The Radioprotective 105/MD-1 Complex Links TLR2 and TLR4/MD-2 in Antibody Response to Microbial Membranes1
http://www.100md.com
免疫学杂志 2005年第11期
Abstract
Low-affinity IgG3 Abs to microbial membranes are important for primary immune defense against microbes, but little is known about the importance of TLRs in their production. IgG3 levels were extremely low in mice lacking radioprotective 105 (RP105), a B cell surface molecule structurally related to TLRs. RP105–/– B cells proliferated poorly in response to not only the TLR4 ligand LPS ut also TLR2 ligand lipoproteins, both of which mediate the immunostimulatory activity of microbial membranes. RP105–/– mice were severely impaired in hapten-specific Ab production against LPS or lipoproteins. CD138 (syndecan-1)-positive plasma cells were detected after lipid A injection in wild-type spleen but much less in RP105–/– spleen. RP105 ligation in vivo induced plasma cell differentiation. RP105 expression was 3-fold higher on marginal zone B cells than on follicular and B1 cells and was down-regulated on germinal center cells. These results demonstrate that a signal via RP105 is uniquely important for regulating TLR-dependent Ab production to microbial membranes.
Introduction
Innate immunity provides a first line of defense against microbial pathogens (1) and is dependent on a restricted set of nonvariant, germline-encoded molecules that include secreted opsonins, C-type lectins, and scavenger receptors (2). B cells can contribute to innate immunity by secreting Abs with similarities to innate immune receptors in that they are semi-invariant and reactive with both self and microbial membrane glycolipids (3). These are principally IgM and IgG3 Abs that directly bind to microbial membranes, activate complement, facilitate their phagocytosis, and enhance their immunogenicity through Ag-trapping in secondary lymphoid organs (4, 5). They are produced as natural Abs to self-antigens and microbial flora or produced in primary, T-independent (TI)5 responses during microbial infections. B1 cells and splenic marginal zone (MZ) B cells are important for producing these protective IgM and IgG3 Abs. However, the recognition molecules on B cells that mediate these TI Ab responses have been poorly understood.
Microbial membranes were known to stimulate immune cells. LPS and lipoproteins were identified as principal components of the immunostimulating activity of microbial membranes. These components have turned out to be ligands for TLRs, a family of innate immune receptors for microbial products (6, 7). Heterodimers such as TLR1/TLR2 or TLR2/TLR6 mediate responses to particular membrane lipoproteins (8), whereas the TLR4/MD-2 complex is essential for LPS recognition (9, 10). LPS is a prototypical TI type 1 Ag. That is, it elicits Ab production without T cell help. Since B cells lacking TLR4 or MD-2 do not respond to LPS (10, 11), TLR4/MD-2 is required for LPS-induced Ab production. Much less is known about the importance of TLR2-mediated Ab responses to microbial lipoproteins.
B cells express another pair of TLR family proteins that are also important for LPS responses. Radioprotective 105 (RP105) forms a complex with MD-1 and can transmit powerful survival as well as proliferation signals when cross-linked by Abs (12, 13, 14). RP105–/– or MD-1–/– mice, like animals deficient in TLR4 or MD-2, are hyporesponsive to LPS (15, 16), but precise mechanisms that functionally couple the two types of complex remain unclear.
This study explored the relative importance of TLR family of receptors including TLR2, TLR4/MD-2, or RP105/MD-1 for innate Ab responses to microbial membranes. We found that RP105/MD-1 controls Ab production mediated via TLR2 and TLR4/MD-2 receptor complexes.
Materials and Methods
Bone marrow-derived macrophages and TNF- ELISA
Bone marrow cells were plated in 10-cm bacteriological plastic plates with 10% FCS-RPMI 1640 supplemented with 100 ng/ml recombinant murine M-CSF (Genzyme Techne). At day 7, adherent cells were harvested by scraper, plated at 1 x 105 cells/ml in 96-well plates, and cultured with or without TLR ligands in 10% FCS-RPMI 1640 without M-CSF for 24 h. TNF- in culture supernatants were determined using ELISA kits (BioSource International).
Results
Serum Ig levels are depressed in some TLR mutant and MyD88–/– mice
To explore a role for TLRs in serum Ab production, we first studied serum Ig levels in mice lacking MyD88, TLR2, TLR4, or RP105. MyD88–/– mice had higher IgG1 but lower IgG2a and IgG3 levels than wild-type mice (Fig. 1a). The immune system of MyD88–/– mice is skewed toward Th2 responses due to a lack of MyD88-dependent Th1 polarization (19, 20), and this could account for their high titers of serum IgG1 and their low titers of IgG2a. However, Th2 polarization would not explain the low IgG3 titers.
In contrast to MyD88–/– mice, we found no significant alterations in either IgG1 or IgG2a in TLR2–/– or TLR4–/– mutant mice (Fig. 1, b and c). Modest reductions in serum IgG3 were observed in TLR4–/– (p = 0.0016) mice. RP105–/– mice were severely impaired with respect to serum IgG3 on two genetic backgrounds (p = 0.0015 for C57BL/6 and p = 0.000037 for BALB/c background, respectively) (Fig. 1, d and e). Production of IgM and IgG2b were additionally decreased in BALB/c RP105–/– mice. These findings suggest that RP105 is important in regulating serum levels of innate antibodies, and especially those of the IgG3 class.
B cells lacking RP105 are impaired in proliferative responses to TLR2 but not TLR9 ligand
We previously demonstrated that RP105–/– B cells are hyporesponsive to LPS (15). Given that RP105–/– mice were more impaired in serum IgG3 production than TLR4–/– mice (Fig. 1), we asked a possibility that RP105/MD-1 has a role in responses to TLR ligands other than LPS. We stimulated RP105–/– B cells with Pam3CSK4 (TLR2/TLR1 ligand), MALP-2 (TLR2/TLR6 ligand), or the TLR9 ligand CpG. B cell proliferation in response to lipid A stimulation was impaired in RP105–/– B cells by at least an order of magnitude (Fig. 2A). We now show that RP105–/– B cells have defective proliferative responses to TLR2 ligands, either Pam3CSK4 or MALP-2 (Fig. 2B). Proliferation induced by these ligands could not be due to LPS contamination, because TLR2 ligands did not stimulate TLR2–/– B cells (Fig. 2B), which can respond to LPS (21). Hyporesponsiveness was more apparent in response to Pam3CSK4 than to MALP-2. As is the case with lipid A, the impairment of B cell proliferation in RP105–/– B cells seemed to be dose-dependent. The impairment was more apparent at lower concentrations of TLR2 ligands but not significant at higher concentration of Pam3CSK4 (Fig. 2, B and C). Interestingly, RP105–/– B cells responded normally to the TLR9 ligand CpG (Fig. 2D). CpG used in the present study did not stimulate B cells lacking TLR9, excluding contamination of LPS or other TLR ligands (data not shown). TLR9 is distinct from TLR2 or TLR4/MD-2 and similar to TLR7 in its ligand recognition in endosomal/lysosomal compartments (17, 22). Therefore, we stimulated RP105–/– B cells with the TLR7 ligand loxoribine (17). RP105–/– B cells proliferated normally (data not shown). RP105/MD-1 regulates B cell responses to LPS or TLR2 ligands but not those involving TLR7 or TLR9 ligands.
RP105–/– mice are impaired with respect to LPS-induced polyclonal Ab production
Serum Ig production is thought to be driven by TLR ligands from microbial flora. To study a role for RP105 in Ab production in vivo, we next investigated whether RP105 is important for polyclonal Ab production in vivo induced by i.p. injection of LPS or TLR2 ligands. Mice at 4–5 wk of age were injected weekly for 4 weeks with LPS (50 μg/mouse) or Pam3CSK4 (50 μg/mouse), and serum Ig levels were determined by ELISA. Significant difference in serum titers between wild-type and RP105–/– mice injected with Pam3CSK4 or MALP-2 was seen only in IgG3 (Fig. 3, right panel, and data not shown). Because such a difference in IgG3 was also seen without any stimulation at 11–12 wk of age (Fig. 1), it was difficult to exclude a possibility that the difference in IgG3 was due to age-dependent increase in IgG3. In contrast, LPS induced significant differences between wild-type and RP105–/– mice not only in IgG3 but also in IgM and IgG2b despite C57BL/6 background (Fig. 3, left panel), which appeared to be distinct from serum Ig titers at 11–12 wk of age (Fig. 1) or from the results with Pam3CSK4 stimulation (Fig. 3, right panel). RP105 is important for mediating polyclonal Ab production in vivo induced by LPS.
RP105–/– mice are impaired in production of Abs specific to LPS, TLR2 ligands, and bacteria
We then studied specific TI responses to the hapten, FITC conjugated to defined TLR2 ligands. These were injected into mice i.p. and FITC-specific Ab production was determined by ELISA (Fig. 4). Mice were also immunized with TNP-LPS, which induces TNP-specific Ab production in wild-type but not RP105–/– mice (15). The TLR2 ligands, FITC-Pam3CSK4 and FITC-MALP-2, induced FITC-specific IgM, IgG3, and IgG2b in normal animals, but these responses were all severely impaired in RP105–/– mice (Fig. 4). Similar results were obtained with TNP-Pam3CSK4 (data not shown). Because the immunostimulatory activity of microbial membranes is mostly explained by LPS and TLR2 ligands, we also assessed the importance of RP105 for primary Ab production against bacteria. Mice were immunized i.p. with heat-killed Escherichia coli and titers of Ab to E. coli were determined by ELISA. IgM, IgG3, and IgG2b Abs to E. coli were produced in wild-type but not in RP105–/– mice (Fig. 4). Thus, RP105/MD-1 is very important for Ab production against microbial membranes.
RP105 is expressed on B1, MZ B, and follicular B, but not germinal center (GC) B cells
Although the above results indicate that RP105 regulates TI type 1 responses, we previously showed that RP105–/– mice were not impaired with respect to thymus-dependent (TD) responses (15). Therefore, we wondered whether this molecule might be differentially expressed by lymphocyte subsets. All B cell subsets expressed similar amounts of TLR2. However, MZ B cells had 3-fold higher mean fluorescence intensities of RP105 than B1 or follicular B cells (Fig. 5, a and b). TLR4/MD-2 was barely detectable. We previously found that GC B cells in human tonsils had less cell surface RP105 than follicular B cells (23). We now show that the same is true for mice. RP105 was almost absent in PNA-positive GC B cells in the spleens of mice immunized with SRBC (Fig. 5c). Whereas GC B cells are committed to T-dependent Ab responses, MZ B cells contribute to TI Ab production against microbial membranes (24), in which IgG3 is a principal Ig isotype. This is consistent with a unique role for RP105 in TI responses.
RP105–/– mice are not impaired in expression of TLR4, MD-2, and TLR2, or in macrophages responses to lipid A or TLR2 ligands
To address the mechanism by which B cell response to TLR4 and TLR2 ligands were impaired in RP105–/– mice, we first studied expression of TLR2, TLR4, and MD-2 in RP105–/– B cells. Although an Ab to the TLR4/MD-2 complex did not give significant staining on normal splenic B cells (Fig. 5a), expression of mRNAs for TLR4 and MD-2 were confirmed in RP105–/– B cells by RT-PCR (Fig. 6a). TLR2 densities were similar on B cells from wild-type and RP105–/– mice (Fig. 6b). RP105 is expressed not only on B cells but also on bone marrow-derived macrophages and dendritic cells (Fig. 6c and data not shown). We could not see any difference between wild-type and RP105–/– macrophages in cell surface TLR4/MD-2 and TLR2 as judged by flow cytometry (Fig. 6c).
RP105–/– bone marrow-derived macrophages showed sharp contrast to B cells in that they were not impaired in TNF- production induced with lipid A or TLR2 ligands (Fig. 6d). RP105–/– bone marrow-derived dendritic cells were similar to macrophages in that they were not impaired in IL-12 production or up-regulation of costimulatory molecules in response to lipid A or TLR2 ligands (data not shown). Coexpression of RP105/MD-1 did not necessarily influence the expression of TLR4/MD-2 and TLR2.
Signals transmitted via RP105/MD-1 induce plasma cell differentiation
CD19 plays an important role in regulating signal transduction through RP105 (25). In keeping with this, mAb-mediated RP105 ligation induced potent activation in B cells (12) but not in macrophages and dendritic cells, as judged by TNF- production in macrophages, IL-12 production in dendritic cells, and up-regulation of costimulatory molecules on dendritic cells (Fig. 6d and data not shown). B cell-specific defect in TLR responses of RP105–/– mice seemed to correlate with cell activation induced by RP105 ligation. RP105-dependent B cell activation is likely to have a role in Ab production in response to LPS or TLR2 ligands. To address this possibility, we studied plasma cell differentiation in vivo. B lymphocyte-induced maturation protein (Blimp-1) is a transcriptional repressor that drives the terminal differentiation of B cells into Ig-secreting plasma cells (26). Furthermore, Blimp-1 is up-regulated early during the transition of mature B cells to IgM-secreting plasma cells (27). Transcripts for Blimp-1 increased dramatically in wild-type but not in RP105–/– spleens following immunization with LPS (Fig. 7a). Much lower but significant up-regulation of transcripts for Blimp-1 was observed in wild-type mice stimulated with Pam3CSK4 but not in RP105–/– mice (Fig. 7a). Soro et al. (27) recently demonstrated that Blimp-1 expression in B cells were up-regulated by LPS but not by anti-CD40 mAb and IL-4. In keeping with this, Blimp-1 expression was not up-regulated by anti-CD40 mAb injection but clearly up-regulated by anti-RP105 mAb injection (Fig. 7b).
Syndecan-1 (CD138) is normally expressed on bone marrow pre-B cells and IgM-secreting plasma cells but not on mature B cells (27). Lipid A injection induced B220+ CD138+ plasma cells in wild-type but much less so in RP105–/– spleen cells (Fig. 7c). Much lower but significant increase in plasma cells were observed with stimulation by TLR2 ligand Pam3CSK4 in wild-type mice but not in RP105–/– mice (Fig. 7c). Furthermore, RP105 mAb injection induced CD138+ plasma cells only in wild-type spleens (Fig. 7d). Although CD40 mAb caused CD86 up-regulation in vivo (Fig. 7e) and drove B cells to proliferate in vitro (28), anti-CD40 alone induced much less B220+ CD138+ plasma cells than anti-RP105 (Fig. 7d), which is consistent with a previous report (27). These results strongly suggested that a signal via RP105/MD-1 drove B cells to differentiate into Ab-secreting cells expressing Blimp-1 and CD138.
Discussion
Low-affinity polyreactive Abs of the IgM and IgG3 classes provide first line protection against microbial pathogens, and the experiments conducted in this study addressed their dependence on particular TLRs. Animals lacking the RP105/MD-1 containing receptors, previously known to be essential to LPS responses, were IgG3 deficient. Normal polyclonal Ig production in response to LPS required the same complex as did synthesis of hapten-specific, thymus-independent Abs. MZ B cells displayed high densities of RP105/MD-1, and ligation of this receptor complex on spleen B cells promoted terminal differentiation. We will now discuss the importance of RP105/MD-1 relative to other TLR family-containing receptors and speculate how they might be functionally coupled.
The lack of RP105 did not seem to influence expression of TLR4, MD-2, and TLR2, as revealed by RT-PCR and flow cytometry (Fig. 6, a–c). RP105/MD-1 is expressed not only on B cells but also on macrophages and dendritic cells, but the phenotypes of RP105–/– mice are B cell specific. That is, macrophages and dendritic cells from RP105–/– mice were not impaired in response to LPS or TLR2 ligands (Fig. 6d and data not shown). It is of note that the agonistic activity of the RP105 mAb is also B cell specific in that it did not activate macrophages and dendritic cells (Fig. 6d and data not shown). We prefer a possibility that a signal via RP105/MD-1 is required for TI responses to LPS or TLR2 ligands. This is supported by the following results. Blimp-1high CD138+Ab-secreting cells appeared in response to lipid A stimulation in wild-type but much less so in RP105–/– mice (Fig. 7). Injection of RP105 mAb induced Blimp-1high CD138+ plasmacytic cells in the spleen (Fig. 7).
RP105–/– B cells were able to proliferate in response to higher concentrations of lipid A or TLR2 ligands (Fig. 2). In contrast, we could not detect polyclonal Ab production even after repeated injection of LPS (Fig. 3). Moreover, RP105–/– mice were comparable with MyD88–/– mice in serum IgG3 deficiency. TLR signals without RP105 are able to induce B cell proliferation particularly at high concentration of LPS or TLR2 ligands but not differentiation into plasma cells. We would like to conclude that RP105/MD-1-mediated signal is indispensable for B cell differentiation into plasma cells during TI responses.
The present findings beg the important question of what molecular mechanisms couple RP105/MD-1 with TLR4/MD-2 or with TLR2. Given that not only LPS but also TLR2 ligands stimulate the RP105/MD-1 complex, it is unlikely that RP105/MD-1 directly interacts with these TLR ligands. Tsuneyoshi et al. (29) recently demonstrated that LPS interacts with MD-2 but not with MD-1. In keeping with this, LPS was coprecipitated with TLR4/MD-2 but not with coexpressed RP105/MD-1 (data not shown). LPS or TLR ligands stimulate TLR4/MD-2 and TLR2, and these activated TLRs might then activate RP105/MD-1 by an as yet unknown mechanism. TLR9 and TLR7 have no obvious functional link with RP105/MD-1. TLR9 and TLR7 reside in the endoplasmic reticulum (17, 22), and TLR9 is recruited to early endosomes, where interaction with ligand and subsequent signaling occurs. Cell surface expression seems to be important for use of RP105/MD-1. As one possibility, RP105/MD-1 could be physically associated with TLR4/MD-2 and/or TLR2 on the cell surface. Physical association may lead to coclustering of RP105/MD-1 with TLR4/MD-2 or with TLR2 upon stimulation with TLR ligands.
There were marked differences between Pam3CSK4 and MALP-2 in inducing B cell responses. MALP-2 required 100–1000 times higher concentration than lipid A or Pam3CSK4 (Fig. 2). Ab titers induced by FITC-MALP-2 were also lower than those induced by FITC-Pam3CSK4 (Fig. 4). Given that TLR6 is required for response to MALP-2, cell surface expression of TLR6 may be low on the splenic B cell surface. Although FITC-specific Ab production was significantly induced by Pam3CSK4 in a manner dependent on RP105 (Fig. 4a), Pam3CSK4 was much lower than lipid A in inducing plasmablasts (Fig. 7) and we were able to detect polyclonal Ab production with LPS but not with Pam3CSK4 (Fig. 3). In contrast, Pam3CSK4 was as potent as lipid A in inducing B cell proliferation (Fig. 2). Pam3CSK4-induced B cell proliferation might be differentially regulated from B cell differentiation. It is of note that CD19 is shown to regulate RP105-signaling (25). CD19–/– B cells were reported to be lower than wild-type B cells in Ab production to TNP-LPS (30), suggesting that CD19 activation in LPS response are directly linked with Ab production. It is possible that CD19 activation during Pam3CSK4 response is much lower than that in LPS response. CD19 was shown to be phosphorylated when a B cell lymphoma A20 was stimulated with LPS (25). We tried to examine CD19 phosphorylation in normal B cells activated by LPS or Pam3CSK4, but we could not detect it (data not shown). CD19 phosphorylation may be too low in normal B cells for biochemical detection. Further studies are under way.
For primary defense, Abs in the serum or produced during TI responses must bind to microbial membranes. The immunostimulatory activity of these membranes is explained by the presence of ligands for TLR4/MD-2 and TLR2. LPS and TLR2 ligands were both able to induce Ag-specific Ab production in vivo (Figs. 3 and 4). RP105–/– mice were impaired in mounting Ab production not only against LPS and TLR2 ligands but also against heat-killed bacteria (Figs. 3 and 4). Taken together, primary Ab production to microbial membranes is regulated by TLR4/MD-2, TLR2, and RP105/MD-1. TLR4–/– and TLR2–/– mice were only modestly impaired in serum IgG3 production when compared with RP105–/– and MyD88–/– mice (Fig. 1). TLR2 and TLR4 would compensate for a lack of TLR4 or TLR2, respectively. In contrast, TI responses to TLR4 and TLR2 ligands were all impaired in RP105–/–, leading to low serum IgG3 titer. It is important to study serum Ig titers of mice lacking both TLR4 and TLR2.
RP105–/– mice have normal levels of serum IgG1 and IgG2a (Fig. 1), and were not impaired with respect to TD Ab production (15). In keeping with this, RP105/MD-1 was nearly absent from GC B cells that were committed to TD Ab production (Fig. 5c). In sharp contrast, the density of RP105/MD-1 was, among B cell subsets, brightest on MZ B cells, which greatly contribute to TI responses (Fig. 5a) (24). RP105–/– mice are quite different from CD40–/– mice that are impaired in TD but not in TI responses (31). An additional difference between these two types of receptors was found by injection of mAb. Injection of RP105 but not CD40 mAb induced plasma cells in spleen (Fig. 7). These differences between RP105/MD-1 and CD40 could represent important distinctions between TI and TD responses.
In conclusion, TLR2- and TLR4/MD-2-containing receptors are essential for recognition of lipoproteins and LPS. Whereas RP105/MD-1 may not directly recognize these microbial products, it is displayed at high levels on MZ B cells and is essential for both polyclonal and specific IgG3 Ab responses to TI type 1 vaccines. These observations add to our understanding of innate immune mechanisms and should be important for developing new immunization strategies.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This study was supported by Special Coordination Funds of the Ministry of Education, Culture, Sports, Science and Technology of the Japanese Government, Uehara Memorial Foundation, the Naito Foundation, and Sankyo Co.
2 Y.N. and T.Ko. contributed equally to this study.
3 Current address: Immunobiology and Cancer Research Program, Oklahoma Medical Research Foundation, Oklahoma City, OK 73104.
4 Address correspondence and reprint requests to Dr. Kensuke Miyake, Division of Infectious Genetics, Institute of Medical Science, University of Tokyo, 4-6-1 Shirokanedai, Minatoku, Tokyo 108-8639, Japan. E-mail address: kmiyake{at}ims.u-tokyo.ac.jp
5 Abbreviations used in this paper: TI, T independent; MZ, marginal zone; RP105, radioprotective 105; PNA, peanut agglutinin; TNP. trinitrophenyl; HPRT, hypoxanthine-guanine phosphoribosyltransferase; GC, germinal center; TD, thymus dependent.
Received for publication August 30, 2004. Accepted for publication March 29, 2005.
References
Janeway, C. A., Jr, R. Medzhitov. 2002. Innate immune recognition. Annu. Rev. Immunol. 20: 197-216.
Gordon, S.. 2002. Pattern recognition receptors: doubling up for the innate immune response. Cell 111: 927-930.
Bendelac, A., M. Bonneville, J. F. Kearney. 2001. Autoreactivity by design: innate B and T lymphocytes. Nat. Rev. Immunol. 1: 177-186.
Baumgarth, N., O. C. Herman, G. C. Jager, L. E. Brown, L. A. Herzenberg, J. Chen. 2000. B-1 and B-2 cell-derived immunoglobulin M antibodies are nonredundant components of the protective response to influenza virus infection. J. Exp. Med. 192: 271-280.
Ochsenbein, A. F., T. Fehr, C. Lutz, M. Suter, F. Brombacher, H. Hengartner, R. M. Zinkernagel. 1999. Control of early viral and bacterial distribution and disease by natural antibodies. Science 286: 2156-2159.
Takeda, K., S. Akira. 2004. Microbial recognition by Toll-like receptors. J. Dermatol. Sci. 34: 73-82.
Medzhitov, R., C. A. Janeway, Jr. 2002. Decoding the patterns of self and nonself by the innate immune system. Science 296: 298-300.(Yoshinori Nagai2,3,*, Tos)
Low-affinity IgG3 Abs to microbial membranes are important for primary immune defense against microbes, but little is known about the importance of TLRs in their production. IgG3 levels were extremely low in mice lacking radioprotective 105 (RP105), a B cell surface molecule structurally related to TLRs. RP105–/– B cells proliferated poorly in response to not only the TLR4 ligand LPS ut also TLR2 ligand lipoproteins, both of which mediate the immunostimulatory activity of microbial membranes. RP105–/– mice were severely impaired in hapten-specific Ab production against LPS or lipoproteins. CD138 (syndecan-1)-positive plasma cells were detected after lipid A injection in wild-type spleen but much less in RP105–/– spleen. RP105 ligation in vivo induced plasma cell differentiation. RP105 expression was 3-fold higher on marginal zone B cells than on follicular and B1 cells and was down-regulated on germinal center cells. These results demonstrate that a signal via RP105 is uniquely important for regulating TLR-dependent Ab production to microbial membranes.
Introduction
Innate immunity provides a first line of defense against microbial pathogens (1) and is dependent on a restricted set of nonvariant, germline-encoded molecules that include secreted opsonins, C-type lectins, and scavenger receptors (2). B cells can contribute to innate immunity by secreting Abs with similarities to innate immune receptors in that they are semi-invariant and reactive with both self and microbial membrane glycolipids (3). These are principally IgM and IgG3 Abs that directly bind to microbial membranes, activate complement, facilitate their phagocytosis, and enhance their immunogenicity through Ag-trapping in secondary lymphoid organs (4, 5). They are produced as natural Abs to self-antigens and microbial flora or produced in primary, T-independent (TI)5 responses during microbial infections. B1 cells and splenic marginal zone (MZ) B cells are important for producing these protective IgM and IgG3 Abs. However, the recognition molecules on B cells that mediate these TI Ab responses have been poorly understood.
Microbial membranes were known to stimulate immune cells. LPS and lipoproteins were identified as principal components of the immunostimulating activity of microbial membranes. These components have turned out to be ligands for TLRs, a family of innate immune receptors for microbial products (6, 7). Heterodimers such as TLR1/TLR2 or TLR2/TLR6 mediate responses to particular membrane lipoproteins (8), whereas the TLR4/MD-2 complex is essential for LPS recognition (9, 10). LPS is a prototypical TI type 1 Ag. That is, it elicits Ab production without T cell help. Since B cells lacking TLR4 or MD-2 do not respond to LPS (10, 11), TLR4/MD-2 is required for LPS-induced Ab production. Much less is known about the importance of TLR2-mediated Ab responses to microbial lipoproteins.
B cells express another pair of TLR family proteins that are also important for LPS responses. Radioprotective 105 (RP105) forms a complex with MD-1 and can transmit powerful survival as well as proliferation signals when cross-linked by Abs (12, 13, 14). RP105–/– or MD-1–/– mice, like animals deficient in TLR4 or MD-2, are hyporesponsive to LPS (15, 16), but precise mechanisms that functionally couple the two types of complex remain unclear.
This study explored the relative importance of TLR family of receptors including TLR2, TLR4/MD-2, or RP105/MD-1 for innate Ab responses to microbial membranes. We found that RP105/MD-1 controls Ab production mediated via TLR2 and TLR4/MD-2 receptor complexes.
Materials and Methods
Bone marrow-derived macrophages and TNF- ELISA
Bone marrow cells were plated in 10-cm bacteriological plastic plates with 10% FCS-RPMI 1640 supplemented with 100 ng/ml recombinant murine M-CSF (Genzyme Techne). At day 7, adherent cells were harvested by scraper, plated at 1 x 105 cells/ml in 96-well plates, and cultured with or without TLR ligands in 10% FCS-RPMI 1640 without M-CSF for 24 h. TNF- in culture supernatants were determined using ELISA kits (BioSource International).
Results
Serum Ig levels are depressed in some TLR mutant and MyD88–/– mice
To explore a role for TLRs in serum Ab production, we first studied serum Ig levels in mice lacking MyD88, TLR2, TLR4, or RP105. MyD88–/– mice had higher IgG1 but lower IgG2a and IgG3 levels than wild-type mice (Fig. 1a). The immune system of MyD88–/– mice is skewed toward Th2 responses due to a lack of MyD88-dependent Th1 polarization (19, 20), and this could account for their high titers of serum IgG1 and their low titers of IgG2a. However, Th2 polarization would not explain the low IgG3 titers.
In contrast to MyD88–/– mice, we found no significant alterations in either IgG1 or IgG2a in TLR2–/– or TLR4–/– mutant mice (Fig. 1, b and c). Modest reductions in serum IgG3 were observed in TLR4–/– (p = 0.0016) mice. RP105–/– mice were severely impaired with respect to serum IgG3 on two genetic backgrounds (p = 0.0015 for C57BL/6 and p = 0.000037 for BALB/c background, respectively) (Fig. 1, d and e). Production of IgM and IgG2b were additionally decreased in BALB/c RP105–/– mice. These findings suggest that RP105 is important in regulating serum levels of innate antibodies, and especially those of the IgG3 class.
B cells lacking RP105 are impaired in proliferative responses to TLR2 but not TLR9 ligand
We previously demonstrated that RP105–/– B cells are hyporesponsive to LPS (15). Given that RP105–/– mice were more impaired in serum IgG3 production than TLR4–/– mice (Fig. 1), we asked a possibility that RP105/MD-1 has a role in responses to TLR ligands other than LPS. We stimulated RP105–/– B cells with Pam3CSK4 (TLR2/TLR1 ligand), MALP-2 (TLR2/TLR6 ligand), or the TLR9 ligand CpG. B cell proliferation in response to lipid A stimulation was impaired in RP105–/– B cells by at least an order of magnitude (Fig. 2A). We now show that RP105–/– B cells have defective proliferative responses to TLR2 ligands, either Pam3CSK4 or MALP-2 (Fig. 2B). Proliferation induced by these ligands could not be due to LPS contamination, because TLR2 ligands did not stimulate TLR2–/– B cells (Fig. 2B), which can respond to LPS (21). Hyporesponsiveness was more apparent in response to Pam3CSK4 than to MALP-2. As is the case with lipid A, the impairment of B cell proliferation in RP105–/– B cells seemed to be dose-dependent. The impairment was more apparent at lower concentrations of TLR2 ligands but not significant at higher concentration of Pam3CSK4 (Fig. 2, B and C). Interestingly, RP105–/– B cells responded normally to the TLR9 ligand CpG (Fig. 2D). CpG used in the present study did not stimulate B cells lacking TLR9, excluding contamination of LPS or other TLR ligands (data not shown). TLR9 is distinct from TLR2 or TLR4/MD-2 and similar to TLR7 in its ligand recognition in endosomal/lysosomal compartments (17, 22). Therefore, we stimulated RP105–/– B cells with the TLR7 ligand loxoribine (17). RP105–/– B cells proliferated normally (data not shown). RP105/MD-1 regulates B cell responses to LPS or TLR2 ligands but not those involving TLR7 or TLR9 ligands.
RP105–/– mice are impaired with respect to LPS-induced polyclonal Ab production
Serum Ig production is thought to be driven by TLR ligands from microbial flora. To study a role for RP105 in Ab production in vivo, we next investigated whether RP105 is important for polyclonal Ab production in vivo induced by i.p. injection of LPS or TLR2 ligands. Mice at 4–5 wk of age were injected weekly for 4 weeks with LPS (50 μg/mouse) or Pam3CSK4 (50 μg/mouse), and serum Ig levels were determined by ELISA. Significant difference in serum titers between wild-type and RP105–/– mice injected with Pam3CSK4 or MALP-2 was seen only in IgG3 (Fig. 3, right panel, and data not shown). Because such a difference in IgG3 was also seen without any stimulation at 11–12 wk of age (Fig. 1), it was difficult to exclude a possibility that the difference in IgG3 was due to age-dependent increase in IgG3. In contrast, LPS induced significant differences between wild-type and RP105–/– mice not only in IgG3 but also in IgM and IgG2b despite C57BL/6 background (Fig. 3, left panel), which appeared to be distinct from serum Ig titers at 11–12 wk of age (Fig. 1) or from the results with Pam3CSK4 stimulation (Fig. 3, right panel). RP105 is important for mediating polyclonal Ab production in vivo induced by LPS.
RP105–/– mice are impaired in production of Abs specific to LPS, TLR2 ligands, and bacteria
We then studied specific TI responses to the hapten, FITC conjugated to defined TLR2 ligands. These were injected into mice i.p. and FITC-specific Ab production was determined by ELISA (Fig. 4). Mice were also immunized with TNP-LPS, which induces TNP-specific Ab production in wild-type but not RP105–/– mice (15). The TLR2 ligands, FITC-Pam3CSK4 and FITC-MALP-2, induced FITC-specific IgM, IgG3, and IgG2b in normal animals, but these responses were all severely impaired in RP105–/– mice (Fig. 4). Similar results were obtained with TNP-Pam3CSK4 (data not shown). Because the immunostimulatory activity of microbial membranes is mostly explained by LPS and TLR2 ligands, we also assessed the importance of RP105 for primary Ab production against bacteria. Mice were immunized i.p. with heat-killed Escherichia coli and titers of Ab to E. coli were determined by ELISA. IgM, IgG3, and IgG2b Abs to E. coli were produced in wild-type but not in RP105–/– mice (Fig. 4). Thus, RP105/MD-1 is very important for Ab production against microbial membranes.
RP105 is expressed on B1, MZ B, and follicular B, but not germinal center (GC) B cells
Although the above results indicate that RP105 regulates TI type 1 responses, we previously showed that RP105–/– mice were not impaired with respect to thymus-dependent (TD) responses (15). Therefore, we wondered whether this molecule might be differentially expressed by lymphocyte subsets. All B cell subsets expressed similar amounts of TLR2. However, MZ B cells had 3-fold higher mean fluorescence intensities of RP105 than B1 or follicular B cells (Fig. 5, a and b). TLR4/MD-2 was barely detectable. We previously found that GC B cells in human tonsils had less cell surface RP105 than follicular B cells (23). We now show that the same is true for mice. RP105 was almost absent in PNA-positive GC B cells in the spleens of mice immunized with SRBC (Fig. 5c). Whereas GC B cells are committed to T-dependent Ab responses, MZ B cells contribute to TI Ab production against microbial membranes (24), in which IgG3 is a principal Ig isotype. This is consistent with a unique role for RP105 in TI responses.
RP105–/– mice are not impaired in expression of TLR4, MD-2, and TLR2, or in macrophages responses to lipid A or TLR2 ligands
To address the mechanism by which B cell response to TLR4 and TLR2 ligands were impaired in RP105–/– mice, we first studied expression of TLR2, TLR4, and MD-2 in RP105–/– B cells. Although an Ab to the TLR4/MD-2 complex did not give significant staining on normal splenic B cells (Fig. 5a), expression of mRNAs for TLR4 and MD-2 were confirmed in RP105–/– B cells by RT-PCR (Fig. 6a). TLR2 densities were similar on B cells from wild-type and RP105–/– mice (Fig. 6b). RP105 is expressed not only on B cells but also on bone marrow-derived macrophages and dendritic cells (Fig. 6c and data not shown). We could not see any difference between wild-type and RP105–/– macrophages in cell surface TLR4/MD-2 and TLR2 as judged by flow cytometry (Fig. 6c).
RP105–/– bone marrow-derived macrophages showed sharp contrast to B cells in that they were not impaired in TNF- production induced with lipid A or TLR2 ligands (Fig. 6d). RP105–/– bone marrow-derived dendritic cells were similar to macrophages in that they were not impaired in IL-12 production or up-regulation of costimulatory molecules in response to lipid A or TLR2 ligands (data not shown). Coexpression of RP105/MD-1 did not necessarily influence the expression of TLR4/MD-2 and TLR2.
Signals transmitted via RP105/MD-1 induce plasma cell differentiation
CD19 plays an important role in regulating signal transduction through RP105 (25). In keeping with this, mAb-mediated RP105 ligation induced potent activation in B cells (12) but not in macrophages and dendritic cells, as judged by TNF- production in macrophages, IL-12 production in dendritic cells, and up-regulation of costimulatory molecules on dendritic cells (Fig. 6d and data not shown). B cell-specific defect in TLR responses of RP105–/– mice seemed to correlate with cell activation induced by RP105 ligation. RP105-dependent B cell activation is likely to have a role in Ab production in response to LPS or TLR2 ligands. To address this possibility, we studied plasma cell differentiation in vivo. B lymphocyte-induced maturation protein (Blimp-1) is a transcriptional repressor that drives the terminal differentiation of B cells into Ig-secreting plasma cells (26). Furthermore, Blimp-1 is up-regulated early during the transition of mature B cells to IgM-secreting plasma cells (27). Transcripts for Blimp-1 increased dramatically in wild-type but not in RP105–/– spleens following immunization with LPS (Fig. 7a). Much lower but significant up-regulation of transcripts for Blimp-1 was observed in wild-type mice stimulated with Pam3CSK4 but not in RP105–/– mice (Fig. 7a). Soro et al. (27) recently demonstrated that Blimp-1 expression in B cells were up-regulated by LPS but not by anti-CD40 mAb and IL-4. In keeping with this, Blimp-1 expression was not up-regulated by anti-CD40 mAb injection but clearly up-regulated by anti-RP105 mAb injection (Fig. 7b).
Syndecan-1 (CD138) is normally expressed on bone marrow pre-B cells and IgM-secreting plasma cells but not on mature B cells (27). Lipid A injection induced B220+ CD138+ plasma cells in wild-type but much less so in RP105–/– spleen cells (Fig. 7c). Much lower but significant increase in plasma cells were observed with stimulation by TLR2 ligand Pam3CSK4 in wild-type mice but not in RP105–/– mice (Fig. 7c). Furthermore, RP105 mAb injection induced CD138+ plasma cells only in wild-type spleens (Fig. 7d). Although CD40 mAb caused CD86 up-regulation in vivo (Fig. 7e) and drove B cells to proliferate in vitro (28), anti-CD40 alone induced much less B220+ CD138+ plasma cells than anti-RP105 (Fig. 7d), which is consistent with a previous report (27). These results strongly suggested that a signal via RP105/MD-1 drove B cells to differentiate into Ab-secreting cells expressing Blimp-1 and CD138.
Discussion
Low-affinity polyreactive Abs of the IgM and IgG3 classes provide first line protection against microbial pathogens, and the experiments conducted in this study addressed their dependence on particular TLRs. Animals lacking the RP105/MD-1 containing receptors, previously known to be essential to LPS responses, were IgG3 deficient. Normal polyclonal Ig production in response to LPS required the same complex as did synthesis of hapten-specific, thymus-independent Abs. MZ B cells displayed high densities of RP105/MD-1, and ligation of this receptor complex on spleen B cells promoted terminal differentiation. We will now discuss the importance of RP105/MD-1 relative to other TLR family-containing receptors and speculate how they might be functionally coupled.
The lack of RP105 did not seem to influence expression of TLR4, MD-2, and TLR2, as revealed by RT-PCR and flow cytometry (Fig. 6, a–c). RP105/MD-1 is expressed not only on B cells but also on macrophages and dendritic cells, but the phenotypes of RP105–/– mice are B cell specific. That is, macrophages and dendritic cells from RP105–/– mice were not impaired in response to LPS or TLR2 ligands (Fig. 6d and data not shown). It is of note that the agonistic activity of the RP105 mAb is also B cell specific in that it did not activate macrophages and dendritic cells (Fig. 6d and data not shown). We prefer a possibility that a signal via RP105/MD-1 is required for TI responses to LPS or TLR2 ligands. This is supported by the following results. Blimp-1high CD138+Ab-secreting cells appeared in response to lipid A stimulation in wild-type but much less so in RP105–/– mice (Fig. 7). Injection of RP105 mAb induced Blimp-1high CD138+ plasmacytic cells in the spleen (Fig. 7).
RP105–/– B cells were able to proliferate in response to higher concentrations of lipid A or TLR2 ligands (Fig. 2). In contrast, we could not detect polyclonal Ab production even after repeated injection of LPS (Fig. 3). Moreover, RP105–/– mice were comparable with MyD88–/– mice in serum IgG3 deficiency. TLR signals without RP105 are able to induce B cell proliferation particularly at high concentration of LPS or TLR2 ligands but not differentiation into plasma cells. We would like to conclude that RP105/MD-1-mediated signal is indispensable for B cell differentiation into plasma cells during TI responses.
The present findings beg the important question of what molecular mechanisms couple RP105/MD-1 with TLR4/MD-2 or with TLR2. Given that not only LPS but also TLR2 ligands stimulate the RP105/MD-1 complex, it is unlikely that RP105/MD-1 directly interacts with these TLR ligands. Tsuneyoshi et al. (29) recently demonstrated that LPS interacts with MD-2 but not with MD-1. In keeping with this, LPS was coprecipitated with TLR4/MD-2 but not with coexpressed RP105/MD-1 (data not shown). LPS or TLR ligands stimulate TLR4/MD-2 and TLR2, and these activated TLRs might then activate RP105/MD-1 by an as yet unknown mechanism. TLR9 and TLR7 have no obvious functional link with RP105/MD-1. TLR9 and TLR7 reside in the endoplasmic reticulum (17, 22), and TLR9 is recruited to early endosomes, where interaction with ligand and subsequent signaling occurs. Cell surface expression seems to be important for use of RP105/MD-1. As one possibility, RP105/MD-1 could be physically associated with TLR4/MD-2 and/or TLR2 on the cell surface. Physical association may lead to coclustering of RP105/MD-1 with TLR4/MD-2 or with TLR2 upon stimulation with TLR ligands.
There were marked differences between Pam3CSK4 and MALP-2 in inducing B cell responses. MALP-2 required 100–1000 times higher concentration than lipid A or Pam3CSK4 (Fig. 2). Ab titers induced by FITC-MALP-2 were also lower than those induced by FITC-Pam3CSK4 (Fig. 4). Given that TLR6 is required for response to MALP-2, cell surface expression of TLR6 may be low on the splenic B cell surface. Although FITC-specific Ab production was significantly induced by Pam3CSK4 in a manner dependent on RP105 (Fig. 4a), Pam3CSK4 was much lower than lipid A in inducing plasmablasts (Fig. 7) and we were able to detect polyclonal Ab production with LPS but not with Pam3CSK4 (Fig. 3). In contrast, Pam3CSK4 was as potent as lipid A in inducing B cell proliferation (Fig. 2). Pam3CSK4-induced B cell proliferation might be differentially regulated from B cell differentiation. It is of note that CD19 is shown to regulate RP105-signaling (25). CD19–/– B cells were reported to be lower than wild-type B cells in Ab production to TNP-LPS (30), suggesting that CD19 activation in LPS response are directly linked with Ab production. It is possible that CD19 activation during Pam3CSK4 response is much lower than that in LPS response. CD19 was shown to be phosphorylated when a B cell lymphoma A20 was stimulated with LPS (25). We tried to examine CD19 phosphorylation in normal B cells activated by LPS or Pam3CSK4, but we could not detect it (data not shown). CD19 phosphorylation may be too low in normal B cells for biochemical detection. Further studies are under way.
For primary defense, Abs in the serum or produced during TI responses must bind to microbial membranes. The immunostimulatory activity of these membranes is explained by the presence of ligands for TLR4/MD-2 and TLR2. LPS and TLR2 ligands were both able to induce Ag-specific Ab production in vivo (Figs. 3 and 4). RP105–/– mice were impaired in mounting Ab production not only against LPS and TLR2 ligands but also against heat-killed bacteria (Figs. 3 and 4). Taken together, primary Ab production to microbial membranes is regulated by TLR4/MD-2, TLR2, and RP105/MD-1. TLR4–/– and TLR2–/– mice were only modestly impaired in serum IgG3 production when compared with RP105–/– and MyD88–/– mice (Fig. 1). TLR2 and TLR4 would compensate for a lack of TLR4 or TLR2, respectively. In contrast, TI responses to TLR4 and TLR2 ligands were all impaired in RP105–/–, leading to low serum IgG3 titer. It is important to study serum Ig titers of mice lacking both TLR4 and TLR2.
RP105–/– mice have normal levels of serum IgG1 and IgG2a (Fig. 1), and were not impaired with respect to TD Ab production (15). In keeping with this, RP105/MD-1 was nearly absent from GC B cells that were committed to TD Ab production (Fig. 5c). In sharp contrast, the density of RP105/MD-1 was, among B cell subsets, brightest on MZ B cells, which greatly contribute to TI responses (Fig. 5a) (24). RP105–/– mice are quite different from CD40–/– mice that are impaired in TD but not in TI responses (31). An additional difference between these two types of receptors was found by injection of mAb. Injection of RP105 but not CD40 mAb induced plasma cells in spleen (Fig. 7). These differences between RP105/MD-1 and CD40 could represent important distinctions between TI and TD responses.
In conclusion, TLR2- and TLR4/MD-2-containing receptors are essential for recognition of lipoproteins and LPS. Whereas RP105/MD-1 may not directly recognize these microbial products, it is displayed at high levels on MZ B cells and is essential for both polyclonal and specific IgG3 Ab responses to TI type 1 vaccines. These observations add to our understanding of innate immune mechanisms and should be important for developing new immunization strategies.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 This study was supported by Special Coordination Funds of the Ministry of Education, Culture, Sports, Science and Technology of the Japanese Government, Uehara Memorial Foundation, the Naito Foundation, and Sankyo Co.
2 Y.N. and T.Ko. contributed equally to this study.
3 Current address: Immunobiology and Cancer Research Program, Oklahoma Medical Research Foundation, Oklahoma City, OK 73104.
4 Address correspondence and reprint requests to Dr. Kensuke Miyake, Division of Infectious Genetics, Institute of Medical Science, University of Tokyo, 4-6-1 Shirokanedai, Minatoku, Tokyo 108-8639, Japan. E-mail address: kmiyake{at}ims.u-tokyo.ac.jp
5 Abbreviations used in this paper: TI, T independent; MZ, marginal zone; RP105, radioprotective 105; PNA, peanut agglutinin; TNP. trinitrophenyl; HPRT, hypoxanthine-guanine phosphoribosyltransferase; GC, germinal center; TD, thymus dependent.
Received for publication August 30, 2004. Accepted for publication March 29, 2005.
References
Janeway, C. A., Jr, R. Medzhitov. 2002. Innate immune recognition. Annu. Rev. Immunol. 20: 197-216.
Gordon, S.. 2002. Pattern recognition receptors: doubling up for the innate immune response. Cell 111: 927-930.
Bendelac, A., M. Bonneville, J. F. Kearney. 2001. Autoreactivity by design: innate B and T lymphocytes. Nat. Rev. Immunol. 1: 177-186.
Baumgarth, N., O. C. Herman, G. C. Jager, L. E. Brown, L. A. Herzenberg, J. Chen. 2000. B-1 and B-2 cell-derived immunoglobulin M antibodies are nonredundant components of the protective response to influenza virus infection. J. Exp. Med. 192: 271-280.
Ochsenbein, A. F., T. Fehr, C. Lutz, M. Suter, F. Brombacher, H. Hengartner, R. M. Zinkernagel. 1999. Control of early viral and bacterial distribution and disease by natural antibodies. Science 286: 2156-2159.
Takeda, K., S. Akira. 2004. Microbial recognition by Toll-like receptors. J. Dermatol. Sci. 34: 73-82.
Medzhitov, R., C. A. Janeway, Jr. 2002. Decoding the patterns of self and nonself by the innate immune system. Science 296: 298-300.(Yoshinori Nagai2,3,*, Tos)