Membrane Cholesterol Content Accounts for Developmental Differences in Surface B Cell Receptor Compartmentalization and Signaling*

Recent studies argue for an important role for cholesterol in maintaining plasma membrane heterogeneity and influencing a variety of cellular processes, including signaling, adhesion, and permeability. Here, we document that tolerance-sensitive transitional immature B cells maintain significantly lower membrane unesterified cholesterol levels than mature-stage splenic B cells. In addition, the relatively low level of cholesterol in transitional immature B cells impairs compartmentalization of their B cell receptor (BCR) into cholesterol-enriched domains following BCR aggregation and re-duces their ability to sustain certain aspects of BCR signaling as compared with mature B cells. These studies establish an unexpected difference in the lipid composition of peripheral transitional immature and mature B cells and point to a determining role for development-associated differences in cholesterol content for the differential responses of these B cells to BCR engagement. The spatial relationships between proteins in the inner and outer leaflet the plasma roles in a of diverse biological initiation entry, Organization pro-tein-protein Isolation and Definition of Mature and Transitional Immature B Cell Subsets— Transitional immature B cells were isolated from spleens of 6-to 10-week-old sublethally irradiated, auto-reconstituting BALB/c mice 14 days post sub-lethal irradiation at 500 rads, as described previously (18). The same protocol was applied to non-irradiated mice to isolate a population enriched in mature B cells. For these studies, the isolated populations ranged between 94 and 98% B cells, as assessed by B220 expression. Non-sorted mature populations generally contain 10–15% transitional immature B cells (19). The non-sorted transitional imma- ture populations used in these studies were devoid of detectable mature B cells, as assessed by the absence of sIgM pos , B220 pos , and AA4.1 lo cells. Fluorescence-activated Cell Sorting and Filipin Staining— To determine the amount of membrane-associated cholesterol in each cell type, 1 (cid:1) 10 6 cells were fixed in 120 (cid:5) l of 4% paraformaldehyde at room temperature for 12 min, washed three times with PBS, and incubated with filipin (1:100 of 25 mg/ml stock, Sigma) fo r 1 h at 4 °C.Samples of cells were also simultaneously analyzed for expression of B220, CD23 (both BD Pharmingen), AA4.1, and level of filipin binding and fluorescence. The cells were subject to flow cytometric analysis, and the filipin binding levels of mature (B220 pos , AA4.1 lo , and CD23 pos ), T1 (B220 pos , AA4.1 hi , and CD23 neg ), and T2 (B220 pos , AA4.1 hi , and CD23 pos ) were determined. Gas Chromatography— Mature and transitional immature B cells were purified as above. To each of the samples of purified B cells, a known of cholesteryl methyl ether was added as an internal standard for the gas-liquid chromatographic assay, and then lipid inverted micro-scope receptor (Zymed Laboratories Inc.), and detected by ECL. Northern Blot Analysis— RNA was extracted using TRIzol (Invitro-gen) according to the manufacturer’s instructions. Total RNA was sub- jected to Northern blot analysis on a 1% agarose/formaldehyde gel. The RNA was transferred to Hybond-N (Amersham Biosciences) by capil- lary action. Northern blots were probed with actin and c- myc inserts radioactively labeled to high specific activity with a Prime-It II kit and hybridized in QuikHyb, both according to the manufacturer’s instruc-tions (Stratagene). for we analyzed the partitioning of the BCR within the membrane in transitional immature and mature B cells.

Recent studies argue for an important role for cholesterol in maintaining plasma membrane heterogeneity and influencing a variety of cellular processes, including signaling, adhesion, and permeability. Here, we document that tolerance-sensitive transitional immature B cells maintain significantly lower membrane unesterified cholesterol levels than mature-stage splenic B cells. In addition, the relatively low level of cholesterol in transitional immature B cells impairs compartmentalization of their B cell receptor (BCR) into cholesterolenriched domains following BCR aggregation and reduces their ability to sustain certain aspects of BCR signaling as compared with mature B cells. These studies establish an unexpected difference in the lipid composition of peripheral transitional immature and mature B cells and point to a determining role for development-associated differences in cholesterol content for the differential responses of these B cells to BCR engagement.
The spatial relationships between proteins in the inner and outer leaflet of the plasma membrane play important roles in a number of diverse biological processes, including the initiation and transduction of receptor generated signals, protein sorting, viral entry, and membrane internalization (1)(2)(3)(4). Organization of the plasma membrane into discrete compartments determined by lipid content provides additional regulation for protein-protein and protein-lipid interactions. The primary classes of lipids that reside in the plasma membrane are glycerophospholipids, sphingolipids, and membrane-active sterols (5). Glycerophospholipids contain two acyl chains, one of which is often unsaturated. In contrast, sphingolipids contain two acyl chains, which are both larger and saturated. Membrane-active sterols, mainly represented by cholesterol in mammals, are defined by a flat fused-ring system. Cholesterol has a greater affinity for the saturated acyl chains of sphingolipids and its incorporation into membranes provides rigidity via its fourring structure. The higher concentration of cholesterol within sphingolipids also leads to strong van der Waals interactions between lipid acyl chains, which decreases membrane permeability and stabilizes the lipid packing. These properties result in a distinct membrane compartment characterized by a higher ordering of the lipid membrane, as opposed to disordered, unsaturated, acyl-containing membrane domains (6,7).
Many recent studies have focused on these higher ordered sphingolipid/cholesterol-enriched membrane microdomains, termed lipid rafts. Rafts are operationally defined as regions of the membrane that are resistant to some non-ionic detergents and are enriched in the ganglioside GM1 1 (8). It has been suggested that raft domains play important roles in signaling, because several receptors, including the B cell receptor (BCR) (2), T cell receptor (9), Fc⑀ R (10), and insulin receptor (11), are observed to redistribute from the detergent-soluble, lipid-disordered to the GM1-enriched, detergent-insoluble compartment upon ligand binding.
B lymphocyte responses to similar stimuli differ dramatically at discrete stages of maturation and in different microenvironments. Notably, bone marrow and peripheral immature B cells are not activated by signals generated through the BCR but rather undergo apoptosis (deletion) or receptor modification (editing) (12). Transitional immature B cells are a population of peripheral immature B cells that have recently left the bone marrow and are a critical link to immunocompetent mature B cells (13)(14)(15). Our laboratory has previously shown that, upon BCR cross-linking, transitional immature B cells undergo apoptosis, whereas mature B cells enter and progress through cell cycle, up-regulate proteins necessary to activate T cells, and manifest other processes associated with activation (12,16,17). Importantly, the ligand-induced redistribution of the BCR is developmentally regulated, such that, although it preferentially associates with the lipid raft marker GM1 in mature cells after BCR aggregation, this inducible redistribution does not occur in transitional immature B cells (2,14). The relationship between these responses and plasma membrane composition has not been explored. We speculated that this phenomenon might be related to differential membrane composition. Because of the unique function of cholesterol to facilitate membrane microdomain formation within the plasma membrane, we have compared cholesterol levels between transitional immature and mature B cells and have determined the association of membrane cholesterol levels to BCR proximal signaling and lipid raft compartmentalization.

Isolation and Definition of Mature and Transitional Immature B Cell
Subsets-Transitional immature B cells were isolated from spleens of 6to 10-week-old sublethally irradiated, auto-reconstituting BALB/c mice 14 days post sub-lethal irradiation at 500 rads, as described previously (18). The same protocol was applied to non-irradiated mice to isolate a population enriched in mature B cells. For these studies, the isolated populations ranged between 94 and 98% B cells, as assessed by B220 expression. Non-sorted mature populations generally contain 10 -15% transitional immature B cells (19). The non-sorted transitional immature populations used in these studies were devoid of detectable mature B cells, as assessed by the absence of sIgM pos , B220 pos , and AA4.1 lo cells.
Fluorescence-activated Cell Sorting and Filipin Staining-To determine the amount of membrane-associated cholesterol in each cell type, 1 ϫ 10 6 cells were fixed in 120 l of 4% paraformaldehyde at room temperature for 12 min, washed three times with PBS, and incubated with filipin (1:100 of 25 mg/ml stock, Sigma) for 1 h at 4°C. Samples of cells were also simultaneously analyzed for expression of B220, CD23 (both BD Pharmingen), AA4.1, and level of filipin binding and fluorescence. The cells were subject to flow cytometric analysis, and the filipin binding levels of mature (B220 pos , AA4.1 lo , and CD23 pos ), T1 (B220 pos , AA4.1 hi , and CD23 neg ), and T2 (B220 pos , AA4.1 hi , and CD23 pos ) were determined.
Gas Chromatography-Mature and transitional immature B cells were purified as described above. To each of the samples of purified B cells, a known amount of cholesteryl methyl ether was added as an internal standard for the gas-liquid chromatographic assay, and then each group was subject to a Bligh-Dyer lipid extraction (20). 4 ϫ 10 7 B cells of each type were suspended in 0.8 ml of PBS to which 3 ml of CHCl 3 /MeOH (1:2) was added, and the samples were vigorously vortexed. Then, 1 ml of CHCl 3 alone was added, and the samples were vortexed, followed by the addition of 1 ml of H 2 O and vortexing. The samples were then spun at 3000 rpm to separate the phases, and the lower CHCl 3 phase was retained, split into aliquots, and dried under nitrogen. The amount of free and total cholesterol in each sample was then determined by gas chromatography as previously described (21).
Addition of Cholesterol to Cells-Transitional immature B cells (1 ϫ 10 7 cells/ml) were incubated in RPMI containing either cholesterol alone (300 g/ml) or cholesterol (150 or 300 g/ml) plus 5 mM M␤CD. These samples were incubated at 37°C for 30 -120 min as indicated in individual experiments. The cells were then incubated with filipin as previously described to determine their relative membrane cholesterol content.
Immunofluorescence and Microscopy-Purified B cells (2 ϫ 10 6 in 200 L of PBS) were placed in the wells of a 96 well plate and incubated at 37°C. Cells were stimulated with a goat F(abЈ) 2 anti-BCR antibody (IgG-and liquid chromatography-directed, Jackson Laboratories) for 5 min, and the reaction was stopped by addition of cold PBS/0.1% bovine serum albumin/0.02% azide. Cells were fixed in 4% paraformaldehyde/ 0.1% glutaraldehyde, followed by washing and resuspension in 0.5 mg/ml sodium borohydride to stop fixation. Cells were then incubated in staining buffer containing both a Cy3-conjugated secondary antibody directed to the anti-BCR primary antibody and filipin at 4°C. The cells were adhered to glass slides by centrifugation at 800 rpm using a Cytospin centrifuge (Shandon) and then mounted with Prolong (Molecular Probes) anti-fade agent. BCR-expressing, filipin-binding cells were visualized using a Zeiss Axiovert 200M inverted epifluorescence microscope equipped with a Sensicam QE high performance camera. Images were captured and analyzed using Slidebook image analysis software (Intelligent Imaging Innovations). The No Neighbors deconvolution module of the software was used to remove out-of-focus light.
Cells were scored according to the following criteria: 1) Cells were only counted if they were not in contact with other cells. 2) The BCR was scored as aggregated when the BCR polarized on less than half the circumference of the cell. 3) Co-polarization of the BCR aggregates and cholesterol were scored when peak cholesterol intensity overlapped with a BCR aggregate. 4) At least 35 cells were scored for each condition per experiment. For representative cells in Figs. 1 and 3, relative filipin distribution was determined by calculating the sum pixel intensities at the poles of each cell. For cells that have been stimulated with anti-BCR, the halves were determined using the BCR cap as a reference point. For unstimulated cells that lacked this point of reference, the cell was divided arbitrarily.
Detergent-soluble and Detergent-insoluble Domain Isolation for Silver Staining-Cells (20 -30 ϫ 10 6 cells) were pelleted and then lysed in 200 l of 1% Triton X-100/25 mM MES/150 mM NaCl/1 mM phenylmethylsulfonyl fluoride/1 mM vanadate/protease inhibitor mixture (Roche Applied Science) on ice for 15 min. Lysates were homogenized in a 1-ml Dounce homogenizer for 10 strokes. Lysates were then centrifuged at 500 ϫ g for 7 min at 4°C. The post nuclear supernatant was centrifuged at 100,000 ϫ g for 60 min at 4°C. The resultant supernatant was removed and labeled the detergent-soluble fraction (DS). The 100,000 ϫ g pellet was resuspended in 200 l of 50 mM ␤-octylglucopyranoside (Sigma)/20 mM Tris/500 mM NaCl/protease inhibitor mixture (Roche Applied Science), kept on ice for 15 min, and then centrifuged at 10,000 ϫ g for 12 min at 4°C. The resultant supernatant was removed and labeled as the Triton X-100 detergent-insoluble fraction (DI). Equivalent amounts for each cell fraction were subjected to SDS-PAGE, and the proteins were detected with Bio-Rad Silver Stain Plus (Bio-Rad) per the manufacturer's instructions.
Sucrose Gradient Centrifugation and Immunoblot Analysis-CH27, untreated WEHI-231, and M␤CD/cholesterol-treated WEHI-231 (5-10 ϫ 10 7 cells/ml in PBS) cells were incubated at 37°C. Cells were either unstimulated or stimulated with anti-BCR for 2 min at 37°C. Cells were pelleted and lysed in 1% Triton X-100/25 mM MES/150 mM NaCl/1 mM phenylmethylsulfonyl fluoride/1 mM vanadate/protease inhibitor mixture (Roche Applied Science) on ice for 15 min. Equal amounts of protein in 1 ml of lysis buffer were combined with 1 ml of 80% sucrose in MES-buffered saline and overlaid with 2 ml of 30% sucrose in MES-buffered saline, followed by 1 ml of 5% sucrose in MES-buffered saline. Sucrose gradients were subject to ultracentrifugation (Beckman) at 43,000 rpm for 18 h at 4°C. Gradients were divided into 12 fractions, and the fractions were separated by 9% SDS-PAGE. After transfer onto nitrocellulose, immunoblots were probed with anti-IgM (Jackson ImmunoResearch) and anti-transferrin receptor (Zymed Laboratories Inc.), and detected by ECL.
Northern Blot Analysis-RNA was extracted using TRIzol (Invitrogen) according to the manufacturer's instructions. Total RNA was subjected to Northern blot analysis on a 1% agarose/formaldehyde gel. The RNA was transferred to Hybond-N (Amersham Biosciences) by capillary action. Northern blots were probed with actin and c-myc inserts radioactively labeled to high specific activity with a Prime-It II kit and hybridized in QuikHyb, both according to the manufacturer's instructions (Stratagene).

BCR Aggregates Co-polarize with Cholesterol in Mature B
Cells-We have previously shown that anti-BCR-induced BCR aggregates in mature B cells (B220 pos , CD23 pos , and AA4.1 lo ) co-polarize with the lipid raft marker GM1 to a significantly greater extent than do those in transitional immature B cells (B220 pos and AA4.1 hi ) (14). Because cholesterol is a key component in lipid rafts, we wanted to determine if anti-BCRinduced BCR aggregates localized to areas relatively enriched in unesterified cholesterol in transitional immature and mature B cells. We used filipin as a marker for unesterified (free) cholesterol (22)(23)(24). In both unstimulated mature and transitional immature B cells, cholesterol and the BCR were evenly distributed around the cells. Anti-BCR-induced BCR aggregates in mature B cells co-polarized with cholesterol in 48 Ϯ 4% of observed cells (Fig. 1A, top panel, and Table I). Anti-BCRinduced BCR aggregates in transitional immature B cells copolarized with cholesterol in 16 Ϯ 3% of observed cells (Fig. 1B, bottom panel, and Table I). Therefore, in agreement with the observed co-polarization of BCR aggregates with GM1, BCR aggregates co-polarized with cholesterol to a greater extent in mature B cells compared with transitional immature B cells.
Cholesterol Content of Transitional Immature B Cell Subsets Is Equivalent and Different from Mature Splenic B Cells-We next wanted to determine if the relative inability of BCR aggregates to co-polarize with cholesterol in transitional immature B cells was due to lower levels of cholesterol in their plasma membrane. In the studies to follow, the fluorescence emission of filipin bound to cholesterol was utilized to detect free cholesterol in fixed cells while allowing for the simultaneous measurement of other markers by flow cytometry, thereby permitting subpopulation analysis in heterogeneous populations.
Cholesterol levels were first determined by filipin binding to transitional immature and mature B cells ( Fig. 2A). We observed that the mature population binds ϳ2-3 times the amount of filipin as the transitional immature population. Notably, filipin binding to mature B cells was bimodal. Subsequent gating and phenotypic analysis of the lower intensity peak indicated that this peak represents the transitional immature B cells that are found normally in the adult mouse spleen (10 -15% of B220 pos cells) (Fig. 2, B and C). Populations designated mature in all subsequent studies will contain this small, defined population of transitional immature B cells. Furthermore, transitional immature B cells can be divided into sub-populations based upon CD23 expression (T1, CD23 neg , and AA4.1 hi ; and T2, CD23 pos , and AA4.1 hi ). The T1 and T2 sub-populations exhibit the same level of relative filipin bind-ing (Fig. 2C). Lastly, cytometric light scatter analysis was performed on all of the populations to compare their cell size.
The forward scatter of all the populations was indistinguishable (Fig. 2D). Therefore, differences in the cholesterol content between these populations are not likely attributable to differences in plasma membrane area. Of note, the total BCR expression (IgM and IgD combined) are equivalent for mature and transitional immature and mature B cells (data not shown). Importantly, in all our studies we are stimulating through BCR using an antibody that reacts with the light chain component common to both isoforms so that all surface BCR complexes are engaged. The result depicted in Fig. 2A indicates that transitional immature B cells contain less unesterified cholesterol than do mature B cells. To corroborate this finding more quantitatively and to address the issue that the cholesterol content of the plasma membrane in these cells may not be at equilibrium, we determined the relative levels of unesterified (free) and total cholesterol by gas-liquid chromatography from both populations (Table II). Esterified cholesterol does not have the ability to bind to the plasma membrane, and it is typically found in cells such as hepatocytes that store cholesterol (25). Lymphocytes do not contain appreciable amounts of esterified cholesterol (26 -29). As in previous studies by others (22,24), we observed a close agreement in the cholesterol levels determined by biochemical analysis and filipin binding. Regardless of whether the results were normalized to cell number or protein level, the mature B cells contained 2-3 times the amount of cholesterol relative to that found in transitional immature B cells. In addition, this analysis confirmed that total and free cholesterol pools were identical, indicating that all the cholesterol in these cells is unesterified. Therefore, the cholesterol levels quantitated by filipin binding levels represent plasma membrane or plasma membrane precursor pools and strongly support the interpretation that the cholesterol content of the transitional immature B cell plasma membranes is reproducibly lower than in mature B cells.

Manipulation of the Cholesterol Content of Transitional Immature B Cells Facilitates Anti-BCR Induced Co-localization of Cholesterol-enriched Compartments with Aggregated BCR
Complexes-M␤CD is a carbohydrate that binds selectively and reversibly to cholesterol. Most studies have utilized M␤CD to deplete plasma membrane cholesterol. However, it can also be used to transfer cholesterol from a high cholesterol-containing medium into the relatively low cholesterol-containing plasma membrane. Fig. 3A shows the effect of incubating transitional immature B cells with varying levels of extracellular free cholesterol at 37°C in the presence of M␤CD at various times of incubation at each condition. The results depicted in Fig. 3A indicate a dose-dependent increase in membrane-associated cholesterol, as determined by filipin binding. At 5 mM M␤CD with 300 g/ml cholesterol, transitional immature cells exhibited similar filipin binding to mature cells. This addition of cholesterol was dependent upon the presence of M␤CD, because transitional cells treated with cholesterol alone showed no increase in filipin binding (second panel from left). To determine if the filipin binding analyzed by fluorescence-activated cell sorting was at the limit of detection, we added cholesterol to mature B cells in the presence of M␤CD. The results showed that the cholesterol level of mature cells could be detectably increased (data not shown). Therefore, the identical filipin binding of M␤CD/cholesterol-treated transitional immature cells and untreated mature cells was not due to the limit of detection of filipin binding. Furthermore, we noted no detectable change in the level of BCR expression on transitional immature B cells after cholesterol addition (data not shown).  One characteristic of cholesterol-enriched domains is that they are resistant to solubilization in 1% Triton X-100 at 4°C. We next determined if the addition of M␤CD/cholesterol to transitional immature B cells affected the distribution of proteins in 1% Triton X-100-soluble and -insoluble domains. Primary mature, transitional immature, and transitional immature B cells treated with M␤CD/cholesterol were lysed in 1% Triton X-100. The cells were then centrifuged at 100,000 ϫ g for 1 h at 4°C, and the resultant supernatant was labeled DS. The 100,000 ϫ g pellet was resuspended in octylglucoside and then centrifuged at 10,000 ϫ g for 12 min at 4°C. The resultant supernatant was labeled Triton X-100 DI. Equivalent cell amounts were loaded onto SDS-PAGE, and the proteins were detected with silver stain. Fig. 3B indicates that the protein profile in the DS fraction is similar in each sample tested. However, there is less protein detected in transitional immature DI fraction than the mature DI fraction. M␤CD/cholesterol treatment of transitional immature B cells resulted in a similar protein profile in the DI fraction as compared with the mature DI fraction.
We next wanted to determine if M␤CD/cholesterol treatment of transitional immature B cells resulted in greater co-polarization of cholesterol with anti-BCR-induced BCR aggregates. Anti-BCR-induced BCR aggregates in M␤CD/cholesteroltreated transitional immature B cells co-polarized with cholesterol in 47 Ϯ 3% of observed cells ( Fig. 3C and Table I). Together, these data suggest that M␤CD/cholesterol treatment of transitional immature B cells results in similar content and distribution of protein in DI fractions and similar BCR aggregate co-polarization with cholesterol as compared with mature B cells (47 Ϯ 3% versus 48 Ϯ 4%).

Addition of Cholesterol to the Immature WEHI-231 B Cell Line Promotes Localization of BCR to Detergent-insoluble Membrane Fractions-It
is not possible to recover the BCR from primary B cells on sucrose gradients after stimulation, owing to its strong association with the cytoskeleton. However, it has previously been shown that the BCR in the mature B cell line CH27 is present in the detergent-insoluble raft fractions after anti-BCR-induced receptor aggregation. In contrast, the BCR in the immature WEHI-231 B cell line is isolated in detergentsoluble fractions following identical stimulation (30). Therefore, we used sucrose gradient centrifugation on these B cell lines representative of primary immature and mature B cells to determine if M␤CD/cholesterol addition to WEHI-231 cells resulted in BCR isolation in detergent-insoluble fractions.
M␤CD/cholesterol addition to WEHI-231 B cells resulted in BCR isolation in detergent-insoluble fractions after BCR stimulation (Fig. 4). Isolation of the BCR in detergent-insoluble  fractions was dependent on BCR stimulation, because the BCR was not isolated in the detergent-insoluble fractions of resting WEHI-231 B cells treated with M␤CD/cholesterol. Sucrose gradients of CH27 cells were used as a comparison for mature B cells. The detergent-soluble marker transferrin receptor was not detected in any of the detergent-insoluble fractions tested. One potential caveat with sucrose gradient isolation of cholesterol-enriched domains is that some internal membranes are detergent-insoluble. To determine if the BCR isolated in the detergent-insoluble fraction of cholesterol-enriched WEHI-231 cells was located at the cell surface, we biotinylated WEHI-231 cells after cholesterol addition. The results indicate that the BCR found in the detergent-insoluble fraction was on the surface of the cell (data not shown).

Inducible BCR Co-localization with Cholesterol-enriched Compartments Results in Sustained Signaling Resembling
That of BCR-triggered Mature B Cells-Our laboratory and others have shown that signals generated through the BCR in mature and transitional immature B cells differ in both intensity and duration, specifically hydrolysis of phosphatidylinositol 4,5-bisphosphate and generation of inositol 3,4,5-trisphosphate (12,(31)(32)(33). Because cholesterol levels are different in transitional immature and mature B cells, and because cholesterol-enriched domains have been implicated in amplifying and/or sustaining signaling, we investigated if cholesterol addition to transitional immature B cells affects downstream signaling. In mature B cells, anti-BCR-induced PLC␥2 phosphorylation peaked at 2 min and then remained phosphorylated through 60 min (Fig. 5A). Although transitional B cells can maintain anti-BCR-induced phosphorylation through 60 min, the initial peak activation at 2 min was reduced compared with mature B cells. Cholesterol addition to transitional immature B cells resulted in a peak activation of anti-BCR-induced PLC␥2 phosphorylation that was similar to mature B cells at 2 min and was maintained through 60 min. Therefore, cholesterol addition to transitional immature B cells results in a signal intensity similar to that seen in mature B cells.
Recent studies have shown that cholesterol-enriched domains play a role in NFB signaling (13). This signaling cas-cade is initiated by PLC␥2 hydrolysis of phosphatidylinositol 4,5-bisphosphate, resulting in protein kinase C␤ activation of IB kinases that leads to ubiquitination and subsequent degradation of IB␣. In particular, induction of c-myc expression has been shown to be up-regulated by BCR signaling through this pathway (34). Furthermore, previous studies have shown that the BCR-induced increase in expression of c-myc transcripts in the immature B cell line WEHI-231 are transient compared with mature cell lines (35), suggesting lack of sustained signaling through this pathway in immature B cell lines. We therefore wanted to determine levels of c-myc transcript expression in primary transitional immature and mature B cells following BCR stimulation. For both transitional immature and mature B cells, there was a clear increase in c-myc expression after 1 h (Fig. 5C). However, although the anti-BCR stimulated mature B cells maintained expression through 4 h (74% of peak), levels decreased significantly in the anti-BCR stimulated transitional immature B cells (27% of peak). Thus, as compared with mature B cells, these results argue that transitional immature B cells are unable to sustain signals leading to c-myc expression following BCR aggregation. We next evaluated if M␤CD/cholesterol addition to transitional immature B cells leads to sustained c-myc mRNA expression levels. The addition of M␤CD/cholesterol to transitional immature B cells rescues c-myc expression after 4 h of anti-BCR treatment (63% of peak). Therefore, M␤CD/cholesterol addition to transitional immature B cells not only results in the association of BCR with cholesterol-enriched domains following BCR aggregation but also converts the signaling phenotype of transitional immature B cells to the sustained signaling phenotype of mature B cells. DISCUSSION BCR-linked differences between immature and transitional immature B cells and mature B cells have been documented through work from our laboratory (12,14,16) and others (15,32,33,36). Moreover, several of these studies (37, 38) implicate these differences in the regulation of the responses of immature and mature B cells to BCR cross-linking. In general, these FIG. 5. Cholesterol addition to transitional immature B cells results in a signaling phenotype similar to mature B cells. A, purified splenic B cells from wt BALB/c mice (mature) and day 14 sublethally irradiated mice (transitional immature) were unstimulated or stimulated for 2, 10, 30, and 60 min, lysed, and proteins were separated on SDS-PAGE. PLC␥2 and phospho-PLC␥2 were detected by Western blot. The relative phospho-PLC␥2 levels indicate the relative increase in phospho-PLC␥2 compared with background. B, purified splenic B cells from wt BALB/c mice (mature) and day 14 sublethally irradiated mice (transitional immature) were unstimulated or anti-BCR stimulated for 1 and 4 h. c-myc transcripts were identified by Northern blot analysis. ␤-Actin transcripts are shown as a control. Relative c-myc levels represent the percentage of signal remaining relative to peak induction at 1 h. studies have implicated differences in the ability to engage or maintain the phosphatidylinositol 4,5-bisphosphate hydrolysis 3 protein kinase C␤ 3 IB kinase 3 NF␤ 3 c-myc pathway leading to differential survival/cell cycle progression or apoptosis. The mechanistic basis for the differential regulation of specific aspects of BCR signaling at these two developmental stages remains unresolved. However, studies documenting the differential ability of the BCR in mature and transitional immature B cells and transformed cell line models to translocate into detergent-insoluble (1, 2) and GM1-enriched (14) microdomains are provocative. The studies reported here document distinct and biologically significant differences in the membrane cholesterol content of transitional immature and mature B cells. This observation was unexpected, because there are no other examples where the unesterified cholesterol content of mammalian cells is developmentally regulated. More remarkably, these differences integrate the previous observations regarding BCR signaling at these stages.
We show that cholesterol levels between transitional immature and mature B cells differ. Related studies using neurons have shown that variations in cholesterol levels within individual cells result in signaling changes important for neuronal function (39). Therefore, we hypothesized that the difference in cholesterol content between transitional immature and mature B cells may play a role in the developmentally regulated differences in BCR signal transduction observed at these developmental stages. It is possible that some of the cholesterol measured in these cells will likely be associated with endoplasmic reticulum, mitochondrial, and Golgi membrane fractions. However, in mammalian cells these pools represent a relative small fraction, as little as 10%, of the membrane-associated cholesterol (28,29). We believe that the most likely interpretation of our data is that the differences observed reflect differences in plasma membrane cholesterol.
Because cholesterol plays a role in maintaining membrane heterogeneity, we analyzed the partitioning of the BCR within the membrane in transitional immature and mature B cells. We have previously shown that BCR stimulation of mature B cells leads to receptor aggregation and co-localization with aggregated GM1 by the use of cholera toxin staining. However, although BCR stimulation of transitional immature B cells results in receptor aggregation, the GM1 on BCR-stimulated transitional immature B cells does not aggregate. Because studies have shown that cholera toxin binding is not limited to GM1 and insufficient fixation can result in receptor redistribution, we fixed cells in 4% paraformaldehyde with 0.1% glutaraldehyde and stained with the cholesterol marker filipin. Our initial results indicated greater cholesterol co-localization with the BCR in BCR-aggregated mature cells. The distinction between cholera toxin staining and filipin staining as a marker for membrane heterogeneity is important, because we are primarily interested in cholesterol variances between transitional immature and mature B cells. Inducible BCR co-localization with cholesterol supports the model that membrane cholesterol mediates membrane heterogeneity and that signals transmitted through the BCR may be modulated by differential localization within membrane microdomains.
Instead of basing our studies on a loss of function with the depletion of cholesterol, we determined if cholesterol addition to transitional cells results in a gain of function. To add cholesterol to the plasma membrane, we utilized M␤CD medium containing excess levels of cholesterol to shuttle cholesterol into the plasma membrane. Although the treatment of cells with M␤CD alone results in membrane perturbation and cell death, cell viability is generally not compromised when cells are treated with M␤CD and excess levels of cholesterol (40,41). Therefore, using this treatment we can increase the levels of cholesterol in transitional B cells and look for phenotypic differences in BCR receptor partitioning in the plasma membrane. Our results show that the low level of cholesterol in transitional immature B cells does not result in significant enrichment of cholesterol in BCR aggregates. However, treatment of transitional immature B cells with M␤CD/cholesterol restores enrichment of cholesterol within BCR aggregates.
The implications of these studies extend beyond the regulation of BCR proximal signaling. Together with our previous studies, these results suggest that cholesterol content plays a determining role in the fate decisions of transitional immature and mature B cells in response to antigen. One would predict that individuals with genetic defects affecting cholesterol production, uptake, and/or turnover in the plasma membrane might generate B cells with defective responses to antigen. For example, immature B cells with incrementally elevated plasma membrane cholesterol content might be less sensitive to BCR-induced apoptosis and, therefore, exhibit impaired negative selection. In these individuals, polymorphisms in genes regulating these processes would be considered genetic risk factors for autoimmunity. Similarly, mature B cells with incrementally lower cholesterol levels, as might occur in individuals treated with statins to lower serum cholesterol, could be compromised in their ability to respond to antigen. Our current studies point to these issues as important directions for future studies, including the relevance of these findings to human disease and potential therapeutic targets for modulating B cell responsiveness in autoimmune disorders and allogeneic transplants.