HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by He, B. Articles by Cerutti, A. Search for Related Content PubMed PubMed Citation Articles by He, B. Articles by Cerutti, A.

CpG DNA Induces IgG Class Switch DNA Recombination by Activating Human B Cells through an Innate Pathway That Requires TLR9 and Cooperates with IL-10¹

Bing He^*, Xugang Qiao^* and Andrea Cerutti^2,*,

^* Department of Pathology, Weill Medical College, and Immunology Graduate Program, Weill Graduate School of Medical Sciences, Cornell University, New York, NY 10021

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
TLRs are pattern recognition receptors that initiate innate immune responses. TLR9 detects microbial DNA with hypomethylated CpG motifs and in humans is preferentially expressed by IFN--producing plasmacytoid dendritic cells and B cells. In addition to favoring IFN- release, TLR9 signals B cell activation, proliferation, and IgM production. Recent findings suggest that CpG DNA-TLR9 interaction plays a key role in systemic lupus erythematosus and rheumatoid arthritis, two autoimmune disorders characterized by dysregulated production of DNA-reactive IgG. We show that CpG DNA initiates germline C1, C2, and C3 gene transcription by activating B cells through a TLR9-mediated NF-B-Rel-dependent innate pathway that cooperates with IL-10 through STAT proteins and IFN-responsive factors. This pathway is inhibited by chloroquine, a drug that attenuates the clinical manifestations of IgG-mediated autoimmune disorders. Germline C gene transcription is associated with up-regulation of activation-induced cytidine deaminase, a key element of the B cell class switch-inducing machinery, and is followed by class switch DNA recombination from C_µ to C1, C2, and C3. Subsequent IgG production requires additional signals from BCR and a B cell-activating factor of the TNF family (BAFF), produced by dendritic cells upon exposure to IFN-. Our findings suggest that CpG DNA-TLR9 interaction may be important to initiate or amplify early T cell-independent IgG responses against pathogens. This implies that CpG DNA released during infections may exacerbate autoimmunity by stimulating autoreactive B cells to switch from an IgM to a more pathogenic IgG isotype.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
By substituting the Ig C_H region of IgM and IgD with that of IgG, IgA, or IgE, class switching endows Abs with new effector functions that enhance the clearance of invading pathogens (1). Germinal center B cells undergo IgH class switching through class switch DNA recombination (CSR), ³ a process that promotes recombination of the switch region µ(S_µ) 5' of the C_µ gene with a targeted downstream S, S, or S region 5' of C, C, and C, respectively (2, 3). Complex microbial proteins initiate CSR by up-regulating the TNF family member CD40L on CD4⁺ T cells (1). Engagement of CD40 on B cells by CD40L activates NF-B-Rel (4), which initiates germline I_H-S-C_H transcription by binding cis-acting B elements within intronic I_H promoters 5' of S regions (1). Germline transcription is further enhanced by T cell-derived IL-4 through STAT6, which binds a IFN--activated sequence (GAS) adjacent to B sites (1). By inducing chromatin opening, germline transcription facilitates the recruitment of activation-induced cytidine deaminase (AID), an enzyme that initiates CSR by introducing DNA breaks within S regions (5). Subsequent looping-out deletion of the DNA between the two recombined S regions juxtaposes the somatically recombined V_HDJ_H exon, which encodes the Ag-binding Ig V_H region, to a new C_H gene (3).

In addition to inducing CSR, T cell-dependent (TD) Ags stimulate germinal center B cells to undergo V(D)J gene somatic hypermutation, a process that increases the affinity of Abs for Ag (2, 5). Although generating high affinity Ig-secreting plasma cells and long-lived memory B cells (6), TD B cell responses require 5–7 days, which is too much of a delay to neutralize quickly replicating pathogens. To compensate for this limitation, marginal zone B cells and mucosal B-1 cells rapidly undergo T cell-independent (TI) Ab production in response to highly conserved microbial Ags with repetitive structure (7, 8). These compounds include type 1 TI Ags, such as bacterial wall LPS, and type 2 TI Ags, such as bacterial capsule polysaccharides (9). Both TI-1 and TI-2 Ags encompass pathogen-associated molecular patterns (PAMPs) (10), which stimulate marginal zone and B-1 cells by cross-linking poorly diversified surface BCRs encoded by a restricted set of V(D)J genes carrying no or few mutations (11). TI Ab production also requires interaction of B cells with dendritic cells (DCs) and macrophages (8, 12). These innate immune cells recognize PAMPs through pattern recognition receptors, including TLR family members (13).

In addition to triggering immediate innate responses (10), PAMP-TLR interaction favors TI B cell Ab production. For instance, engagement of TLR4 by LPS stimulates myeloid DCs to up-regulate B cell-activating factor of the TNF family (BAFF) (10, 14), which cooperates with signals emanating from BCR to stimulate B cell survival, CSR, and Ab production (15, 16, 17). Myeloid DCs also up-regulate BAFF upon exposure to IFN- (16), a key innate cytokine released by circulating plasmacytoid DCs upon TLR9 engagement by microbial products, including DNA with unmethylated CpG motifs (18, 19, 20, 21). Unlike TLR4, which is expressed on the cell surface (10), TLR9 is expressed in the endoplasmic reticulum and signals from an endosomal compartment upon CpG DNA internalization (22, 23). Not only do PAMPs stimulate B cells through BAFF, but they also directly activate B cells via TLRs. This is exemplified by LPS, which triggers mouse B cell CSR and Ab production by engaging TLR4 (1, 10). Human B cells lack TLR4, but express TLR9, and undergo proliferation and IgM production in response to CpG DNA (24, 25). The role of CpG DNA in human B cell CSR remains elusive.

Unmethylated microbial CpG DNA is similar to certain CpG DNA islands within the mammalian genome (22, 26, 27). Hypomethylated genomic CpG DNA is increased in subjects with systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA) (28), two autoimmune disorders characterized by abnormal B cell reactivity to a relatively restricted set of self-Ags, including DNA and DNA-bound proteins (29, 30, 31). Recent studies suggest that dual TLR9 and BCR engagement by endogenous CpG DNA released from dying cells and CpG DNA-IgG complexes stimulates autoreactive B cells to produce DNA-reactive IgM and IgM rheumatoid factors (RFs), which recognize DNA-bound IgG (32, 33). Considering that anti-DNA IgG is usually more abundant than anti-DNA IgM and that RF-expressing B cells produce IgG in addition to IgM (29, 30), it is conceivable that CpG DNA stimulates autoreactive IgG CSR in addition to IgM production.

CD40L and BAFF induce CSR through p50, p52, p65, c-Rel, and RelB (1, 16). These NF-B-Rel proteins are retained in a cytoplasmic inactive form by IB, an inhibitor of NF-B (34, 35). By recruiting TNF receptor-associated factors (TRAFs) to their receptors, CD40L and, to a lesser extent, BAFF activate IB kinase (IKK) (4, 17), which is part of an IKK complex that includes two catalytic and subunits and a regulatory subunit (34, 35). Phosphorylation of IB by IKK is followed by IB degradation and p50, p65, and c-Rel nuclear translocation (34, 35). CD40L and, to a greater extent, BAFF also activate IKK through NF-B-inducing kinase (NIK) (34, 35). Phosphorylation of p100 by IKK is followed by p100 processing to p52, which then translocates to the nucleus in association with RelB (34, 35). Similarly to CD40 and BAFF receptors, TLRs activate p50, p52, p65, c-Rel, and RelB through NIK and IKK (10). These kinases are turned on upon recruitment of MyD88, IL-1R-associated kinases (IRAKs), and TRAF6 to a cytoplasmic TLR/IL-1R (TIR) domain (36, 37, 38). This prompted us to hypothesize that TLR9 signaling initiates IgG CSR through NF-B-Rel.

We show that CpG-containing oligodeoxynucleotides (ODNs) and bacterial DNA initiate germline C gene transcription in human B cells by turning on a TLR9 pathway that activates NF-B-Rel, STATs, and IFN-responsive factors (IRFs) in cooperation with IL-10. CpG DNA-induced germline C gene transcription is inhibited by chloroquine, a compound used to treat patients with SLE and RA (39), and ultimately leads to CSR from C_µto C1, C2, and C3. Subsequent IgG production requires additional B cell stimulation through BCR engagement, BAFF, or CD40L.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Cells and reagents

Peripheral blood IgD⁺IgM⁺ naive B cells were obtained as previously described (16). 2E2 B cells are characterized by high TLR9 expression levels and were obtained by subcloning CL-01, an IgD⁺IgM⁺ B cell line that undergoes CSR upon in vitro exposure to appropriate stimuli (40). CL-01 subclones were obtained by limiting dilution and tested for TLR9 expression through semiquantitative RT-PCR. Purified phosphorothioate-modified ODNs (Operon Technologies, Alameda, CA; lowercase letters represent nuclease-resistant phosphorothioate linkage, uppercase letters represent phosphorodiester linkage 3' of the base, and underlined letters represent CpG dinucleotides), including 5'-tcgtcgttttgtcgttttgtcgtt-3' ODN-2006/type B CpG DNA and 5'-ggGGGACGATCGTCgggggg-3' ODN-2216/type-A CpG DNA, were used at 5 µg/ml. Control 5'-tgctgcttttgtgcttttgtgctt-3' GpC DNA and 5'-GCTTGATGACTCAGCCGGAA-3' non-CpG DNA (Operon Technologies) were used at 5 µg/ml. Escherichia coli DNA was purified by extraction with phenol/chloroform/isoamyl alcohol and ethanol precipitation, suspended in DNase-free distillated water, and used at 200 µg/ml. In some experiments E. coli DNA was digested with 50 µg/ml DNase I. Recombinant BAFF (Alexis Biochemicals, San Diego, CA), CD40L (Immunex, Seattle, WA), IL-2 (from Dr. K. A. Smith, Cornell University, New York, NY), IL-4 (Schering-Plough, Kenilworth, NJ), IL-10, IL-15, LPS (Sigma-Aldrich, St. Louis, MO), IL-6, and IL-13 (R&D Systems, Minneapolis, MN) were used at 500 ng/ml, 500 ng/ml, 50 U/ml, 200 U/ml, 200 ng/ml, 100 ng/ml, 1 µg/ml, 5 ng/ml, and 10 ng/ml, respectively, unless otherwise indicated. Immunobead reagent (Irvine Scientifics, Camarillo, CA), which binds both H and L chains of human Igs, was used at 2 µg/ml to mimic extensive BCR cross-linking by a TI Ag. Bafilomycin A and chloroquine (Sigma-Aldrich) were used at 10 nM and 20 ng/ml, respectively.

Flow cytometry and ELISAs

B cell surface CD19 and IgG were labeled with FITC- and PE-conjugated Abs (BD Pharmingen, San Diego, CA). Cells (10⁴) were acquired using a FACSCalibur analyzer. IgG ELISA was performed as previously described (16, 40).

RT-PCRs

cDNA was reverse transcribed from 3 µg of total RNA. TLR9 was RT-PCR amplified for 25 cycles using forward primer 5'-ACAACAACATCCACAGCCAAGTGTC-3' and reverse primer 5'-AAGGCCAGGTAATTGTCACGGAG-3'. I1-C1 (603 bp), I2-C2 (597 bp), I3-C3 (670 bp), I4-C4 (411 bp), I-C (409 bp), I_µ-C_µ (537 bp), I1/2-C_µ (557 bp), I3-C_µ (608 bp), AID (382 bp), and -actin (593 bp) transcripts were amplified for 25 cycles as previously described (16).

Genomic DNA PCRs

Genomic DNA was extracted by using the QIAmp DNA Mini Kit (Qiagen, Valencia, CA). Total S-S_µ switch circles and -actin were amplified from 500 ng of genomic DNA as previously described (16). The conditions were denaturation 1 min at 94°C, annealing 1 min at 68°C, and extension 4 min at 72°C for two rounds of 30 cycles each. Before each PCR, DNA was denatured for 5 min at 94°C. The identity of PCR products with switch circles was confirmed by DNA sequencing.

Southern blots

PCR products were fractionated onto agarose gels, transferred overnight to nylon membranes, and hybridized with radiolabeled probes as described. Switch circles were hybridized with a probe recognizing the recombined S_µ region (16). I-C_µ transcripts were hybridized with a probe encompassing nt 1–250 of the first C_µ exon. I-C transcripts were hybridized with a 5'-CAGGGGGAAGACCGATGG-3' oligoprobe recognizing a consensus C sequence. Finally, I-C transcripts were hybridized with a 5'-TCCACACAGAGCCCATC-3' oligoprobe recognizing C.

Plasmids

Genomic DNA fragments encompassing I1, I2, I3, I4, and I promoters were inserted into a promoterless pGL3-Basic luciferase (Luc) reporter vector (Promega, Madison, WI) as previously reported (40, 41, 42). B1^–, B2^–, IFN-stimulated responsive element (ISRE)-B3^–, B cell-specific activating protein-binding site (BSAP^–), and GAS^– I3 mutants were generated through a PCR-based strategy that disrupts the targeted motif by replacing six nucleotides of the wild-type sequence with a SacI restriction site (40). Briefly, 5' and 3' DNA fragments were amplified using appropriate primers with 5' overhangs containing KpnI (sense)/SacI (antisense) and SacI (sense)/BglII (antisense) sites, respectively. Then, DNA fragments were digested with KpnI/SacI, and SacI/BglII, respectively, and cloned into the KpnI/BglII-digested pGL3-Basic vector. Reporter vectors driven by two B sites from the Ig gene (from Dr. H.-C. Liou, Cornell University, New York, NY), three GAS sites from the Ly-6/E gene (from Dr. J. J. Zhang, Cornell University), four GAS sites from the CD23b gene (from Dr. J. J. Zhang), and ISRE sites from the ISG15 gene (from Dr. K. Horvath, Mount Sinai Medical Center, New York, NY) are reported elsewhere (43, 44, 45). Expression vectors for TLR9, Toll-like interacting protein (InvivoGen, San Diego, CA), dominant negative (DN)-IB (S32/36A; from Dr. E. Cesarman, Cornell University), DN-MyD88 (residues 152–296), DN-IRAK1 (residues 1–208; Tularik, San Francisco, CA), DN-TRAF2 (residues 248–501), DN-TRAF6 (residues 301–530; ScienceReagents, El Cajon, CA), DN-IKK (K44M), DN-IKK (K49A), DN-NIK (K429/430A; from Dr. J.-D. Li, University of South California, Los Angeles, CA), DN-STAT1 (Tyr701F), DN-STAT3 (Y705F; from Drs. J. E. Darnell, Jr., and J. Bromberg, Rockefeller University, New York, NY), and DN-IRF4 (residues 1–150; from Dr. A. B. Pernis, Columbia University) were described previously (18, 20, 36, 44, 46, 47, 48). TIR-TLR9 (residues 1–870) was PCR-amplified from a TLR9 plasmid (InvivoGen) using forward primer 5'-GGGGATATCTACTAGATTTATCAAAAAGAG-3' and reverse primer 5'-GGGGCTAGCTAGGGCAGGGCATCCTCATC-3'.

Luciferase reporter assays

2E2 B cells (20 x 10⁶/ml) were transfected with 40 µl of plasmid DNA-TE solution containing 10 µg of I-Luc, I-Luc, ISRE-Luc, or STAT-Luc plasmid expressing firefly luciferase and 200 ng of control pRL-TK plasmid expressing Renilla luciferase under control of the thymidine kinase promoter (Promega). Similar transfections were performed in the presence or the absence of 10 µg of empty expression plasmid or 10 µg of DN-MyD88, DN-IRAK1, DN-TRAF6, DN-TRAF2, Toll-interacting protein (TOLLIP), IB, DN-IKK, DN-IKK, DN-NIK, DN-STAT1, DN-STAT3, or DN-IRF4 expression plasmid. Electroporations were performed at 625 V/cm and 950 µF using a Gene Pulser II apparatus (Bio-Rad, Hercules, CA). After electroporation, B cells were resuspended in complete medium (2 x 10⁶/ml), split into aliquots, and cultured with or without stimuli. After 48 h, firefly and Renilla luciferase activities were measured using the dual luciferase assay system (Promega) to assess promoter activity and transfection efficiency, respectively. Luciferase activity was expressed as relative light units normalized to a cotransfected pRL-TK control plasmid or as fold induction, i.e., normalized luciferase activity of extracts from stimulated B cells/normalized luciferase activity of extracts from unstimulated B cells.

Immunoblots

Equal amounts of total, cytoplasmic, or nuclear proteins (10 µg) were fractionated onto a 10% SDS-polyacrylamide gel and transferred to nylon membranes (Bio-Rad). After blocking, membranes were probed with primary Abs to TLR9 (InvivoGen), p50, p52, p65, c-Rel, RelB, IB, STAT1, STAT2, STAT3, STAT6, IRF1, IRF9, actin, Oct1 (Santa Cruz Biotechnologies, Santa Cruz, CA), IRF4 (from Dr. A. B. Pernis, Columbia University, New York, NY), (Ser32^p/Ser36^p)-IB, (Tyr701^p)-STAT1, (Tyr705^p)-STAT3, (Tyr641^p)-STAT6 (Cell Signaling Technology, Beverly, MA), or (Tyr689^p)-STAT2 (Upstate Biotechnology, Lake Placid, NY). Membranes were then washed and incubated with an appropriate secondary Ab (Santa Cruz Biotechnology). Proteins were detected with an ECL detection system (Amersham Biosciences, Little Chalfont, U.K.).

EMSAs

Nuclear proteins were extracted from 5 x 10⁶ cells (40). Double-stranded oligonucleotide probes (consensus sequences are underlined or, when partially overlapping, underlined and in bold) encompassing B2 cis-I3 (residues –216 to –192, 5'-CCTCTGACACAGAAACCACCAGAAG-3'), ISRE-B3 cis-I3 (residues –198 to –176, 5'-CCAGAAGAAAAGGGAACTTCAGG-3') and GAS cis-I3 (residues –96 to –69, 5'-GAGCTGTGATTTCCTAGGAAGACAAA-3') sites were end-labeled with [-³²P]ATP by T4 kinase and used at 20,000 cpm in each EMSA reaction. A control Oct1-binding oligonucleotide (TGTCGAATGCAAATCACTAGAA; Santa Cruz Biotechnologies) was labeled in a similar fashion. Reaction samples were prepared as described, incubated at room temperature for 15 min, and electrophoresed through a 6% nondenaturing polyacrylamide gel. The compositions of DNA-bound protein complexes were determined by incubating reaction mixtures with 1 µl of a DNA-binding inhibiting/supershifting Ab to p65, c-Rel, Rel-B, p50, p52, STAT1, STAT2, STAT3, STAT6, IRF1, IRF9 (Santa Cruz Biotechnology), or IRF4 (from Dr. A. B. Pernis) for 15 min at room temperature before adding the probe.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
B cells up-regulate TLR9 expression upon exposure to IL-10

IL-10 and IL-4 are two major CSR-inducing cytokines (1, 16). Recent studies indicate that CpG DNA cooperates with autocrine IL-10 to enhance IgG production in B cells (25, 49). Additional reports show that CpG DNA inhibits IgG and IgE production in B cells exposed to IL-4 (49, 50). These observations prompted us to verify whether TLR9 expression differs in B cells exposed to IL-10 or IL-4. Human IgD⁺IgM⁺ naive B cells up-regulated TLR9 transcripts and proteins as early as 1 h after incubation with IL-10 (Fig. 1A). This up-regulation peaked after 48 h and progressively increased upon B cell exposure to growing amounts of IL-10 (Fig. 1B). Conversely, naive B cells down-regulated TLR9transcripts and proteins as early as 1 h after exposure to IL-4. This down-regulation was more evident after 48 h and progressively increased upon B cell exposure to growing amounts of IL-4. Thus, IL-10 and IL-4 may control B cell responses to CpG DNA by up- and down-modulating TLR9 expression, respectively.

View larger version (62K): [in this window] [in a new window]   FIGURE 1. IL-10 up-regulates TLR9 expression in human B cells. A, TLR9 and actin (loading control) transcripts and proteins in naive B cells incubated with IL-10 or IL-4 for different time points. Transcripts were RT-PCR amplified for 30 cycles. B, TLR9 and -actin transcripts in naive B cells incubated with different doses of IL-10 or IL-4 for 2 days. Transcripts were RT-PCR amplified for 25 cycles. Data in A and B depict one of three experiments yielding similar results.

CpG DNA up-regulates germline C1, C1, and C3 transcription in B cells by cooperating with IL-10

CSR from C_µ to a targeted downstream C_H gene is preceded by germline I_H-C_H transcription (1). This early CSR event requires activation of the I_H promoter 5' of the targeted C_H gene and yields a germline I_H-C_H transcript encompassing a noncoding I_H exon (3). To verify whether CpG DNA initiates germline I-C and I-C gene transcription, primary IgD⁺IgM⁺ naive B cells or IgD⁺IgM⁺ 2E2 B cells, a subclone of the CL-01 B cell line (40), were incubated for 48 h with a CpG DNA-containing ODN (ODN-2006). Cultures were conducted in the presence or the absence of IL-10, which facilitates switching to IgG1, IgG2, and IgG3, and with or without IL-4, which favors switching to IgG1, IgG2, IgG3, IgG4, and IgE (1, 51). When exposed to CpG ODN-2006, naive B cells up-regulated germline I1-C1, I2-C2, and I3-C3, but not I4-C4 and I-C transcripts (Fig. 2), whereas 2E2 B cells activated I1, I2, and I3, but not I4 and I promoters (Fig. 3). The combination of CpG ODN-2006 and IL-10 up-regulated I1-C1, I2-C2, and I3-C3 and activated I1, I2, and I3 more effectively than CpG ODN-2006 or IL-10 alone. Control GpC ODN-2006 with inverted CpG motifs did not up-regulate I1-C1, I2-C2, and I3-C3, nor did it enhance the IL-10-induced up-regulation of I1-C1, I2-C2, and I3-C3. Similarly, GpC ODN-2006 did not activate I1, I2, and I3 or enhance IL-10-mediated activation of I1, I2, and I3.

View larger version (66K): [in this window] [in a new window]   FIGURE 2. CpG DNA cooperates with IL-10 to up-regulate germline I1-C1, I2-C2, and I3-C3 transcripts in human B cells. Germline I1-C1, I2-C2, I3-C3, I4-C4, I-C, and Iµ-Cµtranscripts in naive B cells cultured with or without CpG DNA ODN-2006, GpC ODN-2006, IL-10, and/or IL-4 for 2 days. Data depict one of three experiments yielding similar results.

View larger version (20K): [in this window] [in a new window]   FIGURE 3. CpG DNA cooperates with IL-10 to induce germline C1, C2, and C3 gene transcription in human B cells. Induction of I1-Luc, I2-Luc, I3-Luc, I4-Luc, and I-Luc in 2E2 B cells cultured with or without CpG DNA ODN-2006, GpC ODN-2006, IL-10, and/or IL-4 for 2 days. Data correspond to one of three experiments yielding similar results (bars represent the SD of triplicate determinations).

View larger version (21K): [in this window] [in a new window]   FIGURE 4. CpG DNA induces germline C3 gene transcription and NF-B-Rel activation in human B cells through a highly specific mechanism. A, Induction of I3-Luc in 2E2 B cells cultured with or without CpG DNA ODN-2006, IL-2, IL-6, IL-10, IL-12, IL-13, and IL-15 for 2 days. B, Induction of I3-Luc and B-Luc in 2E2 B cells cultured with or without non-CpG DNA, GpC ODN-2006, CpG ODN-2216, CpG ODN-2006, bacterial DNA, DNase I-treated bacterial DNA, LPS, or htCD40L for 2 days. Data correspond to one of three experiments yielding similar results (bars represent the SD of triplicate determinations).

CpG DNA and IL-10 up-regulate AID and induce CSR from C_µ to C1, C2, and C3 in B cells

Additional experiments verified whether CpG DNA triggers IgG CSR in B cells. CSR requires AID and generates a circular DNA upon looping-out deletion of the IgH locus lying between S_µ and the targeted downstream S region, such as S (Fig. 5A). The resulting S-S_µ switch circle transcribes a chimeric I-C_µ circle transcript, which includes the I promoter 5' of the targeted C gene, a noncoding I exon, and the C_µ exon. Both switch circles and circle transcripts constitute specific markers of ongoing CSR (5). Naive B cells exposed to IL-10 and/or control GpC ODN-2006 for 4 days did not contain S-S_µ switch circles (Fig. 5B), a byproduct of IgG CSR, nor did they express I1/2-C_µ and I3-C_µ circle transcripts, a byproduct of IgG1/IgG2 CSR and IgG3 CSR, respectively. Similarly treated B cells lacked AID transcripts. In contrast, naive B cells induced S-S_µ, I1/3-C_µ, I3-C_µ, and AID upon exposure to CpG ODN-2006, and this induction further increased when CpG ODN-2006 was combined with IL-10. Conversely, IL-4 attenuated the expression of S-S_µ, I1/3-C_µ, I3-C_µ, and AID in B cells exposed to CpG ODN-2006 (not shown). These findings indicate that CpG DNA triggers CSR to C1, C2, and C3 by cooperating with IL-10.

View larger version (31K): [in this window] [in a new window]   FIGURE 5. CpG DNA cooperates with IL-10 to induce CSR from Cµ to C1, C2, and C3 in human B cells. A, Diagram of CSR from Cµ to a downstream C gene. Ovals indicate S regions; rectangles are VHDJH, IH and CH exons; V-shaped lines indicate splicing. B, S-SµSCs, genomic -actin, I1/2-Cµ2 and I3-CµCTs, and AID and -actin transcripts in naive B cells cultured with or without CpG DNA ODN-2006, GpC ODN-2006, IL-10, and/or IL-4 for 4 days. C and D, Surface IgG expression and IgG secretion by naive B cells cultured with or without CpG DNA ODN-2006, GpC ODN-2006, IL-10, IL-4, anti-BCR, BAFF, and/or CD40L for 8 days. n.d., not determined. Data in B–D represent one of three independent experiments yielding similar results (bars represent the SD).

To verify whether CpG DNA up-regulates IgG production, naive B cells were exposed to CpG ODN-2006 and IL-10 in the presence or the absence of known CpG DNA-costimulating molecules, including anti-BCR and CD40L (24, 25, 49). Cultures also included BAFF, which is produced by myeloid DCs upon exposure to CpG DNA-induced cytokines, including IFN- (16, 19). Naive B cells up-regulated surface IgG upon exposure to CpG ODN-2006 (but not GpC ODN-2006) and IL-10, anti-BCR, BAFF, or CD40L for 5 days (Fig. 5C). The proportion of IgG-expressing B cells further increased when CpG ODN-2006 was combined with either IL-10 and anti-BCR or IL-10 and BAFF, or IL-10 and CD40L. This increase was incremental over that induced by either IL-10 and anti-BCR or IL-10 and BAFF, or IL-10 and CD40L. Finally, CpG ODN-2006 and IL-10 elicited IgG secretion only in B cells concomitantly exposed to anti-BCR, BAFF, or CD40L for 8 days (Fig. 5D). This suggests that CpG DNA requires IL-10 as well as additional TI or TD stimuli, including BAFF or CD40L, to stimulate IgG production in Ag-activated B cells.

CpG DNA requires TLR9, MyD88, IRAK1, and TRAF6 to induce germline C3 gene transcription

CD40 activates initiates NF-B-Rel-dependent germline C transcription by recruiting TRAFs, including TRAF2 and TRAF6 (47, 52). Because TLR9 activates NF-B-Rel by recruiting MyD88, IRAK, and TRAF6 through a cytoplasmic TIR domain (10), it was hypothesized that the TLR9-MyD88-IRAK-TRAF6 axis is crucial for the initiation of IgG CSR by CpG DNA. Consistent with this, enforced expression of TIR-less TLR9 (TIR-TLR9), DN-MyD88, DN-IRAK1, or TOLLIP, an adapter protein that negatively regulates TLR signaling (53), inhibited the activation of I3-Luc in 2E2 B cells exposed to CpG ODN-2006 for 48 h (Fig. 6A). In these cells, TIR-TLR9, DN-MyD88, DN-IRAK1, or TOLLIP inhibited NF-B-Rel-dependent activation of a minimal B-Luc reporter vector. As expected, DN-MyD88, DN-IRAK1, or TOLLIP did not inhibit activation of I3-Luc and B-Luc by CD40L. Furthermore, a DN form of TRAF2 attenuated activation of I3-Luc and B-Luc by CD40L, but not by CpG ODN-2006. Finally, a DN form of TRAF6 attenuated the activation of I3-Luc and B-Luc by either CpG ODN-2006 or CD40L. These findings suggest that CpG DNA activates I promoters as well as NF-B-Rel through an innate pathway that requires TLR9, MyD88, IRAK1, and TRAF6.

View larger version (22K): [in this window] [in a new window]   FIGURE 6. CpG DNA requires TLR9, MyD88, IRAK, and TRAF6 to activate the I3 promoter in human B cells. A, Induction of I3-Luc and B-Luc in 2E2 B cells transfected with empty vector (control), TIR-TLR9, DN-MyD88, DN-IRAK1, DN-TRAF6, DN-TRAF2, or TOLLIP and subsequently incubated with CpG DNA ODN-2006 or CD40L for 2 days. Data are presented as a percentage of the yield (fold activation with expression vector/fold activation without expression vector x 100). B, Induction of I3-Luc and B-Luc in 2E2 B cells incubated for 48 h with CpG DNA ODN-2006 or CD40L in the presence or the absence of control H2O, bafilomycin A (BFLM) or chloroquine (CHLQ). H2O, BFLM or CHLQ were added 6 h before CpG DNA ODN-2006 or CD40L. Data are presented as a percentage of the yield (fold activation with H2O, BFLM or CHLQ/fold activation without H2O, BFLM or CHLQ x 100). Data in A and B correspond to one of three experiments yielding similar results (bars represent the SD of triplicate determinations).

CpG DNA triggers germline C3 gene transcription through a chloroquine-sensitive pathway

Innate immune cells internalize CpG DNA and initiate TLR9 signaling through a mechanism that requires endosomal maturation and acidification (23). Consistent with this, inhibitors of endosomal maturation and acidification, such as bafilomycin-A and chloroquine, attenuate cell activation by CpG DNA (32, 39). Chloroquine and its derivatives might also block CpG DNA signaling by preventing CpG DNA binding to as yet elusive surface receptors (22, 26). Bafilomycin-A and chloroquine attenuated the activation of both I3-Luc and B-Luc in 2E2 B cells exposed to CpG ODN-2006 (Fig. 6B). In contrast, bafilomycin-A and chloroquine did not affect the activation of I3-Luc and B-Luc in 2E2 B cells exposed to CD40L. These data suggest that CpG DNA activates I promoters and NF-B-Rel through a chloroquine-sensitive B cell pathway.

CpG DNA and IL-10 trans-activate the I3 promoter through B, ISRE, and GAS cis-acting elements

Germline C gene transcription requires key cis-acting DNA regulatory elements located within the evolutionarily conserved sequence (ECS) of I promoters (Fig. 7A). The human ECS includes three B sites (B1, B2, and B3), a BSAP, and a STAT-binding GAS (1, 40). CD40L-induced C gene transcription involves binding of NF-B-Rel and BSAP to B1, B2, and BSAP sites, whereas IL-4-induced C gene transcription requires binding of STAT6 to the GAS site. This latter can bind STAT1, STAT2, STAT3, and IRF4 in addition to STAT6 (44, 46, 54, 55, 56). Interestingly, the ECS contains also an ISRE site, which partially overlaps with B3 (ISRE-B3) (40). In general, ISRE binds the STAT1-STAT2-IFR9 complex as well as IRFs, including IRF1 and IRF4 (44, 56). STATs and IRFs are induced by several receptors, including TLRs and IL-10Rs, and cooperate with NF-B-Rel proteins to modulate B cell activation, differentiation, and IgG production (44, 46, 50, 54, 55, 56, 57, 58, 59). Thus, it was postulated that CpG DNA and IL-10 initiate C transcription upon NF-B-Rel, STAT, and IRF binding to cooperative B, ISRE, and GAS cis-I sites. To verify this, 2E2 B cells were transfected with reporter vectors carrying wild-type I3 or mutated I3 with targeted disruptions of B1, B2, ISRE-B3, BSAP, and GAS (43). In 2E2 B cells incubated with medium alone, wild-type and mutated I3 promoters displayed similar transcriptional activity (Fig. 7B). Disruption of B1 or BSAP did not affect activation of I3 by CpG ODN-2006 and/or IL-10. In contrast, disruption of B2, B3-ISRE, or GAS impaired activation of I3 by CpG ODN-2006 and/or IL-10. These findings suggest that CpG DNA and IL-10 require B2, B3, ISRE, and GAS, but not B1 and BSAP, cis-acting motifs to initiate germline C3 transcription.

View larger version (29K): [in this window] [in a new window]   FIGURE 7. CpG DNA and IL-10 require adjacent B2, ISRE-B3, and GAS cis-acting motifs to trans-activate the I3 promoter in human B cells. A, Diagram of the I3 promoter. Boxes depict cis-B, ISRE, BSAP, and GAS motifs within the ECS. –239 and –65 residues span the ECS, and the turned arrow indicates the transcription initiation site. P3 indicates the promoter sequence 5' of the noncoding I3 exon. B, Transcriptional activity of wild-type (wt) I3 or mutated B1–, B2–, ISRE-B3–, BSAP–, and GAS– I3 Luc constructs in 2E2 B cells cultured with or without CpG DNA ODN-2006 and/or IL-10 for 2 days. The luciferase activity of 2E2 cells incubated with medium alone (control) is expressed as relative luciferase units (RLU). Data correspond to one of three independent experiments (bars represent the SD of triplicate determinations).

CpG DNA and IL-10 activate NF-B-Rel

Given the key role of B2 and ISRE-B3 in the activation of I3 by CpG ODN-2006 and IL-10, it was postulated that CpG DNA and IL-10 synergistically induce C gene transcription by cooperatively activating NF-B-Rel. Initial experiments tested the induction of transcriptionally active NF-B-Rel. 2E2 B cells activated a minimal B-Luc reporter vector upon exposure to CpG ODN-2006, but not GpC ODN-2006, for 48 h (Figs. 3B and 8A). When combined, CpG ODN-2006 and IL-10 induced B-Luc more than either CpG ODN-2006 or IL-10 alone (Fig. 8A). Additional studies evaluated NF-B-Rel nuclear translocation. When exposed to CpG ODN-2006 for 3 h, naive B cells up-regulated IB phosphorylation and down-regulated total IB, a hallmark of increased IB degradation (Fig. 8B). After 6 h, CpG ODN-2006 up-regulated the expression of nuclear p65, p50, c-Rel, p52, and RelB (Fig. 8C). Also, IL-10 activated B-Luc and up-regulated IB phosphorylation, IB degradation, and NF-B-Rel nuclear translocation, although less than CpG ODN-2006. When combined, CpG ODN-2006 and IL-10 induced more IB phosphorylation, IB degradation, and NF-B-Rel nuclear translocation than CpG ODN-2006 or IL-10 alone. Thus, CpG DNA and IL-10 cooperatively activate p65, p50, c-Rel, p52, and RelB in B cells.

View larger version (34K): [in this window] [in a new window]   FIGURE 8. CpG DNA and IL-10 induce transcriptionally active NF-B-Rel and up-regulate NF-B-Rel binding to B2 cis-I3 in human B cells. A, B-Luc activity in 2E2 B cells incubated with or without CpG ODN-2006, IL-10, and/or IL-4 for 2 days. B and C, Cytoplasmic pIB, IB, and actin proteins as well as nuclear p65, p50, c-Rel, p52, and RelB proteins in naive B cells cultured as in A for 1 h. The ubiquitous nuclear protein octamer 1 (Oct1) was used as loading control. D, Aligned –232/–176 I3, I1, and I2 DNA sequences. The boxed sequence corresponds to the B2 cis-I3 motif. Left gels show nuclear protein binding to an oligonucleotide encompassing B2 cis-I3 or a consensus Oct1-binding motif. Nuclear extracts were from naive B cells cultured for 6 h as described in A. Italic letters indicate protein-DNA complexes. Right gel shows protein-DNA complexes from naive B cells activated with CpG DNA after nuclear protein incubation with a cold B2 cis-I3 probe or Abs to p50, p52, p65, c-Rel, and RelB. Asterisks indicate lanes in which complexes are attenuated or supershifted by the Ab. Data in A–D correspond to one of three experiments yielding similar results (bars represent the SD of triplicate determinations).

CpG DNA and IL-10 up-regulate NF-B-Rel binding to B2 and ISRE-B3 cis-I3

We next evaluated whether CpG DNA and IL-10 up-regulate the binding of nuclear NF-B-Rel to B2 cis-I3. Naive B cells up-regulated the binding of complexes a, b, c, d, e, f, and g to B2 cis-I3 upon incubation with CpG ODN-2006 for 6 h (Fig. 8D). These complexes were attenuated or supershifted by preincubating nuclear proteins with Abs to p65, p50 and p65, p50 and c-Rel, p52 and RelB, p50 and RelB, p50 and p52, and p50, respectively. IL-10 up-regulated complexes a, b, c, d, e, f, and g to B2 less than CpG ODN-2006, whereas CpG ODN-2006 and IL-10 up-regulated complexes a, b, c, d, e, f, and g more than CpG ODN-2006 or IL-10 alone. Additional experiments evaluated whether CpG DNA and IL-10 up-regulate the binding of nuclear NF-B-Rel to ISRE-B3 cis-I3. Naive B cells up-regulated the binding of nuclear complexes a, b, and c to ISRE-B3 cis-I3 upon incubation with CpG ODN-2006 for 6 h (Fig. 9A). These complexes were attenuated or supershifted by Abs to p50, p52, p65, c-Rel, and RelB; p50, p56, and c-Rel; and p50, respectively (Fig. 9B). IL-10 up-regulated complexes a, b, and c less than CpG ODN-2006, whereas CpG ODN-2006 and IL-10 up-regulated complexes a, b, and c more than CpG ODN-2006 or IL-10 alone. Thus, CpG DNA cooperates with IL-10 to increase p50, p52, p65, c-Rel, and RelB binding to I3.

View larger version (49K): [in this window] [in a new window]   FIGURE 9. CpG DNA and IL-10 up-regulate NF-B-Rel, STAT, and IRF binding to ISRE-B3 cis-I3 in human B cells. A, Aligned –232/–176 I3, I1, and I2 DNA sequences. The boxed sequence corresponds to the ISRE-B3 motif. Gels show nuclear protein binding to a radiolabeled oligonucleotide encompassing ISRE-B3 cis-I3 or a consensus Oct1-binding motif (loading control). Nuclear extracts were from naive B cells cultured with or without CpG ODN-2006, IL-10, and/or IL-4 for 6 h. Italic letters indicate shifted protein-DNA complexes. B, ISRE-B3 cis-I3-binding protein-DNA complexes from naive B cells activated with CpG DNA after nuclear protein incubation with a cold probe or Abs to p50, p52, p65, c-Rel, and RelB. Asterisks indicate lanes in which complexes are attenuated or supershifted by the Ab. C, ISRE-B3 cis-I3-binding protein complexes from naive B cells activated with CpG DNA after nuclear protein incubation with a cold probe or Abs to STAT1, STAT2, STAT3, STAT6, IRF1, IRF4, or IRF9. Asterisks indicate lanes in which complexes are attenuated or supershifted by the Ab. Data in A–C correspond to one of three experiments yielding similar results.

CpG DNA and IL-10 activate STAT1, STAT3, IRF1, and IRF4

Considering the key role of ISRE-B3 and GAS sites in I3 transcription, it was hypothesized that CpG DNA and IL-10 synergistically activate the C3 gene by cooperatively activating STATs and IRFs. Reporter vectors carrying multiple ISRE sites (ISRE-Luc), STAT1/3-binding GAS sites (STAT1/3-Luc), and STAT6-specific GAS sites (STAT6-Luc) were used to test the induction of transcriptionally active STATs and IRFs. 2E2 B cells activated ISRE-Luc, STAT1/3-Luc, and, to a lesser extent, STAT6-Luc upon exposure to CpG ODN-2006 or IL-10, but not GpC ODN2006, for 48 h (Fig. 10A). When combined, CpG ODN-2006 and IL-10 activated ISRE-Luc and STAT1/3-Luc more efficiently than CpG ODN-2006 or IL-10 alone. Although unable to significantly activate ISRE-Luc and STAT1/3-Luc, IL-4 activated STAT6-Luc more efficiently than CpG ODN-2006 and IL-10. This activation was attenuated by CpG ODN-2006, but not by GpC ODN-2006. These findings indicate that CpG DNA cooperates with IL-10, but not IL-4, to induce transcriptionally active STATs and IRFs. They also suggest that CpG DNA attenuates IL-4-induced, STAT6-dependent gene transcription.

View larger version (50K): [in this window] [in a new window]   FIGURE 10. CpG DNA and IL-10 induce transcriptionally active STAT1, STAT3, IRF1, and IRF4 in human B cells. A, Activation of ISRE-Luc, GAS-STAT1/3-Luc, or GAS-STAT6-Luc in 2E2 B cells incubated with or without with or without CpG ODN-2006, GpC ODN-2006, IL-10, and/or IL-4 for 2 days. B, Cytoplasmic pSTAT1 and total STAT1 (91- and 84-kDa isoforms), pSTAT2, total STAT2, pSTAT3, and total STAT3 (88- and 83-kDa isoforms), pSTAT6, and total STAT6 in naive B cells cultured with or without CpG ODN-2006, IL-10, and/or IL-4 for 1 h. Nuclear IRF1, IRF4, and IRF9 are also shown. NS, nonspecific band. Data in A and B correspond to one of three experiments yielding similar results.

CpG DNA and IL-10 up-regulate STAT1, STAT3, IRF1, and IRF4 binding to ISRE and GAS cis-I3

Having shown that CpG DNA cooperates with IL-10 to increase the binding of NF-B-Rel-containing complexes a and b to ISRE-B3 cis-I3 (Fig. 9, A and B), it was determined whether these complexes include IRFs and STATs (Fig. 10). Consistent with this, the binding of complexes a and b to ISRE-B3 was attenuated by incubating nuclear proteins from CpG ODN-2006-stimulated B cells with Abs to STAT1, IRF1, and IRF4 (Fig. 9C). Further assays evaluated STAT and IRF binding to the GAS element of I3. Naive B cells up-regulated the binding of complexes a, f, g, h, and i to GAS cis-I3 upon exposure to CpG ODN-2006 for 6 h (Fig. 11A). These complexes were attenuated or supershifted by Abs to STAT1, STAT3, IRF1, and IRF4; STAT1, STAT3, and IRF1; STAT1, IRF1, and IRF4; STAT1 and IRF4; and IRF4, respectively (Fig. 11B). Finally, IL-10 alone or combined with CpG ODN-2006 induced more binding of complexes a, f, g, h, and i to GAS cis-I3 than did CpG ODN-2006 alone. These findings indicate that CpG DNA and IL-10 cooperatively increase STAT and IRF binding to I3.

View larger version (50K): [in this window] [in a new window]   FIGURE 11. CpG DNA and IL-10 up-regulate STAT and IRF binding to GAS cis-I3 in human B cells. A, Aligned –93/–70 I3, I1 and I2 DNA sequences. The boxed sequence corresponds to the GAS motif. Gel shows binding of nuclear proteins to a radiolabeled oligonucleotide encompassing GAS cis-I3 or a consensus Oct1 motif (loading control). Nuclear proteins were from naive B cells incubated with or without CpG ODN-2006, IL-10, and/or IL-4 for 6 h. Italic letters indicate shifted protein-DNA complexes. B, GAS cis-I3-binding protein-DNA complexes from naive B cells activated with CpG DNA after nuclear protein incubation with a cold probe or Abs to STAT1, STAT2, STAT3, STAT6, IRF1, IRF4, or IRF9. Asterisks indicate lanes in which complexes are attenuated or supershifted by the Ab. C, GAS cis-I3-binding protein complexes from naive B cells activated with IL-4 after nuclear protein incubation with a cold probe or Abs to STAT1, STAT2, STAT3, STAT6, IRF1, IRF4, or IRF9. Asterisks indicate lanes in which complexes are attenuated or supershifted by the Ab. Data in A–C correspond to one of three experiments yielding similar results.

CpG DNA and IL-10 require NF-B-Rel, STAT1, STAT3, and IRF4 to activate I3

The involvement of NF-B-Rel, STAT, and IRF proteins in CpG DNA-induced germline C gene transcription was further evaluated in 2E2 B cells transfected with DN plasmids inhibiting the activation and/or DNA-binding activity of endogenous NF-B-Rel, STAT1, STAT3, and IRF4. Enforced expression of DN-IB, DN-IKK, DN-IKK, or, to a lesser extent, DN-NIK attenuated the activation of both I3-Luc and B-Luc in 2E2 B cells exposed to CpG ODN-2006 and IL-10 for 48 h (Fig. 12). In similar B cells, enforced expression of DN-STAT1, DN-IRF4, or, to a lesser extent, DN-STAT3 attenuated activation of I3-Luc but not that of B-Luc. These findings provide additional evidence that CpG DNA and IL-10 initiate IgG CSR by activating NF-B-Rel, STAT, and IRF proteins.

View larger version (26K): [in this window] [in a new window]   FIGURE 12. CpG DNA and IL-10 require NIK, IKK, STAT1, STAT3, and IRF4 to activate the I3 promoter in human B cells. I3-Luc and B-Luc activities in 2E2 B cells transfected with an empty expression vector (control) or with DN-IB, DN-IKK, DN-IKK, DN-NIK, DN-STAT1, DN-STAT3, and DN-IRF4 expression vectors and subsequently incubated with CpG DNA ODN-2006 and IL-10 for 2 days. Data are presented as a percentage of the yield (fold activation of B cells transfected with an expression vector/fold activation of B cells not transfected with an expression vector x 100). Data correspond to one of three independent experiments yielding similar results (bars represent the SD of triplicate determinations).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

This is exemplified by LPS, which stimulates TI IgM production and IgG CSR by engaging TLR4 on mouse B cells (1). Not only does LPS stimulate mouse B cells to produce bacteria-specific IgM and IgG upon dual BCR and TLR4 engagement, but it also elicits polyclonal Ab production through TLR4 only (1). LPS-induced IgG CSR and Ab production would be further enhanced by BAFF, a B cell-stimulating TNF family member produced by LPS-stimulated myeloid DCs (14, 16). Like BCR and TLR4 (10, 60), BAFF receptors on B cells activate NF-B-Rel proteins (61), which are essential for both CSR and Ab production (1). This implies that early TI IgM and IgG production results from the stimulation of both germline and somatically recombined Ag receptors with intersecting NF-B-Rel pathways.

Human B cells lack TLR4 but express TLR9 (22), an intracellular pattern recognition receptor that detects CpG DNA from viruses and bacteria (18). Previous studies show that CpG DNA triggers B cell proliferation and TI IgM production (20, 24, 25), but do not clarify the role of CpG DNA in IgG CSR. By showing that CpG DNA elicits CSR to C1, C2, and C3, our findings extend to human B cells recent data demonstrating that CpG DNA stimulates CSR to C2a, C2b, and C3 in mouse B cells (62). Unlike LPS (1), CpG DNA does not elicit significant IgG production in the absence of additional stimuli. This might stem from the fact that TLR9 signals through a pathway distinct from that emanating from TLR4 (10, 22). In line with studies showing that CpG DNA induces IL-10 (22, 49) and synergizes with both IL-10 and BCR to activate B cells (24, 25), we found that CpG DNA cooperates with IL-10 and BCR cross-linking to up-regulate IgG. The expression of IgG is also enhanced by BAFF, which is released by myeloid DCs upon exposure to CpG DNA-inducible cytokines, such as IFN- and IL-10 (14, 16, 19, 49).

Low concentrations of microbial CpG DNA are thought to favor the activation of DNA-specific B cells through dual TLR9 and BCR engagement (27, 32, 33). This response would involve B-1 and marginal zone B cells expressing poorly diversified BCRs and might be important to facilitate rapid removal of immunogenic CpG DNA released by invading pathogens. Higher concentrations of microbial CpG DNA would trigger polyclonal B cell activation mainly through TLR9 (25). Although less specific, this response might be important to amplify IgG class switching and production in Ag-primed B cells. Not only would CpG DNA favor the initiation of early TI IgG responses, but it would also amplify later-appearing TD IgG responses. Consistent with this, our results extend previous findings indicating that CpG DNA enhances IgG production in B cells exposed to CD40L (49), a TNF family member expressed by CD4⁺ T cells upon activation by APCs (4). CpG DNA would further enhance TD IgG production by favoring the release of IFN- (19, 22), a powerful inducer of key T cell-costimulating molecules on APCs (21).

By showing that enforced B cell expression of TIR-TLR9, DN-MyD88, DN-IRAK1, or DN-TRAF6 attenuates CpG DNA-induced, but not CD40L-induced, I3 transcription and NF-B-Rel activation, our data suggest that CpG DNA requires TLR9, MyD88, IRAK1, and TRAF6 to initiate IgG CSR. These findings are consistent with recently published data showing that TLR9 and MyD88 are required for the induction of IgG CSR in mouse B cells (62). That TLR9 accounts for CpG DNA-induced IgG CSR is also indicated by our observation that enforced B cell expression of TOLLIP, an adaptor protein that negatively regulates TLR signaling (53), or B cell exposure to chloroquine, an endosomal maturation inhibitor that interferes with CpG DNA-TLR9 interaction (39), attenuates the induction of I3 transcription and NF-