CD36 Is Differentially Expressed on B Cell Subsets during ...

The Journal of Immunology

CD36 Is Differentially Expressed on B Cell Subsets during Development and in Responses to Antigen1 Woong-Jai Won,*† Martin F. Bachmann,‡ and John F. Kearney2*† Of a number of mAbs made by immunization with sort-purified marginal zone (MZ) B cells, one was shown to recognize the mouse scavenger receptor CD36. Although CD36 is expressed by most resting MZ B cells and not by follicular and B1 B cells, it is rapidly induced on follicular B cells in vitro following TLR and CD40 stimulation. In response to T-independent and T-dependent Ag challenge, we found that CD36 was expressed on IgMⴙ plasma cells, but down-regulated on isotype-switched plasma cells in vivo. Although development, localization, and phenotype of MZ B cells in CD36ⴚ/ⴚ mice appeared normal, there was a minor block in the transitional stages of mature B cell development. In both primary and secondary Ab responses to heat-killed Streptococcus pneumoniae (R36A strain), both phosphoryl choline (PC)-specific IgM and IgG levels in CD36ⴚ/ⴚ mice were slightly reduced compared with wild-type mice. In addition, mice deficient in both TLR2 and CD36 produced significantly reduced levels of anti-PC IgG titers than those of single gene-deficient mice, suggesting that they may cooperate in an anti-PC Ab response. Collectively, these results show that CD36 does not affect the development of B cells, but modulates both primary and secondary anti-PC Ab responses during S. pneumoniae infection similarly to TLR2. The Journal of Immunology, 2008, 180: 230 –237.

M

ature B cells in mammals are divided into three major B cell subsets, marginal zone (MZ),3 follicular (FO), and B1 B cells, depending on surface phenotypes, location, and immunological functions (1– 4). Although MZ and B1 B cells represent subpopulations of naive mature B cells, they play an important role in early innate-like and primary Ab responses (1, 2, 5). In particular, they respond quickly to T-independent (TI) Ags, resulting in a rapid production of large numbers of short-lived plasma cells (6, 7). Previously, our laboratory identified two surface markers, CD9 and FCRL5 (Fc receptor-like 5), which help distinguish MZ B and B1 cells from FO B cells in mice (8, 9). Recently, we and others showed by different approaches that CD36 is expressed on resting MZ B cells, but not the majority of FO B cells (10, 11). CD36 is a Type BI scavenger receptor family member, which is expressed by a variety of cell types including endothelium, erythrocytes, platelets, adipocytes, dendritic cells, neutrophils, monocytes/macrophages, microglia, and muscle cells (12–20). CD36 binds multiple ligands including native and modified low density lipoproteins, oxidized phosphoryl choline (PC) epitope, anionic phospholipids, collagens, thrombospondin-1, and fibrillar ␤-amy-

loids (17, 21–27). A role for CD36 has also been suggested in the pathogenesis of Alzheimer’s disease and atherosclerosis (18, 25, 28 –30). Recent studies showed that CD36 enhances TLR2 signaling to induce cytokine secretion by macrophages, although the exact mechanism was not established (31, 32). Because CD36 is expressed uniquely on freshly isolated mature MZ B cells and is induced to high levels on other mature B cell subsets after stimulation in vitro, we investigated the role of CD36 with respect to B cell development and function and examined a possible cooperation between CD36 and TLR2 in Ab responses.

Materials and Methods Mice CD36⫺/⫺ mice were obtained from Dr. Maria Febbraio at the Lerner Research Institute (Cleveland, OH). T15 H chain knock-in and M167Id Ig transgenic mice were obtained from Dr. Jim Kenny at National Institutes of Health (Bethesda, MD), and BALB/cJ, C57BL/6J, and TLR2⫺/⫺ mice were purchased from The Jackson Laboratories (33). CD36⫺/⫺ mice were backcrossed at least to the sixth generation to C57BL/6 mice. All mice were bred, and maintained in our animal facilities at the University of Alabama at Birmingham (UAB). All studies and procedures were approved by the UAB Institutional Animal Care and Use Committee (IACUC).

Flow cytometry *Division of Developmental and Clinical Immunology and †Department of Microbiology, University of Alabama at Birmingham, Bikminghan AL 35294; and ‡Cytos Biotechnology AG, Wagistr 25, Schlieren-Zurich, Switzerland Received for publication August 7, 2007. Accepted for publication October 3, 2007. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported in part by National Institutes of Health Grants CA13148 and AI14782. 2 Address correspondence and reprint requests to Dr. John F. Kearney, Division of Developmental and Clinical Immunology, Department of Microbiology, SHEL 410, 1825 University Boulevard, University of Alabama, Birmingham, AL 35294-2182. E-mail address: [email protected] 3

Abbreviations used in this paper: MZ, marginal zone; FO, follicular; PEC, peritoneal cavity; BM, bone marrow; DC, dendritic cell; PLC, plasma cell; GC, germinal center; WT, wild type; KO, knock-out; TG, transgenic; PC, phosphoryl choline; PEC, peritoneal cavity. Copyright © 2007 by The American Association of Immunologists, Inc. 0022-1767/07/$2.00

www.jimmunol.org

All Abs were purchased from BD Biosciences unless specified. Spleen, peritoneal cavity (PEC), peripheral blood, and bone marrow (BM) cells were prepared as single cell suspensions after RBC lysis, and stained with FITC conjugated anti-mouse CD1d, CD19, CD23, B220, IgM (SBA), IgG1 (SBA), TLR2 (eBioscience), peanut agglutinin (Jackson Immunoresearch Laboratories); PE conjugated anti-mouse CD5, CD23, CD38, syndecan1, IgD (SBA); biotin conjugated anti-mouse syndecan1; allophycocyaninconjugated AA4.1 (eBioscience), Mac1. Alexa 647 and Alexa 488 conjugated CD9 and CD36, PE conjugated CD21, and biotinylated CD36 (MZ1) were made in our laboratory, according to the manufacturer’s instructions (Invitrogen Life Technologies and Pierce). FITC conjugated PC-Dextran and PE conjugated M167 Id mAb were provided by Dr. Louis Rezanka at the National Institute on Aging (Baltimore, MD). Biotinylated mAbs were incubated with streptavidin-allophycocyanin as a secondary reagent. Cells were analyzed using FACSCalibur flow cytometer and plotted with WinMDI (Scripps Institute) software.

Cell sorting and semi-quantitative RT-PCR To examine mRNA transcripts by mouse B cell subsets, total spleen B cells obtained after RBC lysis were stained with FITC-conjugated anti-CD23


FIGURE 1. Unique expression of CD36 by MZ B cells. A, Surface staining of mature B cells with anti-CD36 mAb (MZ1 clone). Spleen cells from BALB/cJ mice were stained with FITC ␣-CD23, PE ␣-CD21, biotinylated ␣-CD36. PEC-exudate cells were stained with FITC ␣-B220, PE ␣-CD5, and biotinylated ␣-CD36, then developed with SA-allophycocyanin. Gray filled histograms indicate isotype-control staining. B, Expression of CD36 mRNA transcripts by splenic mature B cells. mRNA levels of CD36 from MZ and FO B cells were detected by semi-quantitative RTPCR. As a positive control, ␤-actin PCR was amplified with the same number of cycles in parallel. C, Immunohistological staining of CD36 in mouse spleen section. Frozen spleen sections from BALB/cJ mouse were stained with MOMA1, then developed with 7-amino-4-methyl coumarin3-acetic acid goat-anti-rat IgG (Blue). Sections were blocked with normal rat serum, then stained with Alexa488 MZ1 (Green) or tetramethyl rhodamine iso-thiocyanate IgM (Red). Each staining histogram and immunohistologic photograph is a representative of at least six mice.

and PE-conjugated anti-CD21, and sorted using a MoFlo sorter or FACSAria sorter (BD Biosciences). mRNA and cDNA of each B cell subset were prepared according to the manufacturer’s protocol (Invitrogen Life Technologies). RT-PCR of serially diluted cDNA were conducted by the following conditions: CD36 5⬘ primer (GGC ACC ACT GTG TAC AGA CAG), CD36 3⬘ primer (GGA AAG GAG GCT GCG TCT GTG C), Taq polymerase (1 unit/reaction, Invitrogen Life Technologies), OP#4 buffer (Stratagene), preheated at 94°C for 3 min, (94°C 1 min, 54°C 1 min, 72°C 1 min for 35 cycles), 72°C 10 min, and then soaked at 4°C. RT-PCR of mouse TLR1, 2, and 6 was followed as previously described at other laboratories (34).

231

FIGURE 2. Rapid induction of CD36 on B cells. A, Rapid induction of CD36 on both MZ and FO B cells. FACS-sorted MZ and FO B cells were incubated with B cell activating agents, LPS (10 ␮g/ml), or anti-CD40 (1 ␮g/ml), for 1–3 days, and CD36 was detected by flow cytometry. B, Induction of CD36 on PEC B1 cells. FACS-sorted PEC B1a (CD5⫹B220lowMac1⫹) or B1b (CD5⫺B220lowMac1⫹) cells were incubated with LPS for 1–3 days. Gray-filled histogram indicates isotype-control staining.

In vitro mitogenic stimulation The sorted MZ and FO B cells were cultured for 1 to 3 days with LPS (Sigma-Aldrich), anti-CD40 (BD Biosciences), or synthetic TLR2 ligands such as FSL1 and Pam3Cys-SKKKK (EMC microcollections GmbH). CD36 expression was determined by flow cytometry using our Alexa 647conjugated anti-mouse CD36 mAb (MZ1).

Cytokine measurement Sorted cells were stimulated with various activation agents such as LPS, Pam3Cys-SKKKK, and FSL-1, and supernatants from cultures were collected at day 2. Levels of various cytokines such as TNF-␣, IL-2, IL-4, IL-6, IL-10, IL-12 (p40 and p70), and IFN-␥, were measured using BD cytokine-bead assay kits with a FACSArray flow cytometer and calculated using the manufacturer’s software (BD Biosciences).

Immunohistochemistry Portions of mouse spleen and small intestines were embedded in OCT (Optimal Cutting Temperature) compound (Sakura Finetek). The frozen OCT-embedded tissues were cut to a thickness of 4 ␮m, put on the glass slides, fixed in acetone, and stored at ⫺80°C. Slides were first blocked with 10% horse serum and stained with Alexa 488-conjugated anti-mouse CD36; tetra-methyl rhodamine iso-thiocyanate-conjugated goat anti-mouse IgM (SBA); supernatant of MOMA1 hybridoma. Anti-rat IgG 7-amino-4methyl coumarin-3-acetic acid (Jackson Immunoresearch Laboratories) was added as a secondary reagent for MOMA1 staining.

232

CD36 EXPRESSION BY B CELL SUBSETS

FIGURE 3. Differential expression of CD36 by activated B cells and plasma cells in vivo. A, Expression of CD36 by natural plasma cells. Spleen cells from unimmunized C57BL/6J mice were stained with PE Syndecan1, Alexa-647 MZ1, and either FITC PC-Dextran or FITC IgM. B, Expression of CD36 by PC-specific plasma cells induced by Streptococcus pneumoniae R36A strain. C57BL/6J mice were injected i.v. with 108 R36A and the expression of CD36 was examined at day 3. *ND (not detected). C, Expression of CD36 by germinal center B cells. Spleen cells at day 7 after SRBC injection were stained. Germinal center cells were defined, as previously described (9, 46). D, Expression of CD36 by isotype-switched plasma cells. Top panel shows surface staining and lower panel shows an intracytoplasmic staining of spleen cells of SRBC immunized mice at day 7.

Bacteria preparation and immunization Nonencapsulated Streptococcus pneumoniae R36A strain was plated on blood agar plates overnight. Separate colonies were picked, transferred, and grown in Todd Hewitt Broth plus 0.5% yeast extract (BD Biosciences) until OD reaches 0.5 at 420 nm. The cells were harvested and washed with PBS three times. The cells were heat-inactivated at 60°C for 1 h and pepsin-treated (7). To induce anti-PC Ab response, 1 ⫻ 108 heat-killed Streptococcus pneumoniae R36A strain were injected i.v. or i.p. into CD36⫺/⫺ and control littermate mice. To induce soluble TI Ag responses, TNP-LPS (TI-1 Ag), and TNP-Ficoll (TI-2 Ag) (Biosearch Technologies) were tested, except that sera titers were measured by ELISA using alkaline phosphatase-conjugated Abs, as described by others (35). In response to SRBC, C57BL/6J mice were injected i.v. with 200 ␮l of packed SRBC (Vector Laboratories), as previously described in our laboratory (8).

Dendritic cell (DC)-B cell coculture BM-derived DC were generated as previously described in our laboratory (36). DCs were primed with R36A at the ratio of 1:10 (DC:R36A) in 5% complete RPMI 1640 medium at 37°C for 4 h. Spleen B cells were isolated from CD36⫹/⫺ or CD36⫺/⫺ M167 Ig transgenic (TG) mice by magnetic sorting with anti-mouse CD43 beads, and incubated with primed DC at the ratio of 106 : 2 ⫻ 105 cells (B cells:DCs) for 3– 8 days. Generation of M167Id⫹ plasma cells (PLC) in the coculture was calculated as Id⫹ PLC percentage ⫻ total live cell numbers collected for same interval by FACSCalibur flow cytometer.

ELISA Sera were prepared from the blood of immunized mice and the concentrations of anti-PC of either IgM or IgG isotypes were determined by ELISA, as previously described in our laboratory (37). Purified hybridoma Abs of

BH8, 32-17-1, and 59.6C5.5 were used as standards for anti-PC IgM, IgG, and IgG3 isotype determination, respectively.

Statistical analysis Concentrations of serum Ig of different isotypes and anti-PC IgM and IgG Ab were expressed as the arithmetic mean of individual serum samples from four to ten mice ⫾ SD. Concentrations of IgM and cytokine secreted in sorted B cell culture supernatants were expressed as the arithmetic mean of triplicate cultures from pooled spleens from three mice ⫾ SD. Levels of significance of the differences between groups were determined by Student’s t test. A value of p ⬍ 0.05 was considered statistically significant.

Results CD36 is a specific marker of mouse MZ B cells Of three mAbs previously generated by immunizing a rat with sort-purified MZ B cells in our laboratory (38), we found that one reacted with mouse CD36 by using ␣-virus expression cloning (10, 39). As shown in Fig. 1A, CD36 is expressed on most MZ B cells, but on only very small numbers of freshly isolated FO B cells and B1 cells in spleen and peritoneal cavity. CD36 mRNA transcripts are also exclusively expressed in MZ B cells, but not in FO B cells (Fig. 1B). The expression of CD36 was higher on MZ B cells outside of the metallophilic macrophage layer in the MZ (Fig. 1C). Although CD36 is widely expressed in other cell types such as macrophages, endothelial cells, platelets, and erythrocytes, it is clear that CD36 is differentially expressed on freshly isolated mature B cell subsets.


233

CD36 is induced rapidly on mature B cells by TLR and CD40 stimulation Although CD36 is not expressed by the majority of freshly isolated FO B and B1 cells, as shown in Fig. 2A, CD36 is rapidly induced on FO B cells after stimulation with LPS and anti-CD40, peaking at day 3 where it is expressed at a level comparable to freshly isolated MZ B cells. In contrast, only a portion of PEC B1 cells were induced to express CD36 following LPS stimulation by day 3 (Fig. 2B). These results show that CD36 is induced rapidly on mature spleen B cells followed by multiple activation agents, but not on most B1 cells. CD36 is expressed on IgM⫹ plasma cells but down-regulated on IgG isotype-switched plasma cells in vivo We next immunized C57BL/6 mice with heat-killed S. pneumoniae (R36A) and SRBC as particulate TI and T-dependent Ags, respectively, and examined the expression of CD36 by plasma cells induced in each response. In un-immunized mice, natural IgM⫹ PLC did not express CD36 (Fig. 3A). In response to R36A, all syndecan1-positive PC-binding cells in the spleen expressed CD36 (Fig. 3B, top panel). The majority of the total IgM⫹ PLC induced by R36A also expressed CD36, but in a biphasic manner (Fig. 3B, middle panel). In contrast, most IgG1⫹ plasma cells induced by SRBC immunization expressed CD36 at much lower levels, although IgM⫹ plasma cells in response to SRBC expressed CD36 at a level comparable to IgM⫹ plasma cells induced by R36A (Fig. 3D). Most germinal center (GC) B cells induced by SRBC weakly expressed the CD36 except for a small population (Fig. 3C). These results show that CD36 is expressed at high levels on short-lived IgM⫹ plasma cells compared with isotype-switched plasma cells, which appear to slowly lose expression as they terminally differentiate. Lack of CD36 has minimal effects on mature B cell development during transitional B cell stages in BM and spleen Examination of the development and the localization of FO and MZ B cells in the spleens of CD36⫺/⫺ mice revealed only a slight reduction in the total numbers of mature B cells and MZ precursors in CD36⫺/⫺ mice. Splenic and PEC B1 cell compartments were comparable between wild-type (WT) and knock-out (KO) mice (data not shown). However, spleen and BM transitional B cells were slightly increased compared with WT mice (Fig. 4), suggesting that there may be a minor block in the transition of immature B cells into mature B cell compartments. Other cell compartments in various tissues such as peripheral blood, mesenteric lymph node, and Peyer’s patches, also appeared normal (data not shown). Absence of CD36 affects plasma cell generation and humoral Ab response to particulate TI-2 Ags Resting serum Ig levels of CD36⫺/⫺ mice showed small differences in several IgG isotypes (IgG1, IgG2a, IgG2b, and IgG2c), but not IgM and IgG3 isotypes, compared with WT mice (Fig. 5A). Because MZ B cells are important for bacterial TI-2 Ab response, we examined conventional Ab response of CD36⫺/⫺ mice with TNP-LPS and TNP-Ficoll as TI-1 and TI-2 soluble Ags, respectively. Both anti-TNP titers to both Ags were comparable between CD36WT and CD36KO mouse groups (data not shown). Next, we challenged CD36⫺/⫺ mice with heat-killed S. pneumoniae as a particulate Tl-2 Ag and examined the generation of PC-binding plasma cells and anti-PC titers. As shown in Fig. 5B, total anti-PC IgM levels were reduced to ⬃65% of the WT group at day 7 ( p ⬍ 0.05), which is the usual peak of the anti-PC IgM titer (Fig. 5B). Similarly, CD36⫺/⫺ mice also showed lower anti-PC IgG levels compared with WT mice, although the p value is not statistically

FIGURE 4. B cell development in CD36⫺/⫺ mice. To examine a possible role of CD36 in B cell development, B cell compartments of WT CD36⫹/⫺ mice were compared with that of CD36⫺/⫺ mice. A, FACS plot of transitional and mature B cell compartment in spleens and BM of CD36⫺/⫺ mice. The numbers in FACS plots indicate the values of a representative individual mouse. B, Comparison of total cell numbers of each B cell subset in CD36⫺/⫺ mouse spleen. †, MZP (marginal zone precursor) indicates a population of CD21highCD23highCD1dhigh. TR (transitional), SP (spleen), BM (bone marrow). ⴱ, p ⬍ 0.01, ⴱⴱ, p ⬍ 0.05.

significant. We also examined the production of PC-binding PLC at day 3 after R36A challenge. As shown in Fig. 5C, the frequency and actual number of PC-specific PLC in CD36⫺/⫺ mice were ⬃2-fold lower than in wild-type (WT) mice, correlating with serum anti-PC levels. To examine a secondary response, R36A-immunized mice were rechallenged at day 14 with the same dose of bacteria, and anti-PC titers were compared. Both anti-PC IgM and anti-PC IgG titers in CD36⫺/⫺ mice were consistently lower (Fig. 5D). We also found that a majority of anti-PC IgG isotypes was IgG3 (data not shown), which also showed a reduction to the same

234

CD36 EXPRESSION BY B CELL SUBSETS

FIGURE 5. Anti-PC Ab response and plasma cell generation of CD36⫺/⫺ mice after TI-2 bacterial challenge. A, Serum Ig levels of different isotypes in un-immunized CD36⫺/⫺ mice. All mice were 8 –12 wk-old females. ⴱ, p ⬍ 0.05. B, Anti-PC IgM and IgG titers of CD36⫺/⫺ mice. CD36⫹/⫺ and CD36⫺/⫺ mice were immunized i.p. with 108 R36A, and anti-PC titers were measured at different time points. Levels of anti-PC IgM and IgG were calculated by anti-PC IgM (BH8) and IgG1 (32-17-1) mAbs as standards. ⴱ, p ⬍ 0.05. The graph shows the results from the total pool of mice used in three independent experiments. The sample number at day 28 is three per group. C, PC-specific plasma cell generation of CD36⫺/⫺ mice. Each mouse group was injected i.p. with 108 R36A, and the number of PC-specific plasma cells was compared between the WT and KO groups at day 4. The graphs are representative of the plots of three mice and the numbers inside contour plots indicate the percentage of a representative mouse. ⴱ, p ⬍ 0.05. D, Anti-PC secondary responses of CD36⫺/⫺ mice. The upper panel shows anti-PC IgM responses of CD36⫹/⫺ and CD36⫺/⫺ mice and the lower panel shows anti-PC IgG responses. Each mouse group was rechallenged i.p. with the same dose of R36A at 14 days after the first injection.

degree. This result shows that lack of CD36 has marginal effects on anti-PC IgM and IgG levels in response to the bacterial TI Ag associated with S. pneumoniae. Defective anti-PC Ab response is due to the absence of CD36 on B cells To examine the function of CD36 in B cells, we sorted MZ and FO B cells from CD36⫺/⫺ and WT mice and compared several in vitro activities, such as proliferation, plasma cell differentiation, cytokine production by mitogen stimulation, including TLR ligands, anti-CD40, IL-4, and anti-IgM or in combination. All of these responses were comparable between CD36⫺/⫺ and WT B cells (data not shown). To check whether the defective anti-PC response observed in CD36⫺/⫺ mice is intrinsic to B cells, we studied Agspecific plasma cell generation ex vivo in response to Ag-pulsed DC (36). We prepared DC derived from CD36⫹/⫺ or CD36⫺/⫺ BM, primed them with heat-killed R36A, and coincubated them with B cells from either CD36⫹/⫺ or CD36⫺/⫺ M167 Ig TG mouse spleen cells. As shown in Fig. 6, CD36⫺/⫺ M167 TG B

cells showed a significant decrease in plasma cell generation as well as anti-PC IgM production ( p ⬍ 0.05), compared with the WT counterpart, irrespective of whether WT or CD36⫺/⫺ DCs were used. Neither CD36⫺/⫺ macrophages, nor supernatants from CD36⫺/⫺ DC culture, showed any difference in their ability to induce M167Id⫹ plasma cell generation (data not shown). These results suggest that the mild defects in plasma cell generation of Ag-specific plasma cells were regulated by the expression of CD36 on B cells. CD36 cooperates with TLR2 in the anti-PC IgG Ab response Recent studies showed that CD36 enhances TLR2 signaling in the cytokine response of activated macrophages (32). TLR2 forms dimers with TLR1, TLR6, or TLR2 itself, which give different specificities to ligands (40). We tested the mRNA expression of TLR2 and other dimer partners in mouse B cell subsets. RNA transcripts of TLR1, 2, and 6 were detected in both MZ and FO B cells, but MZ B cells express ⬃5 times more mRNA of each than FO B cells (Fig. 7A). TLR2 protein level in MZ B cells is slightly


235

FIGURE 6. Anti-PC Ab production and plasma cell generation is modulated by CD36 on B cells. A, Anti-PC Ab responses in DC-B cell coculture system. Spleen B cells from M167 Id TG mice with either CD36⫹/⫺ or CD36⫺/⫺ background and DC derived from CD36⫹/⫺ or CD36⫺/⫺ mouse BM were cocultured for 3– 8 days, and the generation of PC-specific plasma cells were measured by flow cytometry. B, Anti-PC IgM titer of coculture supernatants. Anti-PC IgM titers in the culture supernatants were measured by ELISA, as described in Materials and Methods. The graph shows a representative of two independent experiments, both showing the same results. ⴱ, p ⬍ 0.01; ⴱⴱ, p ⬍ 0.05.

higher than that of FO B cells, which correlated with RNA message level, although the expression level was low (Fig. 7B). After activation of B cells from CD36⫺/⫺ mice with chemically synthesized TLR2 ligands, proliferation, plasma cell generation, and cytokine production of CD36⫺/⫺ B cells were similar to WT B cells (Fig. 7C, data not shown). However, as expected, TLR2⫺/⫺ B cells showed severe defects in proliferation (Fig. 7C), cytokine expression, and plasma cell generation (data not shown). This result confirms that B cells are strongly dependent on TLR2 in in vitro responses to these ligands, but not on CD36 expression. To test whether CD36 and TLR2 cooperate in the anti-PC Ab response in vivo, we compared the anti-PC humoral response to S. pneumoniae in CD36 TLR2 doubly deficient and singly deficient mice with WT mice. As shown in Fig. 8, anti-PC IgM levels of doubly deficient mice were similar to those of TLR2 singly deficient mice compared with WT or CD36 singly deficient mice. However, anti-PC IgG titers of CD36 TLR2 double KO mice were much lower than those from either TLR2 or CD36 singly deficient

mice ( p ⬍ 0.05), suggesting that CD36 and TLR2 cooperate in the anti-PC IgG response to S. pneumoniae (Fig. 8).

Discussion We previously showed that CD36 is expressed by freshly isolated MZ B cells, but only by a very few FO and B1 cells (10). The differential expression of CD36 by B cell subsets provides another useful marker for discrimination of MZ B cells along with CD9 and FCRL5 (8, 9). However, because CD36 is rapidly induced in FO B cells by stimulation by TLR ligands and ␣-CD40, it may represent an early activation marker of B cells, similar to CD25 and CD69. Of interest, PEC B1a cells expressed CD36 much less than B1b cells following LPS stimulation. In addition to the minor populations of FO B cells that express CD36, it was also expressed at a slightly lower level on most MZ precursors and a subset of transitional B cells, suggesting that these cells may be activated. CD36 expression along with other MZ B cell markers such as CD9 and FcRL5 may be indicative of activation signals involved in the

FIGURE 7. Expression of TLR2 by B cells and function of TLR2 in B cell activation. A, mRNA transcript levels of TLR1, 2, and 6 by MZ and FO B cells. Semi-quantitative RTPCR of each TLR1, 2, and 6 was performed as described (34). ␤-actin was used as a positive control. ⴱ, Loading was less compared with other wells. B, TLR2 surface expression by spleen B cell subsets. Spleen cells from C57BL/6J mice were stained with FITC TLR2 (clone 6C2), CD21-PE, or biotinylated CD23, then developed with SA-allophycocyanin. †, MZP(CD23highCD21high). C, Proliferation of CD36⫺/⫺ and TLR2⫺/⫺ B cells induced by TLR2 ligands. Sorted CD36⫺/⫺ and TLR2⫺/⫺ MZ B cells along with WT MZ B cells were stimulated with LPS and synthetic TLR2 ligands for 2 days and labeled with 1␮Ci [3H]thymidine in 96-well microplate. The incorporation of radioactive thymidine is presented by a gamma scintillation counter as a proliferation index. Pam3 (Pam3Cys-SKKKK).

236

FIGURE 8. TLR2 and CD36 cooperate in anti-PC Ab responses. To test for cooperation between CD36 and TLR2 in anti-PC responses, TLR2⫺/⫺ and CD36⫺/⫺ mice were bred, and four different genotypes of mice were challenged with 108 heat-killed R36A strain i.p. Tested mouse numbers of each group are five, three, four, and five (WT, CD36KO, TLR2KO, and DKO), respectively. The graph shows a representative of two independent experiments, both showing the same results. ⴱ, p ⬍ 0.05.

positive selection of mature B cell precursors from transitional B cells. One interesting observation is the differential expression of CD36 on IgM⫹ PLC from germinal center B cells and Ig-isotype switched PLC derived from SRBC immunization. Additionally, only some IgM⫹ plasma cells expressed CD36. However, all PCbinding PLC express CD36, perhaps indicating the Ag-driven induction of CD36 on B cells in early immune responses. Although there was a strong induction of CD36 on FO B cells by in vitro stimulation, only a small fraction of GC B cells expressed CD36. Low expression of CD36 on GC B cells indicates that it may play a minor role in T-dependent Ab responses. Previously, Konig et al. (41) showed that CD36 expression is strictly regulated by Oct2 in B cells. Thus, it will also be of interest to check Oct2 expression by B cell subsets. Collectively, our results suggest that CD36 is induced at an early stage of B cell activation and its induction in vivo on B cells results from Ag-specific activation. In CD36⫺/⫺ mice, the localization and compartment of mature B cells are minimally affected, while the number of mature B cells is slightly reduced and TR B cell compartments is increased. Although the unique expression of CD36 by MZ B cells predicted a specific role in MZ B cell development, CD36 appears to affect both MZ and FO B cell compartments. This minor difference in B cell development may be due to a small blockage at the stage between transitional and mature B cells. We also tested Ig TG mouse model system such as M167 (anti-PC) Ig TG, J558 (anti␣-1,3 dextran) Ig TG, and T15 (anti-PC) Ig H chain knock-in mice crossed with a CD36⫺/⫺ background. The development of MZ B and Id⫹ cells in each TG line on a CD36⫺/⫺ background also showed only marginal differences in plasma cell and specific Ab production, similarly to WT mice (data not shown). TLR2 recognizes peptidoglycan, lipopeptide, and lipoteichoic acid (LTA) of Gram-positive bacteria (40, 42). Recent reports showed that CD36 enhances TLR2 signaling in cytokine secretion of macrophages (32). However, in our study, stimulation by TLR2 ligands did not show any prominent differences between CD36WT and CD36⫺/⫺ B cells with respect to proliferation, plasma cell

CD36 EXPRESSION BY B CELL SUBSETS generation, IgM secretion, and cytokine secretion in vitro, similar to other studies (13). MZ B cells and B1 cells are responsible for TI Ab response (3, 7, 43). We examined TI Ab responses with conventional soluble Ags such as TNP-LPS and TNP-Ficoll, but anti-TNP titers between CD36⫺/⫺ and WT mouse groups were comparable. Corcoran et al. (13) also reported Ab responses of CD36⫺/⫺ mice, but their studies targeted the anti-Leishmania Ab response, which also showed minor differences between WT and KO mice. Our studies targeted Streptococcus pneumoniae, a Gram-positive bacterium that expresses TLR2 ligands. Different results between anti-PC, conventional TI Ab responses, and anti-Leishimania response may be due to the difference in Ag structure and the degree of cooperation with other TLR ligands on bacterial surfaces. It is of interest that TLR2 and CD36 cooperate to fully induce anti-PC IgG response at early time points. This result suggests that they may physically interact with each other on B cells to directly regulate Ab production or on other cells, such as macrophages or DCs, to indirectly regulate secretion of cytokines, thus affecting IgG switching. However, the mechanism for cooperation remains unknown at this point, and further studies will be needed (44). Anti-PC IgG levels are lower than anti-PC IgM levels, but both isotypes have been shown to protect against S. pneumoniae (45). Collectively, we showed that after R36A challenge, both anti-PC IgM and anti-PC IgG levels in CD36⫺/⫺ mice were lower than CD36WT mice, which suggest that both responses have a synergistic effect in anti-PC production to S. pneumoniae. We showed that the reduction in PC-specific plasma cell (PLC) generation in vitro by CD36⫺/⫺ B cells is B cell intrinsic. Culture of M167Id⫹ MZ and FO B cells with DCs also showed similar results, although only M167Id⫹ MZB cells gave strong Ab production, and as shown previously, Id⫹ FO B cells did not (data not shown) (36). Given the wide range of CD36 ligands, including its function as a receptor for fatty acids, the role of CD36 on B cells may not have been revealed in the experiments we have conducted. Nevertheless, its differential expression on B cell surfaces makes it another B cell subset marker and its rapid up-regulation after activation may make it useful as a B cell activation marker.

Acknowledgments We acknowledge Lisa Jia and Xiaoying Liu for technical assistance, Jeremy Foote for testing J558L Ig TG mice, Dr. Ute Saunders for help in the dendritic cell culture system, Dr. Maria Febbraio for providing CD36⫺/⫺ mice, Dr. Francis Lund for help with cytokine measurement of B cell culture, and Dr. Inho Park and Dr. Nicholas Kin for discussion and reading the manuscript.

Disclosures The authors have no financial conflict of interest.

References 1. Martin, F., and J. F. Kearney. 2001. B1 cells: similarities and differences with other B cell subsets. Curr. Opin. Immunol. 13: 195–201. 2. Martin, F., and J. F. Kearney. 2002. Marginal-zone B cells. Nat. Rev. Immunol. 2: 323–335. 3. Oliver, A. M., F. Martin, G. L. Gartland, R. H. Carter, and J. F. Kearney. 1997. Marginal zone B cells exhibit unique activation, proliferative and immunoglobulin secretory responses. Eur. J. Immunol. 27: 2366 –2374. 4. Pillai, S., A. Cariappa, and S. T. Moran. 2005. Marginal zone B cells. Annu. Rev. Immunol. 23: 161–196. 5. Martin, F., and J. F. Kearney. 2000. B-cell subsets and the mature preimmune repertoire. Marginal zone and B1 B cells as part of a “natural immune memory.” Immunol. Rev. 175: 70 –79. 6. Oliver, A. M., F. Martin, and J. F. Kearney. 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. 7. Martin, F., A. M. Oliver, and J. F. Kearney. 2001. Marginal zone and B1 B cells unite in the early response against T-independent blood-borne particulate antigens. Immunity 14: 617– 629.

The Journal of Immunology 8. Won, W. J., and J. F. Kearney. 2002. CD9 is a unique marker for marginal zone B cells, B1 cells, and plasma cells in mice. J. Immunol. 168: 5605–5611. 9. Won, W. J., J. B. Foote, M. R. Odom, J. Pan, J. F. Kearney, and R. S. Davis. 2006. Fc receptor homolog 3 is a novel immunoregulatory marker of marginal zone and B1 B cells. J. Immunol. 177: 6815. 10. Won, W. J., L. Ute, M. F. Bachmann, and J. F. Kearney. 2005. Expression of CD36 by mouse marginal zone B cells. FASEB J. A358. 11. Zhang, P., W. Li, Y. Wang, L. Hou, Y. Xing, H. Qin, J. Wang, Y. Liang, and H. Han. 2007. Identification of CD36 as a new surface marker of marginal zone B cells by transcriptomic analysis. Mol. Immunol. 44: 332–337. 12. Gordon, S. 2002. Pattern recognition receptors: doubling up for the innate immune response. Cell 111: 927–930. 13. Corcoran, L., D. Vremec, M. Febbraio, T. Baldwin, and E. Handman. 2002. Differential regulation of CD36 expression in antigen-presenting cells: Oct-2 dependence in B lymphocytes but not dendritic cells or macrophages. Int. Immunol. 14: 1099 –1104. 14. Kuniyasu, A., S. Hayashi, and H. Nakayama. 2002. Adipocytes recognize and degrade oxidized low density lipoprotein through CD36. Biochem. Biophys. Res. Commun. 295: 319 –323. 15. Husemann, J., and S. C. Silverstein. 2001. Expression of scavenger receptor class B, type I, by astrocytes and vascular smooth muscle cells in normal adult mouse and human brain and in Alzheimer’s disease brain. Am. J. Pathol. 158: 825– 832. 16. Suchard, S. J., L. A. Boxer, and V. M. Dixit. 1991. Activation of human neutrophils increases thrombospondin receptor expression. J. Immunol. 147: 651– 659. 17. Kehrel, B., S. Wierwille, K. J. Clemetson, O. Anders, M. Steiner, C. G. Knight, R. W. Farndale, M. Okuma, and M. J. Barnes. 1998. Glycoprotein VI is a major collagen receptor for platelet activation: it recognizes the platelet-activating quaternary structure of collagen, whereas CD36, glycoprotein IIb/IIIa, and von Willebrand factor do not. Blood 91: 491– 499. 18. Husemann, J., J. D. Loike, R. Anankov, M. Febbraio, and S. C. Silverstein. 2002. Scavenger receptors in neurobiology and neuropathology: their role on microglia and other cells of the nervous system. Glia 40: 195–205. 19. Dawson, D. W., S. F. Pearce, R. Zhong, R. L. Silverstein, W. A. Frazier, and N. P. Bouck. 1997. CD36 mediates the In vitro inhibitory effects of thrombospondin-1 on endothelial cells. J. Cell Biol. 138: 707–717. 20. Handunnetti, S. M., M. R. van Schravendijk, T. Hasler, J. W. Barnwell, D. E. Greenwalt, and R. J. Howard. 1992. Involvement of CD36 on erythrocytes as a rosetting receptor for Plasmodium falciparum-infected erythrocytes. Blood 80: 2097–2104. 21. Endemann, G., L. W. Stanton, K. S. Madden, C. M. Bryant, R. T. White, and A. A. Protter. 1993. CD36 is a receptor for oxidized low density lipoprotein. J. Biol. Chem. 268: 11811–11816. 22. Calvo, D., D. Gomez-Coronado, Y. Suarez, M. A. Lasuncion, and M. A. Vega. 1998. Human CD36 is a high affinity receptor for the native lipoproteins HDL, LDL, and VLDL. J. Lipid Res. 39: 777–788. 23. Tait, J. F., and C. Smith. 1999. Phosphatidylserine receptors: role of CD36 in binding of anionic phospholipid vesicles to monocytic cells. J. Biol. Chem. 274: 3048 –3054. 24. Boullier, A., P. Friedman, R. Harkewicz, K. Hartvigsen, S. R. Green, F. Almazan, E. A. Dennis, D. Steinberg, J. L. Witztum, and O. Quehenberger. 2005. Phosphocholine as a pattern recognition ligand for CD36. J. Lipid Res. 46: 969 –976. 25. Febbraio, M., D. P. Hajjar, and R. L. Silverstein. 2001. CD36: a class B scavenger receptor involved in angiogenesis, atherosclerosis, inflammation, and lipid metabolism. J. Clin. Invest. 108: 785–791. 26. Silverstein, R. L., M. Baird, S. K. Lo, and L. M. Yesner. 1992. Sense and antisense cDNA transfection of CD36 (glycoprotein IV) in melanoma cells: role of CD36 as a thrombospondin receptor. J. Biol. Chem. 267: 16607–16612. 27. Moore, K. J., K. J. El, L. A. Medeiros, K. Terada, C. Geula, A. D. Luster, and M. W. Freeman. 2002. A CD36-initiated signaling cascade mediates inflammatory effects of ␤-amyloid. J. Biol. Chem. 277: 47373– 47379.

237 28. El Khoury, J. B., K. J. Moore, T. K. Means, J. Leung, K. Terada, M. Toft, M. W. Freeman, and A. D. Luster. 2003. CD36 mediates the innate host response to ␤-amyloid. J. Exp. Med. 197: 1657–1666. 29. Febbraio, M., N. A. Abumrad, D. P. Hajjar, K. Sharma, W. Cheng, S. F. Pearce, and R. L. Silverstein. 1999. A null mutation in murine CD36 reveals an important role in fatty acid and lipoprotein metabolism. J. Biol. Chem. 274: 19055–19062. 30. Nicholson, A. C., J. Han, M. Febbraio, R. L. Silversterin, and D. P. Hajjar. 2001. Role of CD36, the macrophage class B scavenger receptor, in atherosclerosis. Ann. NY Acad. Sci. 947: 224 –228. 31. Hoebe, K., E. Janssen, and B. Beutler. 2004. The interface between innate and adaptive immunity. Nat. Immunol. 5: 971–974. 32. Hoebe, K., P. Georgel, S. Rutschmann, X. Du, S. Mudd, K. Crozat, S. Sovath, L. Shamel, T. Hartung, U. Zahringer, and B. Beutler. 2005. CD36 is a sensor of diacylglycerides. Nature 433: 523–527. 33. Taki, S., M. Meiering, and K. Rajewsky. 1993. Targeted insertion of a variable region gene into the immunoglobulin heavy chain locus. Science 262: 1268 –1271. 34. Caramalho, I., T. Lopes-Carvalho, D. Ostler, S. Zelenay, M. Haury, and J. Demengeot. 2003. Regulatory T cells selectively express toll-like receptors and are activated by lipopolysaccharide. J. Exp. Med. 197: 403– 411. 35. Cariappa, A., T. Shoham, H. Liu, S. Levy, C. Boucheix, and S. Pillai. 2005. The CD9 tetraspanin is not required for the development of peripheral B cells or for humoral immunity. J. Immunol. 175: 2925–2930. 36. Balazs, M., F. Martin, T. Zhou, and J. Kearney. 2002. Blood dendritic cells interact with splenic marginal zone B cells to initiate T-independent immune responses. Immunity 17: 341–352. 37. Benedict, C. L., and J. F. Kearney. 1999. Increased junctional diversity in fetal B cells results in a loss of protective anti-phosphorylcholine antibodies in adult mice. Immunity 10: 607– 617. 38. Won, W. J., K. Masuda, and J. F. Kearney. 2000. CD9 is a novel marker that discriminates MZ B cells from FO B cells. FASEB J. 14: A1191. 39. Koller, D., C. Ruedl, M. Loetscher, J. Vlach, S. Oehen, K. Oertle, M. Schirinzi, E. Deneuve, R. Moser, M. Kopf, et al. 2001. A high-throughput alphavirus-based expression cloning system for mammalian cells. Nat. Biotechnol. 19: 851– 855. 40. Takeda, K., and S. Akira. 2005. Toll-like receptors in innate immunity. Int. Immunol. 17: 1–14. 41. Konig, H., P. Pfisterer, L. M. Corcoran, and T. Wirth. 1995. Identification of CD36 as the first gene dependent on the B-cell differentiation factor Oct-2. Genes Dev. 9: 1598 –1607. 42. Takeuchi, O., K. Hoshino, T. Kawai, H. Sanjo, H. Takada, T. Ogawa, K. Takeda, and S. Akira. 1999. Differential roles of TLR2 and TLR4 in recognition of gramnegative and gram-positive bacterial cell wall components. Immunity 11: 443– 451. 43. Briles, D. E., M. Nahm, K. Schroer, J. Davie, P. Baker, J. Kearney, and R. Barletta. 1981. Antiphosphocholine antibodies found in normal mouse serum are protective against intravenous infection with type 3 streptococcus pneumoniae. J. Exp. Med. 153: 694 –705. 44. Triantafilou, M., F. G. Gamper, R. M. Haston, M. A. Mouratis, S. Morath, T. Hartung, and K. Triantafilou. 2006. Membrane sorting of toll-like receptor (TLR)-2/6 and TLR2/1 heterodimers at the cell surface determines heterotypic associations with CD36 and intracellular targeting. J. Biol. Chem. 281: 31002–31011. 45. Briles, D. E., J. L. Claflin, K. Schroer, and C. Forman. 1981. Mouse IgG3 antibodies are highly protective against infection with Streptococcus pneumoniae. Nature 294: 88 –90. 46. Oliver, A. M., F. Martin, and J. F. Kearney. 1997. Mouse CD38 is down-regulated on germinal center B cells and mature plasma cells. J. Immunol. 158: 1108 –1115.