduced numbers of transitional cells, we examined levels of heat-stable antigen (HSA) and B220 on splenic B cells 39. B cells from the SHIP−/− mice showed reduced numbers of HSAhi cells (Fig. 2 C), which further demonstrates that the B cell population has a more mature phenotype. Two additional surface markers expressed on mature cells, CD23 and MHC class II, were also significantly elevated on SHIP−/− cells (data not shown). In conclusion, SHIP deficiency causes developmental abnormalities in B cells leading to a decrease in the number of transitional cells and increased numbers of mature splenic B cells.
Accelerated Development of Transitional B Cells in SHIP-deficient Mice.
Data described above suggest that SHIP regulates B cell development. Specifically, levels of PI(3,4,5)P3, determined by SHIP expression and function, may affect the rate of B cell maturation. As a consequence, B cells may traverse development more rapidly in SHIP−/− mice. To address this possibility, we used the autoreconstitution system described by Cancro and colleagues 39,41 to analyze B cell development. In this system, the B cell compartments of adult mice are depleted by sublethal irradiation, allowing newly generated bone marrow–derived B cells to repopulate the spleen. In normal mice, peripheral lymphoid organs are devoid of lymphocytes 9 d after irradiation, with significant reconstitution appearing at 11–13 d. Virtually all newly formed B cells detected at this time point bear the transitional phenotype, HSAhi, B220lo, IgMhi, and sIgDlo.
B cell development in irradiated mice was evaluated based on quantitation of total numbers of B cells in the spleen, as well as their surface expression of IgM, IgD, CD24 (HSA), and B220. The kinetics of regeneration of B cells in the spleen was compared between SHIP−/− and control littermates (Fig. 3 A). 5 d after irradiation, >99% of the B cells were depleted from the spleens of both SHIP−/− and SHIP+/+ mice, indicating that B cells in both mice were radiation sensitive. Although no production of cells was seen at day 7, by day 10 there were 5–10-fold more B cells in the spleens of the SHIP−/− mice than in wild-type littermates. In SHIP−/− mice, B cells repopulated the spleen 9–11 d after irradiation, while in wild-type littermates, reconstitution occurred after 11–13 d. Through the subsequent period of the experiment, increased numbers of B cells accumulated in the spleens of SHIP−/− mice compared with SHIP+/+ mice.
We were interested in determining whether progression of transitional B cells to the mature stage was also accelerated in irradiated SHIP−/− mice. Typically, this phenotypic transition occurs over a 2–4-d period 39. The kinetics of appearance of the mature phenotype, as determined by decreasing expression of membrane-bound Ig (mIg)M and HSA (Fig. 3 B), revealed changes consistent with accelerated conversion of immature to mature splenic B cells in the SHIP-deficient mice. By day 12 after irradiation, only the SHIP-deficient B cells displayed the lower levels of mIgM and HSA typical of mature B cells. The trend for increased rate of maturation in the SHIP-deficient B cells continued 14 d after irradiation. Furthermore, significant numbers of SHIP−/− B cells rapidly downmodulate mIgM to very low levels. In summary, deletion of SHIP apparently leads to increased entry of immature B cells into the spleen and increased rate of maturation of transitional cells as reflected by downmodulation of IgM and HSA.
SHIP Deficiency Causes Accelerated Development of Immature Cells in the Bone Marrow.
The results presented above indicate that SHIP deficiency leads to prolonged elevation of PI(3,4,5)P3 and altered developmental progression from the transitional to mature stage in B cells, supporting a hypothesis that BCR-derived signals regulate developmental progression in the periphery. As tonic, or constitutive, signals also appear to be required for B cell development in the bone marrow, and because emigration of B cells into the spleen of SHIP−/− mice is accelerated, it seemed plausible that bone marrow development also could be accelerated. To explore this possibility, we examined bone marrow populations during autoreconstitution in SHIP−/− and SHIP+/+ mice. Although total numbers of B220+ cells in the bone marrow were similar between the autoreconstituted B cell populations of SHIP−/− and control littermates, phenotypic analysis revealed differences in the rates of development of early B cell subpopulations. To characterize these populations, we used the 493 mAb, which has been recently reported by Rolink et al. as a marker of early B cell development 43,44. This mAb specifically recognizes pro-, pre-, and immature cells of the B cell lineage, while mature cells are not recognized. Furthermore, its expression is significantly diminished on normal transitional cells (data not shown, and reference 44). Therefore, progression to the transitional/mature phenotype can be followed by decreased mAb 493 binding, and allows a simple way to distinguish between immature and mature B cells in the bone marrow. 493/anti-IgM analysis of normal bone marrow during autoreconstitution reveals that B cells progress through the 493hiIgM− pool (pro/pre), and the 493hiIgM+ (immature) to the 493loIgM+ (transitional) stages. From the data in Fig. 4 (top), mAb 493+ (pro/pre) cells appear with similar kinetics in both the SHIP-deficient and wild-type mice. However, 493hiIgM+ (immature) B cells are increased twofold, and 493loIgM+ (transitional) B cells are increased fourfold in SHIP−/− mice by day 9 (Fig. 4, bottom). In the autoreconstitution model, downmodulation of 493 correlates well with emigration from the bone marrow to the periphery, as repopulation of B cells into the spleen also starts to occur at day 9 in the SHIP-deficient but not the control mice (Fig. 3 A). In control mice, IgM+493lo (transitional) cells appeared ∼1–2 d later. It should be noted that at later time points the percentage of IgM+ cells in the bone marrow of SHIP−/− mice was reduced compared with the SHIP+/+ mice (data not shown), possibly reflecting the accelerated maturation and emigration of IgM+ cells to the spleen. In view of previous studies implicating BCR-mediated signaling in development, these results suggest that increased levels of signaling drive accelerated maturation of the SHIP−/− bone marrow B cells through the immature, transitional, and mature stages of development.
SHIP Attenuates BCR-mediated Activation of CD69 and CD86 Expression.
We next assessed the role of SHIP in modulating active BCR-mediated responses. Aggregation of the BCR has been shown to induce increased expression of CD69 and CD86 on resting mature B cells 45, the latter preparing them to present antigen to T cells. However, the impact of specific signaling pathways on BCR-mediated induction of activation marker expression is not well defined. As BCR-mediated signals are amplified in SHIP-deficient B cells, it seemed plausible that BCR-mediated induction of activation marker expression could be similarly amplified.
As shown in Fig. 5A and Fig. B, ex vivo SHIP−/− B cells expressed slightly elevated levels of CD86 and CD69 and were more sensitive to BCR-mediated induction of these markers than were their wild-type counterparts. In addition, because both molecules were induced at significantly lower Ig concentrations in SHIP−/− cells, SHIP increased signaling thresholds for activation marker expression. These results suggest that both BCR-linked signaling pathways and modulation of gene expression are specifically regulated by SHIP.
Splenic B Cells from SHIP−/− Mice Are Less Susceptible to BCR-mediated Cell Death.
Under certain circumstances, aggregation of the BCR leads to activation-induced cell death via induction of apoptotic signaling pathways 46. As PI(3,4,5)P3 has been shown to be a survival signal in many cell models, SHIP-mediated degradation of PI(3,4,5)P3 could thus promote cell death in response to BCR ligation. To test this possibility, resting splenic B cells were cultured with increasing concentrations of F(ab′)2 anti-Ig, and the percentage of surviving cells was determined 24 h later by propidium iodide exclusion. As shown in Fig. 6, the SHIP−/− cells were more resistant to BCR-mediated cell death than their SHIP+/+ counterparts, especially at elevated concentrations of stimulus. Therefore, SHIP enhances the BCR-mediated apoptotic response of B cells, presumably through elimination of PI(3,4,5)P3.
Results reported here demonstrate an important role for SHIP in regulation of signal transduction through the BCR. Ligation of the BCR alone in SHIP−/− B cells resulted in dramatic and sustained elevations in PI(3,4,5)P3 compared with SHIP+/+ B cells. These results demonstrate that, at least in part, SHIP regulatory function is mediated through its enzymatic activity. Despite the overproduction of PI(3,4,5)P3, SHIP-deficient B cells displayed no detectable increase in IP3 production, and only very modest increases in calcium mobilization upon aggregation of the BCR. Thus, the large increases in PI(3,4,5)P3 do not translate into proportionally large increases in activation of PLCγ. These findings indicate that a distinct signaling component must be limiting for BCR-mediated PLCγ activation. This conclusion is consistent with reports that levels of Btk, an effector necessary for PLCγ activation, has been shown to limit BCR-mediated calcium mobilization in the A20 B cell line 19,32. However, these observations are discordant with studies showing that SHIP ablation in the DT40 chicken lymphoma line significantly enhances BCR-mediated calcium mobilization responses 35. It is possible that in the DT40 system, Btk is expressed at higher levels, making PI(3,4,5)P3 the limiting element in the PLC activation cascade.
Mutations in components of the BCR signaling complex can severely influence B cell population dynamics and function. SHIP ablation is clearly associated with an altered biological phenotype. Characterization of the B cells from SHIP-deficient mice revealed a 50% reduction in the numbers of pre- and immature B cells in the bone marrow, and yet increased numbers of B cells in the spleen 34,37. We found that immature and transitional B cells are significantly reduced in the splenic B cell compartment. Two possible mechanisms could account for this phenotype, and be a consequence of the amplified BCR signaling seen in SHIP−/− B cells: (a) an unusually large proportion of immature B cells may be deleted, with survivors expanding to fill available space in the periphery; and (b) the rate of B cell development may be accelerated, depleting the pre-, immature, and transitional B cell pools. While the former possibility was not formally excluded, we found that SHIP-deficient bone marrow cells were fully competent to reconstitute the mature splenic B cell compartment. SHIP−/− and SHIP+/+ littermates differed primarily with respect to the rapidity with which they traversed the immature and transitional stages. In our studies of autoreconstitution, we found that in SHIP−/− mice B cells more rapidly repopulated the splenic compartment. Although these studies support a hypothesis in which deletion of SHIP results in accelerated maturation, the increased production of B cells could additionally be a result of enhanced survival of transitional B cells as they traverse into the mature B cell pool.
Development from the transitional to the mature B cell stage is an active process that is reportedly dependent on the strength of the constitutive BCR-mediated signal. For example, loss of constitutive signal in CD45-deficient mice 42,47 results in a block at the late transitional stage, and loss of mature cells in the periphery. Conversely, loss of CD22 48,49, a negative regulator of B cell signaling, results in increased numbers of mature splenic B cells. SHIP deficiency appears to amplify the constitutive BCR signal, resulting in a rapid conversion from the transitional to the mature stage, based on downmodulation of sIgM and HSA. Therefore, modifications of the BCR complex that enhance constitutive signaling function also appear to enhance B cell development.
SHIP-deficient B cells also emigrated from the bone marrow at an accelerated rate, thus partially accounting for the reduction in numbers of immature cells in SHIP-deficient bone marrow. It is possible that B cell development and survival could be promoted through increased accumulation of PI(3,4,5)P3 in SHIP-deficient B cells. Although other signaling pathways may be necessary for signaling of development, levels of PI(3,4,5)P3 appear to limit progression.
Previous reports have proposed that elevations in tyrosine phosphorylation, as a consequence of loss of tyrosine phosphatases, can also change the signaling thresholds that effect development and selection 50. Combined, these studies suggest that levels of PI3-K activity and tyrosine kinase activity are central in regulation of signaling and development mediated through the BCR. However, tyrosine kinase activity is linked to PI3-K activity 51. For example, efficient PI(3,4,5)P3 generation, during BCR signaling, requires tyrosine phosphorylation of CD19, which subsequently recruits and activates PI3-K. Consistent with findings reported here, B cells derived from CD19-deficient mice are both deficient in PI3-K activation 15 and unusually immature, expressing the elevated levels of sIgM characteristic of the transitional phenotype 52. Conversely, overexpression of CD19 renders B cells hyperresponsive to transmembrane signals. CD19 transgenic mice have increased levels of autoantibodies, and the B cell population is shifted toward the mature sIgMloIgDhi phenotype 53. Deficiency in the regulatory subunit of PI3-K, p85α, also leads to impaired B cell development. Thus, evidence from several models supports a central role for PI(3,4,5)P3 in driving development.
Although the signaling events that trigger various stages of development are not well defined, at least three signaling molecules downstream of SHIP and PI3-K have recently been implicated as potential regulators of B cell development. Dominant active ras, as well as raf and MAP kinases 42,54,55, can provide signals required for developmental maturation in B cells. Intriguingly, the activation of ras/raf and MAP kinase–mediated signaling events has been linked positively to PI(3,4,5)P3 and negatively to SHIP. For example, the PI3-K inhibitor, wortmannin, inhibits ras/raf-mediated downstream activation of the MAP kinase pathway 56, as does SHIP expression 34. Conversely, generation of PI(3,4,5)P3 can activate downstream signaling pathways, including the induction of gene expression through NF-κB 24. Interestingly, NF-κB is essential for B cell development, as mice deficient in both NF-κB p50 and p52 57 demonstrate a loss of peripheral B cells. Thus, there are multiple parallel downstream pathways linked to PI(3,4,5)P3 that are likely candidates for inhibition by SHIP during B cell development. It is important to note that, in addition to its enzymatic activity, SHIP may also function as a linker protein. In some systems, SHIP has been shown to bind with Shc and/or p62 dok 57a, both of which are implicated in ras activation 58.
We also observe that some active BCR-mediated responses are negatively regulated through SHIP, as SHIP-deficient B cells exhibit increased sensitivity to F(ab′)2 anti-Ig stimulation of CD86 and CD69 expression. As upregulation of CD86 is required for efficient T cell–B cell collaboration, this provides an additional mechanism through which SHIP could negatively regulate immune responses and prevent aberrant stimulation. Although previous studies using pharmacological agents indicate that CD86 and CD69 expression can be induced by calcium-activated pathways 45, physiologic activation of these responses appears to be limited by SHIP. SHIP may further mediate its inhibitory effects through a block in ras signaling, as previous studies in T cells have demonstrated inhibition of CD69 induction upon overexpression of dominant negative ras 59.
Finally, we show the positive role of SHIP in regulation of BCR-mediated cell death. Previous studies have shown that in vitro ligation of the BCR results in apoptotic death of B cells 46. Although SHIP negatively influences CD86 and CD69 expression, it is a positive mediator of the apoptotic response 60, suggesting that degradation of PI(3,4,5)P3 promotes cell death. The mechanism of PI(3,4,5)P3-mediated protection is unknown; however, Akt is activated through PI(3,4,5)P3 and provides an antiapoptotic signal in some tissues. Akt is believed to promote cell survival by phosphorylation of Bad 61 and caspase-9 62, and promote cell cycle progression 63. Thus, SHIP may contribute to the apoptotic signal through its degradation of PI(3,4,5)P3 and resulting failed activation of Akt 21,22.
Previous studies have demonstrated a critical role for SHIP in mediating inhibitory signaling through FcγRIIB1 in B cells and mast cells 64. In this report, we demonstrate an active role for SHIP as a critical regulator of B cell development and antigen receptor–mediated activation. Our results indicate that B cell development in both the bone marrow and spleen is regulated by SHIP. We hypothesize that SHIP deficiency enhances both tonic- and antigen-mediated BCR signals driving B cell development, maturation, and activation.
We thank Bill Townend and Shirley Sobus for assistance with flow cytometry. We also thank Antonius Rolink for the gift of the 493 mAb.
This work was supported by grants from the US Public Health Service to J.C. Cambier. J.C. Cambier is an Ida and Cecil Green Endowed Professor of Cell Biology. R.K. Humphries was supported by the National Cancer Institute of Canada (funds from the Canadian Cancer Society and the Terry Fox Run) and the Medical Research Council of Canada.
Abbreviations used in this paper: 7AAD, 7-amino-actinomycin D; BCR, B cell antigen receptor; Btk, Bruton's tyrosine kinase; [Ca2+], intracellular free calcium; HSA, heat-stable antigen; IP3, inositol 1,4,5-triphosphate; MAP, mitogen-activated protein; mIg, membrane-bound Ig; NF, nuclear factor; PI3-K, phosphatidylinositol 3-kinase; PI(3,4)P2, phosphatidylinositol 3,4-biphosphate; PI(3,4,5)P3, phosphatidylinositol 3,4,5-triphosphate; PLC, phospholipase C; SHIP, Src homology 2 domain–containing 5′ inositol phosphatase; sIg, surface Ig.