Recently, the role of B cells in atherosclerosis has gained more attention but studies have mainly focused on B1 and follicular B cell subsets. Therefore, the contribution of marginal zone (MZ) B cells in experimental atherosclerosis remains elusive. In the current study, we examined the MZ B cell compartment in atherosclerotic apoE-deficient (apoE−/−) mice and found that hypercholesterolemia in these mice was associated with an increased number and percentage of MZ B cells. This aberrant accumulation of MZ B cells was not associated with alterations in their development or increased proliferation but was due to decreased apoptotic cell death. This decrease in MZ B cell death in apoE−/− mice was associated with the reduced capacity of invariant NKT (iNKT) cells to produce IFN-γ and IL-4 after activation. Lowering cholesterol plasma levels with ezetimibe in apoE−/− mice reversed iNKT function and MZ B cell accumulation. To elucidate the mechanism whereby iNKT cells control MZ B cell accumulation in apoE−/− mice, we performed an adoptive transfer of iNKT cells and found that only wild-type iNKT cells but not IFN-γ−/− iNKT cells reversed MZ B cell accumulation in apoE−/− recipient mice. Our findings reveal that lipid changes associated with atherosclerotic disease induce decreased production of IFN-γ by iNKT, which in turn leads to aberrant accumulation of MZ B cells. This study further extends the importance of iNKT cells in regulating MZ B cell compartment.

B lymphocyte subsets comprise of B1 and B2 cells, which include follicular (FO) and marginal zone (MZ) B cells. In mice, MZ B cells are a unique subset of non-recirculating, long-lived B cells, which populate the splenic MZ where they represent 5% of total splenic B cells. MZ B cells develop from transitional precursor cell populations known as transitional 1 MZ precursor (T1 MZP) and transitional 2 MZP (T2 MZP) B cells (1). MZ B cells are strategically positioned next to the marginal sinuses, allowing them to respond rapidly to blood-borne bacteria pathogens (2). MZ B cells have been functionally linked to immune responses against type-2 T independent Ags, where they rapidly differentiate into short-lived plasmablasts that secrete large amounts of IgM (2). They express high levels of IgM, complement receptor type II CD21 and low levels of IgD and CD23 (3, 4). As MZ B cells exhibit a preactivated state, they are faster and more efficient than FO B cells in the capture, processing and presentation of Ag, and they deliver costimulatory signals to T cells due to higher basal levels of MHC class II, CD80 and CD86 (5). MZ B cells highly express the non-classical MHC class I molecule CD1d, allowing it to interact and activate invariant NKT (iNKT) cells, which recognize lipid Ags presented on CD1d (4). iNKT cells represent a highly conserved subset of mature T lymphocytes characterized by a TCR repertoire comprising an invariant Vα14/Jα18 chain paired preferentially with Vβ8, 7 or 2β-chains in the mouse, and can be found predominantly in the liver, spleen, and blood (68). The synthetic glycolipid α-galactosylceramide (α-GalCer) isolated from a marine sponge has been shown to selectively activate iNKT cells (9). Upon activation, iNKT cells are able to rapidly secrete large amounts of immuno-modulator cytokines such as IFN-γ and IL-4. This allows for a wide range of regulatory potential by cells of the innate and adaptive immune response such as NK cells, B lymphocytes and dendritic cells (10).

Hypercholesterolemia is characterized by elevated levels of total cholesterol and low density lipoprotein (LDL), which is the hallmark of atherosclerosis, a chronic inflammatory disease of medium and large arteries (11). In recent years, numerous studies have highlighted the importance of B lymphocyte subsets in diseases associated with cholesterol abnormalities such as atherosclerosis (1214), and diabetes (15). It is currently proposed that different B cell-subsets have different or even opposing contributions in atherosclerosis (16). B1 cells have been shown to be protective through their capacity to secrete natural IgM Abs against modified LDL that can block foam cell formation and promote apoptotic cell clearance. In contrast, B2 cells are thought to be proatherogeneic although the underlying mechanisms are still unknown. However, studies on atherosclerosis have mainly focused on total B1 and FO B cell populations and, therefore, the contribution of the MZ B cell subset remains elusive. In particular, the MZ B cell subset was found to be increased in mouse models of autoimmune disorders including lupus and diabetes (15, 17), where they may contribute to disease pathology through autoantigen presentation (15), and the production of autoantibodies (17, 18). In addition, MZ B cell activation and expansion partially contribute to the development of type 1 diabetes in NOD mice (19). Furthermore, the specific depletion of MZ B cells results in an amelioration of autoimmune symptoms in New Zealand Black/White, R4A- γ2b, and Sle1 mice models (18, 20, 21). Conversely, the elimination of MZ B cells in another lupus-prone mouse model leads to the development of autoantibodies that aggravate the disease (21). Finally, MZ B cells may protect against autoimmune disease through the secretion of IL-10 (22, 23). Given the potential contribution of MZ B cells in chronic inflammatory diseases, in this study we sought to investigate whether MZ B cells might be affected during atherosclerosis-associated hypercholesterolemia in apoE-deficient mice (apoE−/−). We found that MZ B cells markedly accumulated in apoE−/− mice from 24 wk of age compared with age-matched wild-type (WT) mice. The unrestrained accumulation of MZ B cells was not due to alterations in their early development or increased proliferation but to decreased apoptotic death mediated by iNKT cells. This iNKT cell-mediated pathway was dependent on IFN-γ. This study demonstrates the importance of iNKT cells in the regulation of MZ B cell accumulation.

Male WT and apoE−/− mice on a C57/BL6 background were obtained from the Jackson Laboratory (Bar Harbor, ME). Animals were maintained on a regular chow diet or switched at 6 wk of age to a high-fat, cholesterol-rich diet (21% fat and 0.15% cholesterol) (Harlan Teklad) until sacrifice. All mice were housed under specific pathogen-free conditions with unrestricted access to food and water in the animal housing unit of the National University of Singapore. All studies were approved by the Institutional Animal Care and Use Committee.

Plasma samples were collected via cardiac-puncture and total cholesterol was determined using Cholesterol/Cholesteryl Ester Quantification Kit K603-100 (Bio Vision) as previously described (24).

Single-cell suspensions were preincubated with 0.5 μg of Fc block (rat anti-mouse CD16/32; BD Pharmingen) per million cells, washed, and stained with fluorochrome conjugated primary or appropriate isotype control Abs: Primary biotin, FITC, PE, PerCP Cy5.5 and APC labeled rat Abs against mouse cell surface Ags B220 (RA-6B2), CD21/CD35 (eBio8D9), TCRβ (H57-597), CD23 (B3B4), IgM (1B4B1), CD1d (1B1), MHC class II (M5/114.15.2), CD69 (H12F3), CD86/B7-2 (GL1), CD80/B7-1 (16-10A1), CD40 (1C10) were purchased from eBioscience and BD Pharmingen. Cells were fixed in 1% formalin if necessary. Flow cytometry was performed either on a FACSCalibur (BD Biosciences, San Diego, CA) with CellQuest data acquisition and analysis software, or on a Cyan flow cytometer (Dako, Denmark) with Summit software. Data were analyzed with Flowjo software (Treestar).

Spleen cells were stained for B220, CD21, CD23 and IgM to sort MZ B cells (B220+ CD21high CD23low), T1 MZP (B220+ CD23neg IgM+ CD21low) and T2 MZP cells (B220+ CD23high IgMhigh CD21high) using a MOFLO cell sorter (DakoCytomation) according to the previous gating strategy (25). Sorted cells were then used for gene expression analysis by real-time PCR (RT-PCR).

All spleens were embedded in O.C.T Compound (Tissue-Tek). Serial 10 μm cryostat sections of spleens were cut and mounted on polysine-coated slides and stored at −20°C. Tissue sections were incubated with 50–100 μl of PBS/10% normal mouse serum containing primary Abs: rat anti-mouse Moma-1 FITC (AbD Serotec), rat anti-mouse CD1d (1B1) (Serotec) in a humidified chamber for 2 h at room temperature or overnight at 4°C (26). Tissue sections were incubated with 50 μl of PBS/10% normal mouse serum containing secondary Abs for 2 h in a humidified chamber at room temperature. Slides were incubated with DAPI (1:20,000) for 5 min. Sections were mounted with mounting media (Dako Cytomation), and examined under a Carl Zeiss Imager.Z1 fluorescence microscope with 60N-C 2/3” 0.63× AxioCam HRm camera.

Spleens were snap frozen in liquid nitrogen, and homogenized in RIPA buffer (Sigma Chemicals) supplemented with a protease inhibitor mixture (Roche Diagnostics). BAFF levels in spleen homogenates and plasma were assessed using a commercial BAFF ELISA kit as per the manufacturer’s protocol. For detection of IL-4 and IFN-γ, serum was collected 4 h post α-GalCer injection and IFN-γ and IL-4 levels were detected using commercial kits from R&D Systems according to the manufacturer’s protocol.

RT-PCRs were performed in triplicate on a 7500 Real-Time PCR System (Applied Biosystems). Each RT PCR contained 10 μl of iTaq SYBR Green supermix with ROX (Biorad), 250 nM forward and reverse primers (Table I), 7 μl of RNase free water, and 1 μl of cDNA template. The cycling conditions for all genes were: denaturation at 95°C for 10 min, 40 cycles of amplification with denaturation at 95°C for 30 s, annealing at 55°C for 1 min, and elongation at 72°C for 1 min. Data were collected at the elongation phase. A melt curve was performed by cooling to 55°C for 1 min, then increasing the temperature to 95°C. Gene expression was normalized to GAPDH.

Table I.
List of primers used in RT-PCR
Target GenePrimer Sequence
BAFF forward 5′-ACACGCCGACTATACGAAAAGGAAC-3′ 
BAFF reverse 5′-ACATCGGAACAGGGTCACCAGGCTCAG-3′ 
DL-1 forward 5′-GGGACAGAGGGGAGAAGATG-3′ 
DL-1 Reverse 5′-CACACCCTGGCAGACAGAT-3′ 
E2A- forward 5′-CGGAGAGCTGCAGATGGTGGC-3′ 
E2A-reverse 5′-AGGCTGCCATCTGCCACGTAGA-3′ 
Fli-1 forward 5′-AATGGATCCAGGGAGTCTCCGGT-3′ 
Fli-1 reverse 5′-TCGAACGTGCTCCTGTGTCCAC-3′ 
ID2- forward 5′-ATCGCCCTGGACTCGCATCC-3′ 
ID2-reverse 5′-GGGAATTCAGATGCCTGCAAGGACA-3′ 
ID3- forward 5′-CTACTCGCGCCTGCGGGAAC-3′ 
ID3- reverse 5′-AGCTCAGCTGTCTGGATCGGG-3′ 
CD79a- forward 5′-TACTTACCTCCGCGTGCGCAATCCA-3′ 
CD79a- reverse 5′-AGTCATCTGGCATGTCCACCCCA-3′ 
MINT- forward 5′-GTCCAAGCATGAAGACTGGAG-3′ 
MINT- reverse 5′-GGACCCTCTTCGTTCCTCTC-3′ 
Notch 2- forward 5′-TGCCTGTTTGACAACTTTGAGT-3′ 
Notch 2- reverse 5′-GTGGTCTGCACAGTATTGTCAT-3′ 
GAPDH- forward 5′-GACGGCCGCATCTTCTTGTG-3′ 
GAPDH- reverse 5′-CTTCCCATTCTCGGCCTTGACTGT-3′ 
Target GenePrimer Sequence
BAFF forward 5′-ACACGCCGACTATACGAAAAGGAAC-3′ 
BAFF reverse 5′-ACATCGGAACAGGGTCACCAGGCTCAG-3′ 
DL-1 forward 5′-GGGACAGAGGGGAGAAGATG-3′ 
DL-1 Reverse 5′-CACACCCTGGCAGACAGAT-3′ 
E2A- forward 5′-CGGAGAGCTGCAGATGGTGGC-3′ 
E2A-reverse 5′-AGGCTGCCATCTGCCACGTAGA-3′ 
Fli-1 forward 5′-AATGGATCCAGGGAGTCTCCGGT-3′ 
Fli-1 reverse 5′-TCGAACGTGCTCCTGTGTCCAC-3′ 
ID2- forward 5′-ATCGCCCTGGACTCGCATCC-3′ 
ID2-reverse 5′-GGGAATTCAGATGCCTGCAAGGACA-3′ 
ID3- forward 5′-CTACTCGCGCCTGCGGGAAC-3′ 
ID3- reverse 5′-AGCTCAGCTGTCTGGATCGGG-3′ 
CD79a- forward 5′-TACTTACCTCCGCGTGCGCAATCCA-3′ 
CD79a- reverse 5′-AGTCATCTGGCATGTCCACCCCA-3′ 
MINT- forward 5′-GTCCAAGCATGAAGACTGGAG-3′ 
MINT- reverse 5′-GGACCCTCTTCGTTCCTCTC-3′ 
Notch 2- forward 5′-TGCCTGTTTGACAACTTTGAGT-3′ 
Notch 2- reverse 5′-GTGGTCTGCACAGTATTGTCAT-3′ 
GAPDH- forward 5′-GACGGCCGCATCTTCTTGTG-3′ 
GAPDH- reverse 5′-CTTCCCATTCTCGGCCTTGACTGT-3′ 

A total of 1 × 106 splenocytes were seeded in 96 round-bottom plates, and cultured at 37°C in complete RPMI-1640 medium (Sigma) for 3 h. MZ B cells were FACS stained for surface markers. The cells were washed once with cold 1 × PBS, and resuspended in 100 μl of Annexin V Binding Buffer at a concentration of 1 × 106 cells/ml. Then 2 μl of Annexin V and 1 μl of PI was added into each tube, and incubated for 15 min at room temperature in the dark. After incubation, 200 μl of Annexin V binding buffer was added to each tube. Cells were analyzed immediately by flow cytometry.

To assess B cell proliferation, mice were fed with drinking water supplemented with BrdU at a concentration of 0.8 mg/ml for a period of 4 wk. The 4 wk treatment period was chosen because MZ B cells are long lived (25). The BrdU solution was changed once every 2 d. After isolating the spleen cells, surface Ag staining of B cell subsets was performed prior to staining of intracellular BrdU with the BD PharmingenTM BrdU Flow Kit (BD Biosciences) according to the manufacturer’s instructions.

For activation of iNKT cells, 8–10 wk old mice were injected with 1 μg of α-GalCer (Enzo) in 200 μl of 1 × PBS i.p. For induction of iNKT cell hyporesponsiveness, mice were injected with 5 μg of α-GalCer (Enzo) in 200 μl of 1 × PBS i.p. followed by a second injection with 1 μg of α-GalCer 1 mo later as described in the previous study (27). Mice were bled 4 h after injection, and the cytokines IL-4 and IFN-γ were measured in the serum. After 3 d, the mice were sacrificed, spleens were harvested, and single-cell suspensions were prepared for analysis of MZB cells by flow cytometry.

To measure the effects of lowering plasma cholesterol on MZB and NKT cells, 6 wk old male WT and apoE−/− mice on a C57/BL6 background were fed a high-fat diet for 6 wk. At 12 wk of age, the mice were treated daily with ezetimibe (Kemprotec) via oral gavage at a dose of 5 mg/kg/d dissolved in 200 μl of corn oil (24). The mice continued to be fed a high-fat diet during the treatment period.

To assess the involvement of CD1d in our system, 500 μg of anti-CD1d blocking Ab (clone 19G11; BioXcell) was injected i.p. into 20 wk old apoE−/− mice once every 3 d for 4 wk as previously described (28). After 4 wk, the spleen was harvested and cell suspensions were prepared for FACS staining of surface markers.

Splenocytes from 8 wk old WT or IFN-γ−/− mice were stained with APC labeled α-GalCer/CD1d tetramers (NIH tetramer facility), and PerCP-Cy5.5 labeled TCR-β (eBioscience) to detect and sort iNKT cells. At day 0, 5 × 105 selectively sorted iNKT cells with a viability >98% as determined by Trypan blue were adoptively transferred into recipient apoE−/− mice intravenously. One day later (day 1), mice were stimulated with 1 μg of α-GalCer and cytokines, and MZB cells were analyzed 4 h and 3 d later (day 3) after α-GalCer injection, respectively, as described above. In one experiment, iNKT cells isolated from CD45.2 WT or IFN-γ−/− mice were adoptively transferred into congenic CD45.1 WT mice. The adoptively transferred iNKT cells were identified in the spleen by flow cytometry as described above, and the total number of transferred iNKT was calculated by multiplying the percentage of iNKT cells by the total number of spleen cells.

Statistical analysis was performed using an unpaired, two-tailed t test for comparison between the two groups. Comparisons of three or more groups were performed using the Mann–Whitney U test with a post hoc test using Bonferroni’s method. All the data are presented as means ± SD. Prism 5 software (Graphpad) was used for all statistical analysis. A p value <0.05 was considered significant.

To investigate the possible effects of hypercholesterolemia associated with atherosclerosis on splenic B cell subpopulations, we quantified the percentages of MZ and FO B cells in the spleen of apoE−/− mice at 6, 16 and 24 wk of age. Mice were fed a diet rich in cholesterol (Western diet), which is known to accelerate atherosclerosis (29), and hypercholesterolemia (30) in apoE−/− mice. Consistent with previous reports, total plasma cholesterol levels reached 2000 mg/dl in apoE−/− at 24 wk of age (Table II). MZ B cells and FO B cells were identified as B220+ CD21high CD23low and B220+ CD23int CD21low/int, respectively (Fig. 1A). At 6 wk of age, there were no significant differences in the percentages of MZ B cells and FO B cells between WT and apoE−/− mice (Fig. 1B). In contrast, the percentage and the number of MZ B cells significantly increased 2-fold in 24 wk old apoE−/− mice compared with age-matched WT mice (Fig.1B). This increase in MZ B cells in 24 wk old apoE−/− mice was also accompanied by a significant decrease in the FO B cell percentage and number compared with their respective WT counterparts (Fig. 1B). There were no significant differences in spleen cellularity between WT and apoE−/− mice at 6, 16 and 24 wk of age, suggesting that hypercholesterolemia does not affect the cellularity of the spleen (Fig 1B) (31). Immunohistological analysis of the spleen confirmed the accumulation of MZ B cells in 24 wk old apoE−/− mice, as shown by an enlarged region of CD1dbright cells, which colocalized with Moma-1+ metallophilic macrophages surrounding the B cell follicles (Fig. 1C). These results suggest that hypercholesterolemia in apoE−/− mice was associated with an increase in the percentage and number of MZ B cells.

Table II.
Total plasma cholesterol (milligrams per deciliter) in WT and apoE−/− across different ages
6 wk (n = 6)16 wk (n = 5)24 wk (n = 6)
WT 124.56 ± 17.95 172.20 ± 91.34 181.70 ± 72.39 
apoE−/− 413.65 ± 56.71** 1733.68 ± 746.02* 1840.64 ± 382.27** 
6 wk (n = 6)16 wk (n = 5)24 wk (n = 6)
WT 124.56 ± 17.95 172.20 ± 91.34 181.70 ± 72.39 
apoE−/− 413.65 ± 56.71** 1733.68 ± 746.02* 1840.64 ± 382.27** 

Values are mean ± SD.

*p ≤ 0.05 and **p ≤ 0.005.

FIGURE 1.

MZ B cells accumulate in the spleen of hyperlipidemic apoE−/− mice. (A) Representative FACS plots of MZ and FO B cells in 24 wk old WT and apoE−/− mice. Splenocytes were isolated and stained with anti-B220-PerCP Cy5.5, anti-CD21-PE, and anti-CD23-FITC, and analyzed by flow cytometry. Numbers indicate the percentage of MZ and FO B cells among B220+ cells. MZ and FO B cells were gated as B220+ CD21 high CD23low and B220+ CD23int CD21low/int, respectively. (B) Percentage and number of MZ and FO B cells in 6, 16 and 24 wk old WT, and apoE−/− mice. Spleen cellularity in 16 and 24 wk old WT and apoE−/− mice. Results were pooled from three independent experiments with n = 8–10 mice per group. Results are shown as mean ± SD and ***p < 0.0001. (C) Immunofluorescence staining of spleen sections from 24 wk old WT and apoE−/− mice. MZ B cells (red) were stained for CD1d whereas metallophilic macrophages (green) were stained for Moma-1. Scale bars, 100 μm. Images are representative of two repeated experiments with n = 3 mice per group.

FIGURE 1.

MZ B cells accumulate in the spleen of hyperlipidemic apoE−/− mice. (A) Representative FACS plots of MZ and FO B cells in 24 wk old WT and apoE−/− mice. Splenocytes were isolated and stained with anti-B220-PerCP Cy5.5, anti-CD21-PE, and anti-CD23-FITC, and analyzed by flow cytometry. Numbers indicate the percentage of MZ and FO B cells among B220+ cells. MZ and FO B cells were gated as B220+ CD21 high CD23low and B220+ CD23int CD21low/int, respectively. (B) Percentage and number of MZ and FO B cells in 6, 16 and 24 wk old WT, and apoE−/− mice. Spleen cellularity in 16 and 24 wk old WT and apoE−/− mice. Results were pooled from three independent experiments with n = 8–10 mice per group. Results are shown as mean ± SD and ***p < 0.0001. (C) Immunofluorescence staining of spleen sections from 24 wk old WT and apoE−/− mice. MZ B cells (red) were stained for CD1d whereas metallophilic macrophages (green) were stained for Moma-1. Scale bars, 100 μm. Images are representative of two repeated experiments with n = 3 mice per group.

Close modal

MZ B cells derive from precursors present in the spleen including T1 MZP and T2 MZP cells (1). Therefore, the increased number of MZ B cells in apoE−/− mice might result from a preferential differentiation of precursor cells into MZ B cells. To address this possibility, we determined the percentage of T1 MZP and T2 MZP cells in 16 and 24 wk old WT and apoE−/− mice. T1 MZP and T2 MZP were identified as B220+ CD23neg IgM+ CD21low, and cells B220+ CD23high IgMhigh CD21high as reported previously (32) (Fig. 2A). We did not detect any differences in the percentage of T1 MZP and T2 MZP cells in apoE−/− mice at both 16 and 24 wk of age compared with WT mice (Fig. 2B). In T1 MZP and T2 MZP cells as well as MZ B cells sorted from 24 wk old WT and apoE−/− mice, we also examined the expression of factors controlling MZ B cell development including Notch 2, Δ like 1, MINT, E2A, ID2, ID3, Fli-1, CD79a, and mb-1. We did not detect any significant changes in gene expression of these factors by quantitative RT-PCR (Supplemental Fig. 1), suggesting that the increased number of MZ B cells in apoE−/− mice was unlikely to result from abnormal development.

FIGURE 2.

MZ B cell accumulation in apoE−/− mice is not associated with changes in transitional precursor cell populations. (A) Representative plots of T1 MZP and T2 MZP cells in 24 wk old WT, and apoE−/−mice. T1 MZP and T2 MZP cells were identified as B220+ CD23neg IgM+ CD21low, and B220+ CD23high IgMhigh CD21high, respectively. (B) Percentages of T1 MZP and T2 MZP cells in 16 and 24 wk old WT and apoE−/−mice. Results were pooled from two independent experiments with n = 4 mice per group. Results are shown as mean ± SD.

FIGURE 2.

MZ B cell accumulation in apoE−/− mice is not associated with changes in transitional precursor cell populations. (A) Representative plots of T1 MZP and T2 MZP cells in 24 wk old WT, and apoE−/−mice. T1 MZP and T2 MZP cells were identified as B220+ CD23neg IgM+ CD21low, and B220+ CD23high IgMhigh CD21high, respectively. (B) Percentages of T1 MZP and T2 MZP cells in 16 and 24 wk old WT and apoE−/−mice. Results were pooled from two independent experiments with n = 4 mice per group. Results are shown as mean ± SD.

Close modal

Because increased proliferation may account for the accumulation of MZ B cells in apoE−/− mice, we investigated the proliferative capacity of apoE−/− MZ B cells in vivo by measuring the incorporation of the thymidine analog BrdU supplemented in the drinking water for 4 wk. Consistent with a previous study (33), 40% of splenic MZ B cells proliferated in WT mice. However, we did not find any significant difference in the percentage of proliferating MZ B cells between WT and apoE−/− mice (Fig. 3A), suggesting that the increase in the MZ B cell compartment in apoE−/− mice was not due to an increased proliferation of MZ B cells. Next, we assessed the activation state of MZ B cells by analyzing the expression levels of costimulatory molecules and the activation marker CD69 in 24 wk old WT, and apoE−/− mice. We did not find any significant differences in the expression of CD40, CD80, CD86, and MHC class II on MZ B cells between WT and apoE−/− mice (Fig. 3B). We also failed to detect any upregulation of CD69 on MZ B cells in 24 wk old apoE−/− mice compared with their matching WT counterparts (Fig. 3B), indicating that activation of MZ B cells from apoE−/− mice is comparable to those in WT mice.

FIGURE 3.

MZ B cell accumulation in apoE−/− mice is not associated with their increased proliferation and activation. (A) Representative histogram of BrdU-positive MZ B cells. Proliferating MZ B cells gated as B220+CD21highCD23low were identified as positive for BrdU. Percentage of BrdU-positive MZ B cells in 24 wk old WT and apoE−/−mice. Results are representative of two independent experiments with n = 4 mice per group. (B) Flow cytometric analysis of costimulatory molecules expression (CD40, CD80, CD86), MHC class II and activation marker CD69 on MZ B cells from 24 wk old WT and apoE−/−mice. Results are representative of two independent experiments with n = 3–4 mice per group.

FIGURE 3.

MZ B cell accumulation in apoE−/− mice is not associated with their increased proliferation and activation. (A) Representative histogram of BrdU-positive MZ B cells. Proliferating MZ B cells gated as B220+CD21highCD23low were identified as positive for BrdU. Percentage of BrdU-positive MZ B cells in 24 wk old WT and apoE−/−mice. Results are representative of two independent experiments with n = 4 mice per group. (B) Flow cytometric analysis of costimulatory molecules expression (CD40, CD80, CD86), MHC class II and activation marker CD69 on MZ B cells from 24 wk old WT and apoE−/−mice. Results are representative of two independent experiments with n = 3–4 mice per group.

Close modal

Considering that the alternative explanations above were unlikely to account for the accumulation of MZ B cells in the spleen of apoE−/− mice, we hypothesized that the increase in MZ B cell numbers may result from changes in the survival of these cells. BAFF is a trimeric member of the tumor necrosis family that is found at high concentrations in lymphoid follicles where it functions to promote the survival of T2 MZP and MZ B cells by inducing signals through BAFF receptor (BAFFR) (34). Increased BAFF production leads to excess B cell survival, and the localization of autoreactive B cells in the MZ (35). We therefore investigated whether the accumulation of MZ B cells in apoE−/− mice was associated with elevated levels of BAFF. However, there were no significant differences in the mRNA and protein levels of BAFF in the spleen and plasma of 24 wk old WT and apoE−/− mice (Fig. 4A). In addition, T2 MZP cells and MZ B cells in 16 and 24 wk old WT and apoE−/− mice expressed similar levels of BAFFR (Fig. 4B). Collectively, these results suggest that the accumulation of MZ B cells in apoE−/− mice was not associated with changes in BAFF and BAFFR.

FIGURE 4.

Decreased death of MZ B cells is associated with MZ B cell accumulation in apoE−/− mice. (A) mRNA and protein levels of BAFF in the plasma and splenocyte lysates, respectively, from 24 wk old WT and apoE−/− mice. Results were pooled from two independent experiments with n = 6–7 mice per group. Results are shown as mean ± SD. (B) Flow cytometric analysis of BAFFR expression on T2 MZP cells and MZ B cells in 16 and 24 wk old WT and apoE−/−mice. Results are representative of two independent experiments with n = 4 mice per group. (C) Representative plots of apoptotic MZ B cells in 24 wk old WT and apoE−/−mice. Apoptotic MZ B cells identified as B220+ CD21high CD23low were gated as annexin V+ PI+. Percentage of apoptotic MZ B cells in 24 wk old WT and apoE−/− mice. Results were pooled from two independent experiments with n = 6–8 mice per group. Results are shown as mean ± SD with *p < 0.05.

FIGURE 4.

Decreased death of MZ B cells is associated with MZ B cell accumulation in apoE−/− mice. (A) mRNA and protein levels of BAFF in the plasma and splenocyte lysates, respectively, from 24 wk old WT and apoE−/− mice. Results were pooled from two independent experiments with n = 6–7 mice per group. Results are shown as mean ± SD. (B) Flow cytometric analysis of BAFFR expression on T2 MZP cells and MZ B cells in 16 and 24 wk old WT and apoE−/−mice. Results are representative of two independent experiments with n = 4 mice per group. (C) Representative plots of apoptotic MZ B cells in 24 wk old WT and apoE−/−mice. Apoptotic MZ B cells identified as B220+ CD21high CD23low were gated as annexin V+ PI+. Percentage of apoptotic MZ B cells in 24 wk old WT and apoE−/− mice. Results were pooled from two independent experiments with n = 6–8 mice per group. Results are shown as mean ± SD with *p < 0.05.

Close modal

Next, we examined whether the increased number of MZ B cells may be due to decreased death by apoptosis. To address this possibility, splenocytes isolated from 24 wk old WT and apoE−/− mice were cultured in the absence of stimulation to induce cell death prior to staining for annexin V and propidium iodide. A pilot experiment revealed that MZ B cells died beyond 16 h (data not shown), therefore a 6 h time point was chosen for analysis of MZ B cell apoptosis. Cells undergoing late apoptosis were identified as double positive for annexin V and propidium iodide (Fig. 4C). After 6 h of culture, we found a significant decrease in the percentage of apoptotic apoE−/− MZ B compared with WT MZ B cells (Fig. 4C), suggesting that hypercholesterolemia was associated with decreased death of MZ B cells in apoE−/− mice.

A recent report showed that iNKT cells activated with α-GalCer can induce death of MZ B cells following their activation (36). Interestingly, iNKT cells in hyperlipidemic apoE−/− mice have been reported to fail to produce IFN-γ and IL-4 cytokines after exogenous α-GalCer stimulation (37). Together with our results showing decreased death of MZ B cells in apoE−/− mice, these findings prompted us to hypothesize that a loss of iNKT cell function may render apoE−/− mice resistant to iNKT cell–induced MZ B cell death, resulting in unrestrained MZ B cell accumulation. First, we confirmed that the death of MZ B cells after α-GalCer injection was dependent on iNKT cells by evaluating the percentage of MZ B cells in Jα18−/− mice specifically lacking iNKT cells (38) after α-GalCer treatment. As expected, the administration of α-GalCer into WT mice led to a significant decrease in the MZ B cell percentage, which was observed 3 d (Fig. 5A) but not 1 d (data not shown) after α-GalCer injection. In contrast, we did not observe any changes in the percentage of MZ B cells in Jα18−/− mice treated with α-GalCer at day 3 (Fig. 5A). These results provide direct evidence that the death of MZ B cells was mediated by activated iNKT cells.

FIGURE 5.

The accumulation of MZ B cells in hyperlipidemic apoE−/− mice is associated with hyporesponsive iNKT cells. (A) Percentage of MZ B cells in WT and Jα18−/− mice treated or untreated with α-GalCer. (B) Percentage of iNKT cells in 24 wk old WT and apoE−/− mice treated or untreated with α-GalCer. (C) Serum IFN-γ and IL-4 levels in 24 wk old WT and apoE−/−mice treated or untreated with α-GalCer. (D) Percentage of MZ B cells in 24 wk old WT and apoE−/− mice treated or untreated with anti-CD1d Ab. (E) Percentage of MZ B cells in 24 wk old WT and apoE−/− mice treated or untreated with α-GalCer. Results are representative of two independent experiments with n = 4 mice per group. Results are shown as mean ± SD with *p < 0.05. (F) Percentage of iNKT cells in 24 wk old WT and apoE−/− mice after induction of iNKT cell hyporesponsiveness. (G) Serum IFN-γ and IL-4 levels in 24 wk old WT and apoE−/−mice after induction of iNKT cell hyporesponsiveness. (H) Percentage of MZ B cells in 24 wk old WT and apoE−/− mice after induction of iNKT cell hyporesponsiveness. Results were pooled from two independent experiments with n = 8 mice per group. Results are shown as mean ± SD with ***p < 0.0001.

FIGURE 5.

The accumulation of MZ B cells in hyperlipidemic apoE−/− mice is associated with hyporesponsive iNKT cells. (A) Percentage of MZ B cells in WT and Jα18−/− mice treated or untreated with α-GalCer. (B) Percentage of iNKT cells in 24 wk old WT and apoE−/− mice treated or untreated with α-GalCer. (C) Serum IFN-γ and IL-4 levels in 24 wk old WT and apoE−/−mice treated or untreated with α-GalCer. (D) Percentage of MZ B cells in 24 wk old WT and apoE−/− mice treated or untreated with anti-CD1d Ab. (E) Percentage of MZ B cells in 24 wk old WT and apoE−/− mice treated or untreated with α-GalCer. Results are representative of two independent experiments with n = 4 mice per group. Results are shown as mean ± SD with *p < 0.05. (F) Percentage of iNKT cells in 24 wk old WT and apoE−/− mice after induction of iNKT cell hyporesponsiveness. (G) Serum IFN-γ and IL-4 levels in 24 wk old WT and apoE−/−mice after induction of iNKT cell hyporesponsiveness. (H) Percentage of MZ B cells in 24 wk old WT and apoE−/− mice after induction of iNKT cell hyporesponsiveness. Results were pooled from two independent experiments with n = 8 mice per group. Results are shown as mean ± SD with ***p < 0.0001.

Close modal

Having confirmed that the activation of iNKT cells was required for inducing MZ B cell death in WT mice in response to α-GalCer, we went on to determine whether alterations in splenic iNKT cells in hyperlipidemic apoE−/− mice (37) was associated with the accumulation of MZ B cells in these mice. We found that α-GalCer treatment induced a similar expansion of splenic iNKT cells in WT and apoE−/− mice (Fig. 5B). However, compared with WT mice, the serum levels of IL4 and IFN-γ were blunted in apoE−/− mice after iNKT cell activation (Fig. 5C). Notably, the IFN-γ response was more suppressed than the IL-4 response. Consistent with a previous report (36), we observed a 2-fold decrease in the percentage of MZ B cells in WT mice treated with α-GalCer (Fig. 5D). In contrast, we did not observe any changes in MZ B cell percentage in apoE−/− mice treated with α-GalCer (Fig. 5D), suggesting that the accumulation of MZ B cells in apoE−/− mice might be associated with a defective function of iNKT cells. Because MZB cells express CD1d and iNKT cells recognize lipids presented via CD1d (4), we evaluated whether the effect on MZB cells in apoE−/− mice was dependent on CD1d using a specific blocking Ab against CD1d. Blocking CD1d only partially decreased the percentage of MZB cells in apoE−/− mice to WT levels (Fig. 5E).

Because we found an association between dysfunctional iNKT cells and the increase in MZ B cells in apoE−/− mice, we decided to test directly whether inducing iNKT cell hyporesponsiveness as established previously (27) may lead to MZ B cell accumulation in WT mice. Briefly, 20 wk old WT mice were preinjected with α-GalCer prior restimulation with a second dose of α-GalCer 1 mo later. Compared to control mice, iNKT cells from mice that received α-GalCer 1 mo earlier failed to expand in response to α-GalCer restimulation (Fig. 5F). Furthermore, these iNKT cells showed a reduced capacity to produce IL-4 and IFN-γ upon α-GalCer challenge (Fig. 5G). Having confirmed that the preinjection of α-GalCer into WT mice 1 mo before challenge with α-GalCer reduced the cytokine response by iNKT cells, we went on to determine whether induction of iNKT cell hyporesponsiveness affected the percentage of splenic MZ B cells. Interestingly, we found a 2-fold increase in the percentage of MZ B cells in WT mice in which iNKT were hyporesponsive (Fig. 5H). These results provide evidence that reduced production of cytokines by iNKT cells can lead to the accumulation of MZ B cells. Thus, iNKT cells hyporesponsiveness observed in apoE−/− mice may account for the increased percentage of MZ B cells in these mice.

Because we showed that hypercholesterolemia was associated with iNKT cell dysfunction and MZ B cell accumulation, we investigated whether lowering cholesterol may rescue iNKT cell capacity to produce cytokines, and reverse the accumulation of MZ B cells in apoE−/− mice. ApoE−/− mice at 12 wk of age were treated with ezetimibe by oral gavage for 3 mo as described previously (30, 39). Prior to ezetimibe treatment (week 0), apoE−/− mice at 12 wk of age exhibited marked hypercholesterolemia with total cholesterol levels reaching 2000 mg/dl (Fig. 6A). By 6 wk of treatment, ezetimibe significantly reduced total cholesterol levels in apoE−/− mice, which reached 400 mg/dl after 12 wk of treatment (Fig. 6A). In contrast, plasma cholesterol levels in apoE−/− mice left untreated remained above 2000 mg/dl (Fig. 6A). At the end of the 12 wk treatment, mice were injected with 1 μg of α-GalCer. We showed that the reduction of cholesterol levels in apoE−/− mice restored the percentage and the number of MZ B cells to WT levels (Fig. 6B). This was also confirmed by immunofluorescence staining as shown by a significant reduction in the amount of CD1dbright cells surrounding the ring of Moma-1+ macrophages demarcating the MZ (Fig. 6C). Furthermore, this decrease in the MZ B cell percentage observed in ezetimibe treated apoE−/− mice was associated with a significant production of IFN-γ and IL-4 cytokines in response to α-GalCer stimulation (Fig. 6D, 6E). Collectively, these results indicate that lowering cholesterol in apoE−/− mice restored the function of iNKT cells, which in turn reversed MZ B cell accumulation.

FIGURE 6.

Ezetimibe treatment reverses iNKT cell hyporesponsiveness and restores MZ B cell percentage to WT levels. (A) Total plasma cholesterol in ezetimibe treated and non-treated apoE−/− mice throughout the 12 wk treatment (n = 5–10 mice per group). (B) Percentage and number of MZ B cells in ezetimibe-treated and non-treated WT and apoE−/− mice. Results are representative of three independent experiments with n = 5–6 mice per group. Results are shown as mean ± SD with **p < 0.005, *p < 0.05. (C) Immunofluorescence staining of spleen sections from ezetimibe-treated and non-treated WT and apoE−/− mice. MZ B cells (red) were stained for CD1d and metallophilic macrophages (green) were stained for Moma-1. Scale bars, 100 μm. Images are representative of two repeated experiments with n = 3–4 mice per group. (D) Serum IFN-γ and IL-4 levels in 24 wk old WT and apoE−/− mice treated with ezetimibe, stimulated or unstimulated with α-GalCer. (E) Percentage of MZ B cells in 24 wk old WT and apoE−/− mice treated with ezetimibe, stimulated or unstimulated with α-GalCer. Results were obtained from one independent experiment with n = 4 mice per group. Results are shown as mean ± SD with ***p < 0.0001, **p < 0.005, *p < 0.05.

FIGURE 6.

Ezetimibe treatment reverses iNKT cell hyporesponsiveness and restores MZ B cell percentage to WT levels. (A) Total plasma cholesterol in ezetimibe treated and non-treated apoE−/− mice throughout the 12 wk treatment (n = 5–10 mice per group). (B) Percentage and number of MZ B cells in ezetimibe-treated and non-treated WT and apoE−/− mice. Results are representative of three independent experiments with n = 5–6 mice per group. Results are shown as mean ± SD with **p < 0.005, *p < 0.05. (C) Immunofluorescence staining of spleen sections from ezetimibe-treated and non-treated WT and apoE−/− mice. MZ B cells (red) were stained for CD1d and metallophilic macrophages (green) were stained for Moma-1. Scale bars, 100 μm. Images are representative of two repeated experiments with n = 3–4 mice per group. (D) Serum IFN-γ and IL-4 levels in 24 wk old WT and apoE−/− mice treated with ezetimibe, stimulated or unstimulated with α-GalCer. (E) Percentage of MZ B cells in 24 wk old WT and apoE−/− mice treated with ezetimibe, stimulated or unstimulated with α-GalCer. Results were obtained from one independent experiment with n = 4 mice per group. Results are shown as mean ± SD with ***p < 0.0001, **p < 0.005, *p < 0.05.

Close modal

To provide direct evidence that iNKT cell hyporesponsiveness accounts for MZ B cell expansion in apoE−/− mice, we evaluated the effect of transferring functional WT iNKT in apoE−/− mice on the MZ B cell compartment in these mice. iNKT cells (5 × 105) isolated from WT mice were adoptively transferred into recipient apoE−/− mice 1 d before stimulation with α-GalCer. Compared to non-recipient apoE−/− mice, apoE−/− mice that received WT iNKT cells showed a significant increase in the production of IFN-γ but not IL-4 after α-GalCer stimulation (Fig. 7A). In addition, the adoptive transfer of WT iNKT cells in apoE−/− mice significantly decreased the percentage of MZ B cells to WT levels 3 d after adoptive transfer (Fig. 7B). These results suggest that the adoptive transfer of functional WT iNKT cells into apoE−/− mice significantly increased the production of IFN-γ upon stimulation and reversed the accumulation of MZ B cells.

FIGURE 7.

Adoptive transfer of WT iNKT cells but not IFN-γ −/− iNKT cells in apoE−/− mice reverses MZ B cell expansion. (A) 5 × 105 WT iNKT cells were adoptively transferred into 24 wk old apoE−/− mice. Serum IFN-γ and IL-4 levels 4 h after 1 μg α-GalCer stimulation in 24 wk old recipient apoE−/− mice. (B) Percentage of MZ B cells in 24 wk old recipient apoE−/− mice stimulated or unstimulated with α-GalCer. Results are pooled from two independent experiments with n = 6–8 per group. Results are shown as mean ± SD and *p < 0.05, **p < 0.005. (C) 5 × 105 IFN-γ−/− iNKT cells were adoptively transferred into 24 wk old apoE−/− mice. Serum IFN-γ and IL-4 levels in 24 wk old recipient apoE−/− mice after α-GalCer stimulation. (D) Percentage of MZ B cells in 24 wk old recipient apoE−/− mice stimulated or unstimulated with α-GalCer. Results are shown as mean ± SD with n = 4 mice per group. ***p < 0.0001, **p < 0.005, *p < 0.05.

FIGURE 7.

Adoptive transfer of WT iNKT cells but not IFN-γ −/− iNKT cells in apoE−/− mice reverses MZ B cell expansion. (A) 5 × 105 WT iNKT cells were adoptively transferred into 24 wk old apoE−/− mice. Serum IFN-γ and IL-4 levels 4 h after 1 μg α-GalCer stimulation in 24 wk old recipient apoE−/− mice. (B) Percentage of MZ B cells in 24 wk old recipient apoE−/− mice stimulated or unstimulated with α-GalCer. Results are pooled from two independent experiments with n = 6–8 per group. Results are shown as mean ± SD and *p < 0.05, **p < 0.005. (C) 5 × 105 IFN-γ−/− iNKT cells were adoptively transferred into 24 wk old apoE−/− mice. Serum IFN-γ and IL-4 levels in 24 wk old recipient apoE−/− mice after α-GalCer stimulation. (D) Percentage of MZ B cells in 24 wk old recipient apoE−/− mice stimulated or unstimulated with α-GalCer. Results are shown as mean ± SD with n = 4 mice per group. ***p < 0.0001, **p < 0.005, *p < 0.05.

Close modal

Because IFN-γ production in response to α-GalCer stimulation was more affected than IL-4 response in apoE−/− mice, we hypothesized that activated iNKT cells may induce MZ B cell death via the secretion of IFN-γ. To test this hypothesis, 5 × 105 IFN-γ−/− iNKT cells were adoptively transferred into 24 wk old recipient apoE−/− mice 1 d prior stimulation with α-GalCer. The adoptive transfer of IFN-γ−/− iNKT cells in apoE−/− mice restored neither IFN-γ production (Fig. 7C) nor the percentage of MZ B cells after α-GalCer stimulation (Fig. 7D). However, the number of IFN-γ−/− iNKT cells tracked in spleen after adoptive transfer was comparable to WT iNKT cells (Supplemental Fig. 2). The production of IL-4 remained unchanged (Fig. 7C) as observed after the transfer of WT iNKT cells (Fig.7A). Altogether, these results show that the death of MZ B cells induced by activated iNKT cells was in part mediated by IFN-γ and this regulatory process mediated by iNKT cells is altered in hypercholesterolemic mice.

MZ B cells are strategically positioned next to the marginal sinuses, allowing them to respond rapidly to blood-borne bacteria pathogens. In our study, we showed that MZ B cells accumulate in hypercholesterolemic apoE−/− mice and that this accumulation results from their enhanced survival supported by the loss of IFN-γ released by iNKT cells.

We found that MZ B cell percentages and numbers were normal in the spleens of apoE−/− mice at 6 wk of age, but beyond 24 wk, when mice developed marked hypercholesterolemia, MZ B cells accumulated. Therefore, MZ B cells develop and mature normally in apoE−/− mice, but they accumulate in response to hypercholesterolemia. However, we cannot eliminate the possibility that the primary defect is a lack of apoE, which is manifested late in life. Because we did not observe more MZ B cell precursors or proliferating MZ B cells in the spleen of apoE−/− mice, it seems most likely that the decreased death by apoptosis and thus enhanced survival of MZ B cells explain the accumulation of MZ B cell in apoE−/− mice. The enhanced survival of MZ B cells was not associated with changes in BAFF and its receptor or increased gene expression of the prosurvival factor Bcl-2 (data not shown).

Some reports have suggested that iNKT cells can control MZ B cell numbers by inducing their death by apoptosis upon activation with α-GalCer in WT mice (36). Consistent with this study, we showed that α-GalCer administration induced WT iNKT cells to rapidly secrete large amounts of cytokines IL-4 and IFN-γ, which led to a 2-fold decrease in MZ B cells. This effect on MZB cells was rapid as it was observed as early as 3 d after α-GalCer administration. We also provide direct evidence that this decrease in MZ B cell percentage after α-GalCer administration is iNKT–cell dependent through the use of iNKT cell–deficient Jα18−/− mice. Although we may have expected an expansion of MZ B cells in these mice in the absence of stimulation, a normal MZ B cell compartment was observed. This observation may indicate that prior stimulation is required for iNKT cell to induce MZ B cell death or that Jα18−/− mice have developed a compensatory mechanism to keep the MZ B cell compartment in check. In stark contrast to our observations in WT mice, apoE−/− iNKT cells showed a blunted cytokine response and MZ B cells were not affected after α-GalCer administration.

According to the literature (37), we speculate that the hyporesponsive iNKT cell phenotype observed in apoE−/− mice may result from their chronic activation by elevated levels of lipids. However, the lipid ligand responsible for this chronic activation and defective iNKT cell function in apoE−/− mice remains to be identified. LDL or oxLDL may be involved because elevated levels of Abs against these lipid Ags have been detected in the blood of hypercholesterolemic mice and humans (4043). Furthermore, it was also reported that iNKT cells are not spontaneously activated in non-hypercholesterolemic apoE−/− mice at 5 wk of age (37). In addition, we show that young apoE−/− mice did not develop an expanded MZ B cell compartment, supporting the hypothesis that prolonged exposure to lipids (during life) and hyporesponsive iNKT cells are required for MZ B cells accumulation. This hyporesponsiveness of iNKT cells may also result from the absence of apoE because it was demonstrated that the presentation of CD1d-restricted lipid Ag by human B cells to NKT cells is mediated by apoE (43). In our study, we found that the effects on iNKT cells and MZB cells observed in apoE−/− mice are partially dependent on CD1d.

Interestingly, we were able to reverse the expansion of MZ B cells through decreasing cholesterol levels or transferring functional WT iNKT cells into apoE−/− mice. Cholesterol lowering was achieved through ezetimibe treatment, which decreased the MZ B cell percentage to WT levels. In addition, iNKT cells in ezetimibe-treated apoE−/− mice regained their functional properties, and were able to secrete IFN-γ and IL-4 cytokines upon α-GalCer stimulation. Thus, lowering cholesterol prevented iNKT cells from becoming hyporesponsive, supporting the idea that chronic lipid exposure and activation of iNKT cells promote their development into hyporesponsive phenotype where iNKT cells lose their ability to induce MZ B cell death, leading to their accumulation in apoE−/− mice.

We also demonstrate that induction of iNKT–cell hyporesponsiveness in WT mice supports the expansion of MZ B cells as the administration of a first dose of α-GalCer into WT mice 1 mo prior stimulation with a second dose of α-GalCer led to a 2-fold increase in the MZ B cell percentage compared with non–pre-injected WT mice. In addition, the adoptive transfer of functional WT iNKT cells into apoE−/− mice reversed MZ B cell accumulation, suggesting that reintroducing functional iNKT cells into apoE−/− mice was sufficient to restrain the accumulation of splenic MZ B cells.

Because IL-4 and IFN-γ production was highly suppressed in hyperlipidemic apoE−/− mice upon iNKT cell activation, we speculated that iNKT cells may promote the death of MZ B cells in a cytokine-dependent manner. IFN-γ is a cytokine that controls the Th1 response against viral and intracellular bacterial infections. It negatively regulates B cell differentiation and proliferation by mediating apoptosis (44, 45). Studies have shown that IFN-γ induces apoptosis in normal pre-B cell lines (46). In our study, we show that IFN-γ production by iNKT cells was required to induce the death of MZ B cells because the transfer of IFN-γ−/− iNKT cells into apoE−/− mice failed to reverse the expansion of MZ B cells. In addition, the transfer of functional WT iNKT cells into apoE−/− mice, which rapidly reversed MZ B cell expansion was associated with a marked increase in IFN-γ but not IL-4 cytokine production upon α-GalCer stimulation. In fact, the IFN-γ cytokine response toward α-GalCer was highly suppressed compared with the IL-4 cytokine response in apoE−/− mice. Thus, the absence of MZ B cell death upon α-GalCer stimulation in apoE−/− mice is likely attributed in part to the decrease in IFN-γ production by iNKT cells. However, we cannot rule out the possibility that NK cell–derived IFN-γ could contribute to the control of MZ B cell accumulation in apoE−/− mice as NK cells are also activated and produce IFN-γ rapidly upon α-GalCer injection in vivo (47). IL-4 has been shown to protect B lymphocytes from death by upregulating the transcription of Bcl-xl in a Stat6-dependent manner (48). Although the production of IL-4 after iNKT–cell activation in apoE−/− mice was decreased, low levels of IL-4 were still detected and thus, together with the absence of IFN-γ production, IL-4 may contribute to the accumulation of MZ B cells in apoE−/− mice.

A recent study by Sag et al. (49) described that pretreated iNKT cells with α-GalCer proliferate and produce decreased amount of IL-4 and IFN-γ but acquired the capacity to secrete IL-10. The authors propose that these IL-10–producing iNKT with regulatory potential (NKT10 cells) represent a distinct iNKT cell subset. In contrast, anergic iNKT cells described by Parekh et al. (27) exhibit impaired proliferation and production of IL-4 and IFN-γ upon α-GalCer restimulation. In our study, we showed that iNKT cells in apoE−/− mice expanded upon α-GalCer treatment but they had a diminished capacity to secrete cytokines. Thus, the population of iNKT cells in apoE−/− mice appears to resemble the NKT10 cells more than anergic iNKT cells, which do not proliferate upon α-GalCer restimulation. However, it remains to be determined whether the hyporesponsive iNKT cells described in our study are capable of secreting IL-10. A previous study by Major et al. (37) also reported that iNKT in apoE−/− mice produced less IL-4 and IFN-γ upon α-GalCer restimulation but they also failed to expand. One possible explanation for this discrepancy between this latter study and ours could be the concentration of α-GalCer used to restimulated iNKT cells.

iNKT cell–mediated MZ B cell death could also be contact dependent. Activated iNKT cells express elevated levels of Fas-L, which could interact with Fas on MZ B cells, inducing their apoptosis (50). iNKT cells promote an increase in the expression of Fas (CD95), and caspase-3 by MZ B cells and limited their proliferation in vitro (36). In addition, a recent study identified a novel mechanism by which iNKT cells kill CD1d expressing cells in atherosclerotic lesions through caspase-3–dependent apoptosis, and this was mediated via perforin and granzyme B (51). Because MZ B cells express high levels of CD1d, the decreased death of MZ B cells in apoE−/− mice could also result from decreased perforin and granzyme B mediated caspase-3–dependent apoptosis resulting from iNKT cell hyporesponsiveness.

What might be the possible biological consequences of MZ B cell accumulation in atherosclerosis? As MZ B cells are optimally positioned to detect pathogens in the bloodstream, it is possible that they may function to monitor changes in LDL and/or oxLDL levels, resulting in an immune response toward these atherogenic Ags. MZ B cells are able to uptake oxLDL Ag (data not shown), and they are known to be one of the major producers of natural IgM Abs. In humans, oxLDL IgM Abs were found to be inversely associated with the outcome of cardiovascular disease (43, 52). Interestingly, a robust extrafollicular response involving increased production of oxLDL-specific IgM Abs was observed in the apoE−/− spleen (31). Given that MZ B cells respond to TI-2 Ags in vivo and an expansion was found in the spleen of apoE−/− mice, it is possible that MZ B cells may participate in the extrafollicular reaction, giving rise to the increase in oxLDL IgM responses. However, there are conflicting views on whether these IgM Abs are atheroprotective or atherogenic. Because a significant increase in plaque size was reported in atherosclerotic mice that lacked serum IgM, it is speculated that the IgM response to oxLDL may be atheroprotective (53). In contrast, another study showed that less advanced athersoclerotic lesion development in apoE−/− CD40L−/− mice did not correlate with higher levels of oxLDL IgM Abs (54). Numerous studies have shown that iNKT cells contribute to atherosclerosis progression, and CD4+ iNKT cells were suggested to be the proatherogenic subset due to their increased ability to secrete proinflammatory cytokines IFN-γ, TNF-α, and IL-2 upon in vitro α-GalCer stimulation (55, 56). The repeated activation of iNKT cells by α-GalCer in vivo led to an increase in aortic lesion formation in apoE−/− mice. Further evidence for a proatherogenic role of iNKT cells using CD1d-deficient mice that lacked iNKT cells was also described. For instance, CD1d−/−apoE−/− mice showed a significant reduction of atherosclerotic lesions in the aortic root (57, 58). Therefore, a defect in iNKT cells in apoE−/− mice during the advanced stages of atherosclerosis could possibly be protective. This protective effect could also be mediated indirectly by regulating the size of the MZ B cell population.

In summary, we propose that the hypercholesterolemic environment in apoE−/− mice may allow for the uptake of an unknown lipid ligand by MZ B cells, resulting in the specific activation of iNKT cells over time. During the early stages of the disease, iNKT cells are functional and keep MZ B cell accumulation in check through the secretion of IFN-γ, which promotes the death of MZ B cells. As the disease progresses, the hyperlipidemic environment in apoE−/− mice may result in the chronic activation of iNKT cells that lose the ability to secrete IFN-γ cytokine, resulting in the accumulation of MZ B cells.

We thank Teo Guohui and Dr. Paul Edward Hutchinson (Flow Cytometry Lab, National University of Singapore) for sharing their expertise. We thank Dr. Hii Chung Shii and Wong Hui Sian Fiona for critical discussion.

This work was supported by National Medical Research Council and National Research Foundation grants to V.A.

The online version of this article contains supplemental material.

Abbreviations used in this article:

BAFFR

BAFF receptor

FO

follicular

α-GalCer

α-galactosylceramid

iNKT

invariant NKT

MZ

marginal zone

MZP

MZ precursor

RT-PCR

real-time PCR

T1 MZP

transitional 1 MZ precursor

T2 MZP

transitional 2 MZP

WT

wild type.

1
Pillai
S.
,
Cariappa
A.
.
2009
.
The follicular versus marginal zone B lymphocyte cell fate decision.
Nat. Rev. Immunol.
9
:
767
777
.
2
Pillai
S.
,
Cariappa
A.
,
Moran
S. T.
.
2005
.
Marginal zone B cells.
Annu. Rev. Immunol.
23
:
161
196
.
3
Oliver
A. M.
,
Martin
F.
,
Gartland
G. L.
,
Carter
R. H.
,
Kearney
J. F.
.
1997
.
Marginal zone B cells exhibit unique activation, proliferative and immunoglobulin secretory responses.
Eur. J. Immunol.
27
:
2366
2374
.
4
Oliver
A. M.
,
Martin
F.
,
Kearney
J. F.
.
1999
.
IgMhighCD21high lymphocytes enriched in the splenic marginal zone generate effector cells more rapidly than the bulk of follicular B cells.
J. Immunol.
162
:
7198
7207
.
5
Attanavanich
K.
,
Kearney
J. F.
.
2004
.
Marginal zone, but not follicular B cells, are potent activators of naive CD4 T cells.
J. Immunol.
172
:
803
811
.
6
Benlagha
K.
,
Weiss
A.
,
Beavis
A.
,
Teyton
L.
,
Bendelac
A.
.
2000
.
In vivo identification of glycolipid antigen-specific T cells using fluorescent CD1d tetramers.
J. Exp. Med.
191
:
1895
1903
.
7
Godfrey
D. I.
,
MacDonald
H. R.
,
Kronenberg
M.
,
Smyth
M. J.
,
Van Kaer
L.
.
2004
.
NKT cells: what’s in a name?
Nat. Rev. Immunol.
4
:
231
237
.
8
Kronenberg
M.
2005
.
Toward an understanding of NKT cell biology: progress and paradoxes.
Annu. Rev. Immunol.
23
:
877
900
.
9
Kawano
T.
,
Cui
J.
,
Koezuka
Y.
,
Toura
I.
,
Kaneko
Y.
,
Motoki
K.
,
Ueno
H.
,
Nakagawa
R.
,
Sato
H.
,
Kondo
E.
, et al
.
1997
.
CD1d-restricted and TCR-mediated activation of valpha14 NKT cells by glycosylceramides.
Science
278
:
1626
1629
.
10
Gumperz
J. E.
,
Miyake
S.
,
Yamamura
T.
,
Brenner
M. B.
.
2002
.
Functionally distinct subsets of CD1d-restricted natural killer T cells revealed by CD1d tetramer staining.
J. Exp. Med.
195
:
625
636
.
11
Hansson
G. K.
,
Libby
P.
.
2006
.
The immune response in atherosclerosis: a double-edged sword.
Nat. Rev. Immunol.
6
:
508
519
.
12
Kyaw
T.
,
Cui
P.
,
Tay
C.
,
Kanellakis
P.
,
Hosseini
H.
,
Liu
E.
,
Rolink
A. G.
,
Tipping
P.
,
Bobik
A.
,
Toh
B. H.
.
2013
.
BAFF receptor mAb treatment ameliorates development and progression of atherosclerosis in hyperlipidemic ApoE(-/-) mice.
PLoS One
8
:
e60430
.
13
Kyaw
T.
,
Tay
C.
,
Hosseini
H.
,
Kanellakis
P.
,
Gadowski
T.
,
MacKay
F.
,
Tipping
P.
,
Bobik
A.
,
Toh
B. H.
.
2012
.
Depletion of B2 but not B1a B cells in BAFF receptor-deficient ApoE mice attenuates atherosclerosis by potently ameliorating arterial inflammation.
PLoS One
7
:
e29371
.
14
Kyaw
T.
,
Tay
C.
,
Krishnamurthi
S.
,
Kanellakis
P.
,
Agrotis
A.
,
Tipping
P.
,
Bobik
A.
,
Toh
B. H.
.
2011
.
B1a B lymphocytes are atheroprotective by secreting natural IgM that increases IgM deposits and reduces necrotic cores in atherosclerotic lesions.
Circ. Res.
109
:
830
840
.
15
Mariño
E.
,
Batten
M.
,
Groom
J.
,
Walters
S.
,
Liuwantara
D.
,
Mackay
F.
,
Grey
S. T.
.
2008
.
Marginal-zone B-cells of nonobese diabetic mice expand with diabetes onset, invade the pancreatic lymph nodes, and present autoantigen to diabetogenic T-cells.
Diabetes
57
:
395
404
.
16
Tsiantoulas
D.
,
Sage
A. P.
,
Mallat
Z.
,
Binder
C. J.
.
2015
.
Targeting B cells in atherosclerosis: closing the gap from bench to bedside.
Arterioscler. Thromb. Vasc. Biol.
35
:
296
302
.
17
Grimaldi
C. M.
,
Michael
D. J.
,
Diamond
B.
.
2001
.
Cutting edge: expansion and activation of a population of autoreactive marginal zone B cells in a model of estrogen-induced lupus.
J. Immunol.
167
:
1886
1890
.
18
Zeng
D.
,
Lee
M. K.
,
Tung
J.
,
Brendolan
A.
,
Strober
S.
.
2000
.
Cutting edge: a role for CD1 in the pathogenesis of lupus in NZB/NZW mice.
J. Immunol.
164
:
5000
5004
.
19
Bashratyan
R.
,
Sheng
H.
,
Regn
D.
,
Rahman
M. J.
,
Dai
Y. D.
.
2013
.
Insulinoma-released exosomes activate autoreactive marginal zone-like B cells that expand endogenously in prediabetic NOD mice.
Eur. J. Immunol.
43
:
2588
2597
.
20
Peeva
E.
,
Michael
D.
,
Cleary
J.
,
Rice
J.
,
Chen
X.
,
Diamond
B.
.
2003
.
Prolactin modulates the naive B cell repertoire.
J. Clin. Invest.
111
:
275
283
.
21
Hwang
S. H.
,
Lee
H.
,
Yamamoto
M.
,
Jones
L. A.
,
Dayalan
J.
,
Hopkins
R.
,
Zhou
X. J.
,
Yarovinsky
F.
,
Connolly
J. E.
,
Curotto de Lafaille
M. A.
, et al
.
2012
.
B cell TLR7 expression drives anti-RNA autoantibody production and exacerbates disease in systemic lupus erythematosus-prone mice.
J. Immunol.
189
:
5786
5796
.
22
Hussain
S.
,
Delovitch
T. L.
.
2007
.
Intravenous transfusion of BCR-activated B cells protects NOD mice from type 1 diabetes in an IL-10-dependent manner.
J. Immunol.
179
:
7225
7232
.
23
Lenert
P.
,
Brummel
R.
,
Field
E. H.
,
Ashman
R. F.
.
2005
.
TLR-9 activation of marginal zone B cells in lupus mice regulates immunity through increased IL-10 production.
J. Clin. Immunol.
25
:
29
40
.
24
Lim
H. Y.
,
Thiam
C. H.
,
Yeo
K. P.
,
Bisoendial
R.
,
Hii
C. S.
,
McGrath
K. C.
,
Tan
K. W.
,
Heather
A.
,
Alexander
J. S.
,
Angeli
V.
.
2013
.
Lymphatic vessels are essential for the removal of cholesterol from peripheral tissues by SR-BI-mediated transport of HDL.
Cell Metab.
17
:
671
684
.
25
Westerberg
L. S.
,
Dahlberg
C.
,
Baptista
M.
,
Moran
C. J.
,
Detre
C.
,
Keszei
M.
,
Eston
M. A.
,
Alt
F. W.
,
Terhorst
C.
,
Notarangelo
L. D.
,
Snapper
S. B.
.
2012
.
Wiskott-Aldrich syndrome protein (WASP) and N-WASP are critical for peripheral B cell development and function.
Blood
119
:
3966
3974
.
26
Duan
B.
,
Niu
H.
,
Xu
Z.
,
Sharpe
A. H.
,
Croker
B. P.
,
Sobel
E. S.
,
Morel
L.
.
2008
.
Intrafollicular location of marginal zone/CD1d(hi) B cells is associated with autoimmune pathology in a mouse model of lupus.
Lab. Invest.
88
:
1008
1020
.
27
Parekh
V. V.
,
Wilson
M. T.
,
Olivares-Villagómez
D.
,
Singh
A. K.
,
Wu
L.
,
Wang
C. R.
,
Joyce
S.
,
Van Kaer
L.
.
2005
.
Glycolipid antigen induces long-term natural killer T cell anergy in mice.
J. Clin. Invest.
115
:
2572
2583
.
28
Teng
M. W.
,
Yue
S.
,
Sharkey
J.
,
Exley
M. A.
,
Smyth
M. J.
.
2009
.
CD1d activation and blockade: a new antitumor strategy.
J. Immunol.
182
:
3366
3371
.
29
Breslow
J. L.
1996
.
Mouse models of atherosclerosis.
Science
272
:
685
688
.
30
Lim
H. Y.
,
Rutkowski
J. M.
,
Helft
J.
,
Reddy
S. T.
,
Swartz
M. A.
,
Randolph
G. J.
,
Angeli
V.
.
2009
.
Hypercholesterolemic mice exhibit lymphatic vessel dysfunction and degeneration.
Am. J. Pathol.
175
:
1328
1337
.
31
Khoo
L. H.
,
Thiam
C. H.
,
Soh
S. Y.
,
Angeli
V.
.
2015
.
Splenic extrafollicular reactions and BM plasma cells sustain IgM response associated with hypercholesterolemia.
Eur. J. Immunol.
45
:
1300
1312
.
32
Evans
J. G.
,
Chavez-Rueda
K. A.
,
Eddaoudi
A.
,
Meyer-Bahlburg
A.
,
Rawlings
D. J.
,
Ehrenstein
M. R.
,
Mauri
C.
.
2007
.
Novel suppressive function of transitional 2 B cells in experimental arthritis.
J. Immunol.
178
:
7868
7878
.
33
Yankee
T. M.
,
Draves
K. E.
,
Clark
E. A.
.
2005
.
Expression and function of the adaptor protein Gads in murine B cells.
Eur. J. Immunol.
35
:
1184
1192
.
34
Batten
M.
,
Groom
J.
,
Cachero
T. G.
,
Qian
F.
,
Schneider
P.
,
Tschopp
J.
,
Browning
J. L.
,
Mackay
F.
.
2000
.
BAFF mediates survival of peripheral immature B lymphocytes.
J. Exp. Med.
192
:
1453
1466
.
35
Mackay
F.
,
Browning
J. L.
.
2002
.
BAFF: a fundamental survival factor for B cells.
Nat. Rev. Immunol.
2
:
465
475
.
36
Wen
X.
,
Yang
J. Q.
,
Kim
P. J.
,
Singh
R. R.
.
2011
.
Homeostatic regulation of marginal zone B cells by invariant natural killer T cells.
PLoS One
6
:
e26536
.
37
Braun
N. A.
,
Mendez-Fernandez
Y. V.
,
Covarrubias
R.
,
Porcelli
S. A.
,
Savage
P. B.
,
Yagita
H.
,
Van Kaer
L.
,
Major
A. S.
.
2010
.
Development of spontaneous anergy in invariant natural killer T cells in a mouse model of dyslipidemia.
Arterioscler. Thromb. Vasc. Biol.
30
:
1758
1765
.
38
Wakao
H.
,
Kawamoto
H.
,
Sakata
S.
,
Inoue
K.
,
Ogura
A.
,
Wakao
R.
,
Oda
A.
,
Fujita
H.
.
2007
.
A novel mouse model for invariant NKT cell study.
J. Immunol.
179
:
3888
3895
.
39
Davis
H. R.
 Jr.
,
Compton
D. S.
,
Hoos
L.
,
Tetzloff
G.
.
2001
.
Ezetimibe, a potent cholesterol absorption inhibitor, inhibits the development of atherosclerosis in ApoE knockout mice.
Arterioscler. Thromb. Vasc. Biol.
21
:
2032
2038
.
40
Andican
G.
,
Seven
A.
,
Uncu
M.
,
Cantaşdemir
M.
,
Numan
F.
,
Burçak
G.
.
2008
.
Oxidized LDL and anti-oxLDL antibody levels in peripheral atherosclerotic disease.
Scand. J. Clin. Lab. Invest.
68
:
473
478
.
41
Salonen
J. T.
,
Ylä-Herttuala
S.
,
Yamamoto
R.
,
Butler
S.
,
Korpela
H.
,
Salonen
R.
,
Nyyssönen
K.
,
Palinski
W.
,
Witztum
J. L.
.
1992
.
Autoantibody against oxidised LDL and progression of carotid atherosclerosis.
Lancet
339
:
883
887
.
42
Palinski
W.
,
Tangirala
R. K.
,
Miller
E.
,
Young
S. G.
,
Witztum
J. L.
.
1995
.
Increased autoantibody titers against epitopes of oxidized LDL in LDL receptor-deficient mice with increased atherosclerosis.
Arterioscler. Thromb. Vasc. Biol.
15
:
1569
1576
.
43
Tsimikas
S.
,
Willeit
P.
,
Willeit
J.
,
Santer
P.
,
Mayr
M.
,
Xu
Q.
,
Mayr
A.
,
Witztum
J. L.
,
Kiechl
S.
.
2012
.
Oxidation-specific biomarkers, prospective 15-year cardiovascular and stroke outcomes, and net reclassification of cardiovascular events.
J. Am. Coll. Cardiol.
60
:
2218
2229
.
44
Garvy
B. A.
,
Riley
R. L.
.
1994
.
IFN-gamma abrogates IL-7-dependent proliferation in pre-B cells, coinciding with onset of apoptosis.
Immunology
81
:
381
388
.
45
Grawunder
U.
,
Melchers
F.
,
Rolink
A.
.
1993
.
Interferon-gamma arrests proliferation and causes apoptosis in stromal cell/interleukin-7-dependent normal murine pre-B cell lines and clones in vitro, but does not induce differentiation to surface immunoglobulin-positive B cells.
Eur. J. Immunol.
23
:
544
551
.
46
Yoshikawa
H.
,
Nakajima
Y.
,
Tasaka
K.
.
2001
.
IFN-gamma induces the apoptosis of WEHI 279 and normal pre-B cell lines by expressing direct inhibitor of apoptosis protein binding protein with low pI.
J. Immunol.
167
:
2487
2495
.
47
Cerundolo
V.
,
Silk
J. D.
,
Masri
S. H.
,
Salio
M.
.
2009
.
Harnessing invariant NKT cells in vaccination strategies.
Nat. Rev. Immunol.
9
:
28
38
.
48
Wurster
A. L.
,
Rodgers
V. L.
,
White
M. F.
,
Rothstein
T. L.
,
Grusby
M. J.
.
2002
.
Interleukin-4-mediated protection of primary B cells from apoptosis through Stat6-dependent up-regulation of Bcl-xL.
J. Biol. Chem.
277
:
27169
27175
.
49
Sag
D.
,
Krause
P.
,
Hedrick
C. C.
,
Kronenberg
M.
,
Wingender
G.
.
2014
.
IL-10-producing NKT10 cells are a distinct regulatory invariant NKT cell subset.
J. Clin. Invest.
124
:
3725
3740
.
50
Taniguchi
M.
,
Harada
M.
,
Kojo
S.
,
Nakayama
T.
,
Wakao
H.
.
2003
.
The regulatory role of Valpha14 NKT cells in innate and acquired immune response.
Annu. Rev. Immunol.
21
:
483
513
.
51
Li
Y.
,
To
K.
,
Kanellakis
P.
,
Hosseini
H.
,
Deswaerte
V.
,
Tipping
P.
,
Smyth
M. J.
,
Toh
B. H.
,
Bobik
A.
,
Kyaw
T.
.
2015
.
CD4+ natural killer T cells potently augment aortic root atherosclerosis by perforin- and granzyme B-dependent cytotoxicity.
Circ. Res.
116
:
245
254
.
52
Tsimikas
S.
,
Brilakis
E. S.
,
Lennon
R. J.
,
Miller
E. R.
,
Witztum
J. L.
,
McConnell
J. P.
,
Kornman
K. S.
,
Berger
P. B.
.
2007
.
Relationship of IgG and IgM autoantibodies to oxidized low density lipoprotein with coronary artery disease and cardiovascular events.
J. Lipid Res.
48
:
425
433
.
53
Lewis
M. J.
,
Malik
T. H.
,
Ehrenstein
M. R.
,
Boyle
J. J.
,
Botto
M.
,
Haskard
D. O.
.
2009
.
Immunoglobulin M is required for protection against atherosclerosis in low-density lipoprotein receptor-deficient mice.
Circulation
120
:
417
426
.
54
Smook
M. L.
,
van Leeuwen
M.
,
Heeringa
P.
,
Damoiseaux
J. G.
,
Theunissen
R.
,
Daemen
M. J.
,
Lutgens
E.
,
Tervaert
J. W.
.
2008
.
Anti-oxLDL antibody isotype levels, as potential markers for progressive atherosclerosis in APOE and APOECD40L mice.
Clin. Exp. Immunol.
154
:
264
269
.
55
To
K.
,
Agrotis
A.
,
Besra
G.
,
Bobik
A.
,
Toh
B. H.
.
2009
.
NKT cell subsets mediate differential proatherogenic effects in ApoE-/- mice.
Arterioscler. Thromb. Vasc. Biol.
29
:
671
677
.
56
Major
A. S.
,
Wilson
M. T.
,
McCaleb
J. L.
,
Ru Su
Y.
,
Stanic
A. K.
,
Joyce
S.
,
Van Kaer
L.
,
Fazio
S.
,
Linton
M. F.
.
2004
.
Quantitative and qualitative differences in proatherogenic NKT cells in apolipoprotein E-deficient mice.
Arterioscler. Thromb. Vasc. Biol.
24
:
2351
2357
.
57
Nakai
Y.
,
Iwabuchi
K.
,
Fujii
S.
,
Ishimori
N.
,
Dashtsoodol
N.
,
Watano
K.
,
Mishima
T.
,
Iwabuchi
C.
,
Tanaka
S.
,
Bezbradica
J. S.
, et al
.
2004
.
Natural killer T cells accelerate atherogenesis in mice.
Blood
104
:
2051
2059
.
58
Tupin
E.
,
Nicoletti
A.
,
Elhage
R.
,
Rudling
M.
,
Ljunggren
H. G.
,
Hansson
G. K.
,
Berne
G. P.
.
2004
.
CD1d-dependent activation of NKT cells aggravates atherosclerosis.
J. Exp. Med.
199
:
417
422
.

The authors have no financial conflicts of interest.

Supplementary data