Cell-Dependent Immune Responses Memory B Cell Compartment ...

3 downloads 2463 Views 2MB Size Report
Receive free email-alerts when new articles cite this article. Sign up at: ..... VH sequences analyzed using Finch TV (Geospiza) software were aligned and ...
The Journal of Immunology

Exclusion of Natural Autoantibody-Producing B Cells from IgG Memory B Cell Compartment during T Cell-Dependent Immune Responses1 Agata Matejuk,* Michael Beardall,† Yang Xu,† Qi Tian,† Daniel Phillips,† Boris Alabyev,* Kaiissar Mannoor,* and Ching Chen2* In healthy individuals, a substantial proportion of circulating Abs exhibit polyreactivity and self-reactivity. These Abs are referred to as natural autoantibodies (NAAs). As part of the innate immunity, NAAs play an important role in eliminating pathogens. However, inherent to their poly/autoreactivity is the potential for NAAs to differentiate to high-affinity autoantibodies during an immune response. We recently generated site-directed transgenic mice that express a prototypic NAA, ppc1-5, which binds a variety of self- and non-self-Ags including DNA and phosphocholine. We have shown previously that B cells expressing the ppc1-5 NAA are positively selected during their primary development. In this study, we demonstrate that following immunization with the T-dependent Ag, phosphocholine conjugated to keyhole limpet hemocyanin, ppc1-5 NAA B cells mounted a quick IgM Ab response and entered germinal centers, but they failed to differentiate to IgG-producing cells during late primary and memory responses. Hybridomas and cDNA clones derived from the immunized mice included many IgM NAA-producing cells, but IgG NAA clones were extremely rare. Instead, many of the IgG B cells replaced their IgH transgene with an endogenous VH gene and produced non-autoreactive Abs. These results indicate that although NAA B cells are positively selected in the preimmune repertoire and can participate in early IgM Ab response, they are subjected to regulatory mechanisms that prevent them from developing to high-affinity IgG autoantibody production. This would explain, at least in part, why NAAs do not cause autoimmunity in most individuals. The Journal of Immunology, 2009, 182: 7634 –7643.

B

cells that produce disease-associated autoantibodies are subject to stringent negative selection, namely, deletion, functional silencing (anergy), and conversion of self-reactive receptors to non-self-reactive receptors (receptor editing) (1–7). However, it remains a paradox that Abs reactive with selfconstituents are found in sera of healthy individuals. In the 1980s, several groups reported that a substantial proportion of serum Abs in healthy humans and rodents as well as in lower phylogenetic species were autoreactive and polyreactive (8 –15). These Abs are thought to arise naturally without Ag stimulation because they are present in cord blood (16) and in newborn humans (17) and mice (18), as well as in mice housed in germfree conditions and fed an Ag-free diet (19). They are therefore referred to as natural autoantibodies (NAAs).3 NAAs are mainly of the IgM class and are encoded by germline VH and VL genes (13, 20, 21). Studies over

*Department of Pathology, University of Maryland School of Medicine, Baltimore, MD 21201; and †Department of Pathology, Oregon Health & Science University, Portland, OR 97239 Received for publication May 15, 2008. Accepted for publication April 18, 2009. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1

This work was supported by grants to C.C. from the National Institutes of Health (AI061487), The Pew Charitable Trust, and the Cancer Research Institute.

2

Address correspondence and reprint requests to Dr. Ching Chen, Department of Pathology, University of Maryland School of Medicine, MSTF 7-34B, Baltimore, MD 21201. E-mail address: [email protected]

3

Abbreviations used in this paper: NAA, natural autoantibody; sd-tg, site-directed transgene; PC, phosphocholine; KLH, keyhole limpet hemocyanin; SHM, somatic hypermutation; CSR, class switch recombination; EF, extrafollicular; AFC, Ab-forming cell; LM-PCR, ligation-mediated PCR; TD, T dependent; AP, alkaline phosphatase; GC, germinal center; FWR, framework region. Copyright © 2009 by The American Association of Immunologists, Inc. 0022-1767/09/$2.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.0801562

the past 20 –30 years have demonstrated that NAAs provide critical early protection against pathogens (22–25), and they may also play a role in self-tolerance by facilitating clearance of apoptotic cells and other self-Ags (26). However, it is not clear whether B cells producing NAAs can actively participate in foreign Ag-induced immune responses, in particular, a T cell-dependent (TD) memory response, and, if so, whether the inherent polyreactivity and autoreactivity of such Abs constitute a threat to the host. We recently generated site-directed transgenic (sd-tg) mice in which the H chain gene of a prototypic NAA is inserted into the IgH locus (27). This NAA, named ppc1-5, is encoded by the VH7183.14 and V␭1 germline genes and was derived from the liver of a neonatal mouse. It binds DNA, actin, phosphocholine (PC), as well as a variety of self-Ags and foreign Ags (20). Despite their autoreactivity, B cells expressing ppc1-5H/␭1 NAA are not eliminated in the ppc-1-5H sd-tg mice, but instead they are positively selected during their primary development (27). In this regard, they are similar to B-1 B cells (28). However, the ppc1-5 B cells are located predominantly in the follicles of the spleen and lymph nodes, are also circulating through the peripheral blood, and exhibit a phenotype (IgMintIgDlowCD23highCD5⫺) that is different from typical B-1 or follicular B cells (27). It is not clear whether they represent a subtype of CD5⫺ B-1 B cells or Ag-experienced follicular B cells. Regardless of their cellular origin, the ppc1-5 NAA B cells do not appear to be “anergic” in terms of response to BCR cross-linking and LPS stimulation (27). In this study, we investigate the immune response of ppc1-5H sd-tg mice to the TD Ag PC conjugated to keyhole limpet hemocyanin (KLH). The ppc1-5H mice represent a particularly useful model to examine TD-immune responses of NAA B cells because the transgene is inserted in the IgH locus and can undergo somatic hypermutation (SHM) and class switch recombination (CSR), both

The Journal of Immunology of which are defining features of the TD response. We show that the ppc1-5H/␭1 NAA B cells can participate in an early anti-PC IgM Ab response, but they neither sustain this response nor do they mount a significant memory IgG response. Immunohistologic examination of the spleen revealed many IgMa-␭1 Ab-forming cells (AFCs) but few, if any, IgG1a-␭1 AFCs. Moreover, efforts to isolate IgG ppc1-5H/␭1 B cells from memory response by hybridoma generation and cDNA cloning have demonstrated the extreme paucity of such B cells. Instead, most of the recovered IgG-␭1 B cell clones had replaced the ppc1-5 VH with an endogenous VH. Taken together, we conclude that the ppc1-5H/␭1 NAA B cells are excluded from the memory IgG Ab repertoire during a TD-immune response and that a molecular mechanism for eliminating NAA BCR is VH replacement.

7635 Ig H chain RT-PCR Total RNA from sorted ␭1⫹ splenic cells (⬃107) was extracted using the RNeasy Mini Kit (Qiagen) and dissolved in 20 ␮l of water. Reverse transcription was conducted using 2 ␮l of total RNA and a Superscript II cDNA synthesis kit (Invitrogen) according to the manufacturer’s instructions. The following primers were used for cDNA synthesis: IgG1/2B (5⬘GGACAGGGATCCAGATTCC-3⬘) and IgG3B (5⬘-GGACAGGGCTCC ATAGTTCC-3⬘). PCR amplification of cDNA was conducted using VH7183 primer (5⬘-GGAGTCTGG(A⫹G)GGAGGCTT-3⬘) and downstream IgG constant region primers IgG1/2A (5⬘-GGGCCAGTGGATAG AC(A⫹C⫹T)GATGG-3⬘) and IgG3A (5⬘-GGACCAAGGGATAGACAG ATGG-3⬘), respectively. PCR conditions include an initial denaturation at 92oC for 9 min followed by 40 cycles at 94oC for 45 s, 62oC for 40 s, and 72oC for 60 s and finished with a final elongation step at 72oC for 7 min.

Cloning and sequencing

The ppc1-5 H-chain sd-tg mice have been described previously (27). The ppc1-5 VDJH is joined to the constant region of the 129/Sv mouse strain and therefore the transgene-encoded BCR is of the IgHa allotype. The C57BL/6 (referred to as B6) mice were purchased from The Jackson Laboratory. All mice were housed under specific pathogen-free conditions. In this study, heterozygous ppc1-5H sd-tg mice were used for all of the experiments.

PCR products were analyzed on 1% agarose gels and DNA amplicons of interest were excised from the gel and purified using a QIAquick gel extraction kit (Qiagen). PCR fragments were cloned directly into the PCR4TOPO vector (Invitrogen) and then transformed into Escherichia coli DH5␣-T1R. Individual colonies were cultured and plasmid DNA was extracted using the QIAprep kit (Qiagen). Sequencing reactions were conducted on an Applied Biosystems PRISM 3130xl Genetic Analyzer. Each clone was sequenced in both directions using M13 forward and T3 reverse primers. VH sequences analyzed using Finch TV (Geospiza) software were aligned and compared with germline sequences found at the National Center for Biotechnology Information (NCBI) database (http://www.ncbi.nlm. nih.gov/igblast/) using a ClustalW 1.8 DNA algorithm.

Ag and immunization

MACS purification of splenic B cells

Materials and Methods Mice

PC-KLH was prepared as described previously (29) and was a gift from Dr. M. Rittenberg (Oregon Health & Science University, Portland, OR). It had a hapten:protein ratio of 80 haptens per 100,000 Da of KLH. The ppc1-5H sd-tg mice and their tg-negative littermates (8 –11 wk old) were immunized i.p. with 200 ␮g of PC-KLH in CFA, followed by 50 ␮g of PC-KLH in CFA or IFA at day 28, and boosted with 20 ␮g of PC-KLH in saline at day 134 after the first immunization. Serum and tissue samples were obtained at various time points after immunization. Mice receiving PC-KLH in CFA during secondary immunization had slightly better serum anti-PC Ab responses than mice receiving IFA. Therefore, the data presented in this study are from CFA twice-injected mice.

Hybridoma generation Three days after PC-KLH boost, mice were sacrificed and spleen cells were isolated and fused with SP2/0 myeloma cells without in vitro stimulation using the established procedure (7). Monoclonal hybridomas that produced Ab were saved for further analysis.

ELISA Binding to Ags was conducted as described previously (20). Briefly, plates were coated with 2 ␮g/ml PC-histone, PC-KLH, or 5 ␮g/ml ssDNA and incubated with sera or hybridoma supernatants. Binding was detected with alkaline phosphatase (AP)-conjugated anti-␬ or anti-␭ (Southern Biotechnology Associates) or with biotinylated anti-IgMa, anti-IgG1a, anti-IgMb, or anti-IgG1b (BD Biosciences) followed by AP-avidin. For measurement of serum IgMa-␭ and IgG1a-␭ Abs, plates were coated with anti-IgMa or IgG1a (BD Biosciences) and Abs were detected by AP-conjugated anti-␭. Determination of Ig isotype and concentration of hybridoma supernatants was as described previously (20). Plates were coated with 2 ␮g/ml goat anti-mouse ␬ or ␭, incubated with cell culture supernatants, and developed with AP-labeled anti-IgM, anti-IgG1, anti-IgG2a, anti-IgG2b, or anti-IgG3 (Southern Biotechnology Associates).

Immunohistochemistry Mouse spleens were embedded in OCT and flash frozen in liquid nitrogen. Five-micrometer sections of frozen tissue were fixed in 2% paraformaldehyde, quenched in 0.3% H2O2, and blocked with 1% BSA. Double stains were conducted with anti-␭1-biotin along with anti-GL7-FITC, anti-IgG1aFITC, or anti-IgMa-FITC (BD Biosciences) for 60 min followed by incubation with AP-labeled anti-FITC (Sigma-Aldrich) and the avidin-biotin complex (Vectastain Elite Kit; Vector Laboratories) for 30 min. AP and HRT were developed with Fast Blue (Sigma-Aldrich) and diaminobenzidine (DakoCytomation), respectively.

To obtain enriched populations of GL-7⫹ and GL-7⫺ cells, splenic cells from the immunized ppc1-5H sd-tg mice were depleted of CD3⫹ cells by MACS using anti-CD3 microbeads (Miltenyi Biotec) per the manufacturer’s instructions. The negative fraction was incubated with anti-GL7-biotin (BD Biosciences) followed by avidin microbeads and separated by MACS column. Flow cytometric analysis showed that the GL7⫹ cells were enriched by 40 – 60% and GL7⫺ cells were ⬎98% pure.

Ligation-mediated PCR (LM-PCR) High molecular mass genomic DNA was isolated from the bone marrow, total spleen, and splenic GL7⫹ and GL7⫺ cells as described previously (30). LM-PCR was performed according to a previously published method (31). Briefly, the BW linker (20 pM) was ligated to genomic DNA (2 ␮g) in a volume of 20 ␮l with 5 U of T4 DNA Ligase (Invitrogen) at 16oC for 12–14 h. After ligation, the DNA was denatured at 95°C for 5 min and subjected to a seminested PCR. For the first-round PCR, 30 – 40 ng of linker-ligated DNA was used as template and for the second-round PCR, 2 ␮l of the primary PCR product was used. The conditions for primary and secondary PCR were 94oC for 30 s, 63oC for 45 s, 72oC for 45 s, for 25 cycles, followed by an extension at 72oC for 10 min. The 3⬘ primer for both PCR was BW-1H, which recognizes the BW linker and the first three nucleotides of the heptamer (31). The 5⬘ primers were specific for the ppc1-5 VH leader intron sequence (5⬘-GTTGTATGCACATGAGACA GAGA-3⬘) and the framework 1 sequence (5⬘-GGAGTCTGGRGGAG GCTT-3⬘). The PCR for ␣-actin (32) was included as a control for DNA input.

Southern blotting PCR products were run on 1% agarose gel and transferred to nylon membrane (Invitrogen) using an XCell II Blot Module (Invitrogen) according to the manufacturer’s instructions. Membranes were hybridized with a biotinylated oligonucleotide probe (ppc1-5H CDR2 probe: 5⬘-TAGTTACAC CTACTATCCAGACAGTGTGAAGGGCC-3⬘) at 60oC overnight using hybridization solution from DakoCytomation, followed by a stringent wash with 2⫻ SSC. A Phototope chemiluminescent detection kit with SDP-Star substrate (New England Biolabs) was used for signal detection. Finally, membranes were exposed to Thermo Fisher Scientific CL-XPosure Film.

RAG1 and RAG2 RT-PCR Total RNA was extracted from bone marrow and splenic cells as described above. Reverse transcription was conducted using 0.3 ␮g of total RNA and a Superscript II cDNA synthesis kit (Invitrogen) in 20 ␮l of reaction mix. We used 2 ␮l of cDNA for subsequent PCR amplification. The primers and conditions for RAG1 and RAG2 RT-PCR were as previously published (33). ␤-Actin RT-PCR (32) was used as a control for RNA input.

7636

EXCLUSION OF NATURAL Ab FROM IgG RESPONSE

FIGURE 1. Ab response kinetics following PC-KLH immunization. A, PC-specific Abs in ppc1-5H/B6 and non-tg/B6 mice were measured by a solid-phase ELISA. The plates were coated with PC-histone and the Abs were detected with anti-IgMa or antiIgG1a for ppc1-5H sera and with antiIgMb or anti-IgG1b for non-tg B6 sera. Sera were diluted 1/500, and the A405 values were compared with the pooled preimmune sera included in each plate to calculate the titers of antiPC Ab: 1 U was defined as the fold increase in A405. B, IgMa-␭1 and IgG1a-␭1 Abs were determined by ELISA. Plates were coated with antiIgMa or anti-IgG1a and the serum Abs were detected with anti-␭. Sera were diluted 1/1000 and the A405 values were compared with the pooled preimmune sera included in each plate to calculate the titers of IgHa-␭ Ab: 1 U was defined as the fold increase in A405. The data represent the mean ⫾ SEM from three to eight mice per time point. Arrows indicate the time of immunization.

Results Ab responses in ppc1-5H sd-tg mice immunized with PC-KLH We have shown previously (27) that in the heterozygous H chainonly ppc1-5H sd-tg mice, ⬃95% of the mature B cells express the ppc1-5H tg, of which ⬃30% express ␭ and the remaining express ␬ L chains. The ppc1-5H/␭1 B cells represent NAA-producing B cells while the majority of the ␬-expressing B cells and the non-tg B cells are non-NAAs (27). To determine whether NAA B cells can participate in TD-immune responses, we chose to immunize mice with a well-studied TD Ag, PC-KLH, because the ppc1-5 NAA binds p-nitrophenyl PC, the precursor of diazo-linked PCprotein conjugates (20). Use of PC-KLH therefore guarantees engagement of the ppc1-5H/␭1 BCR by the Ag. Groups of heterozygous ppc1-5H sd-tg and non-tg B6 mice were immunized with PC-KLH three times over a period of 4 and half months. The mice were bled at various time points and the serum anti-PC Ab levels were measured using allotype-specific reagents because the tg-encoded Abs are of the IgHa allotype, whereas the non-tg-encoded Abs from B6 mice are IgHb. Increased levels of anti-PC IgM Abs were readily detected after initial immunization in both tg and non-tg mice, although the latter achieved a higher concentration (Fig. 1A). However, different kinetics were observed between tg and non-tg IgM Abs during the course of primary and secondary responses: the tg-encoded IgMa Abs peaked at day 7 and decreased rapidly afterward, whereas the non-tg Abs reached a peak titer at day 14 and remained at high levels until at least day 28, when the mice received the second immunization. The tg IgMa Abs rose again after each subsequent immunization, whereas the non-tg IgM Abs did not further increase following secondary immunization, likely due to their high titers at that point. These results indicate that the ppc1-5 B cells were able to give rise to an early IgM Ab response upon Ag stimulation, but the response disappeared quickly. In contrast, the non-tg B cells had a slower but more sustainable IgM response.

As expected, the IgG response lagged behind the IgM response (Fig. 1A). In non-tg mice, the anti-PC IgG1 Abs did not appear until day 28 and reached their peak shortly after the secondary immunization and remained at high levels until day 134. In tg mice, however, the anti-PC IgG1a Abs only became detectable after secondary immunization and slowly rose to reach high levels at day 134, mainly due to an increase in the IgG ␬ Abs (see below). Since the L chain is not predetermined in the ppc1-5H sd-tg mice, both ␭ and ␬ L chains can contribute to the IgHa Abs. To estimate the relative contribution of ␭ Abs to the anti-PC response, we measured the kinetics of IgHa-␭ Abs. (We could not measure IgHa-␬ because the anti-allotypic reagents are themselves mouse IgG/␬ and would cause high background with a secondary antimouse ␬ Ab.) The IgMa-␭ Abs increased quickly and significantly (⬃6-fold) after each immunization (Fig. 1B), a response similar to that of the anti-PC IgMa Abs. In contrast, the IgG1a-␭ Abs only increased slightly (⬍2-fold) during the entire course of the immune response. These results suggest that the ppc1-5H/␭1 NAA B cells did not contribute significantly to the IgG Ab response and that the majority of the IgG Abs in the late phase of the response were of the non-NAA ␬ Abs. To visualize the formation of AFCs, splenic sections from ppc1-5H sd-tg mice were stained with ␭1 in combination with IgMa or IgG1a. Seven days after the primary immunization, clusters of darkly stained IgMa-␭1 AFCs were seen in the T zone/red pulp bridge channels (Fig. 2A). IgG1a AFCs were rare in the primary response (data not shown), but they were frequent in the memory response. However, few IgG1a cells coexpressed ␭1, consistent with the serum data. Germinal center reaction in immunized ppc1-5H mice To determine whether ppc1-5H/␭1 B cells can participate in germinal center (GC) reactions, consecutive spleen sections from the

The Journal of Immunology

7637

FIGURE 3. Analysis of hybridomas derived from the ppc1-5H sd-tg mice following PC-KLH immunization. Hybridomas were generated from splenic B cells of five ppc1-5H sd-tg mice 3 days after PC-KLH memory challenge (see Materials and Methods for details). Ig isotypes were determined by ELISA. The presence or absence of the ppc1-5H sd-tg was determined by PCR and/or sequencing.

Analysis of hybridomas from the immunized ppc1-5H mice

FIGURE 2. AFC formation and GC reaction in ppc1-5H sd-tg mice. A, EF AFC formation in ppc1-5H mice during primary (7 days after the first immunization) and memory (9 days after the third immunization) responses to PC-KLH. Splenic sections were stained with anti-␭1 (brown) along with anti-IgMa or IgG1a (blue). B, Immunohistochemical analysis of GC reaction in ppc1-5H sd-tg mice during memory response (9 days after the third immunization). Consecutive splenic sections were stained with anti ␭1-GL7 or anti ␭1-IgMa Abs. Many ␭1⫹IgMa⫹ B cells are present in GL7⫹ GCs. The lack of staining in the GC dark zone is likely due to the down-regulation of surface Ig on centrablasts. These are representative sections of three to five mice of each group. Original magnifications, ⫻100 and ⫻400. C, Flow cytometric analysis of ␭1-IgMa B cells isolated from the ppc1-5 sd-tg mice before and after PC-KLH immunization (8 days after secondary immunization). Levels of GL7 expression are shown as a histogram. At least five mice from each group were examined; plots shown are representative.

immunized mice were stained with anti-GL7-␭1, and IgMa-␭1. GCs were seen 9 –12 days after each immunization and were most numerous during memory responses. B cells expressing IgMa and ␭1 were readily identified in the GL7⫹ GCs (Fig. 2B). Consistent with the immunohistochemical results, flow cytometric analysis of the splenic cells from the immunized ppc1-5H mice showed a small but significant population of IgMa-␭1 B cells expressing a high level of GL-7 (Fig. 2C). These results indicate that the ppc15H/␭1 NAA B cells can enter GCs.

To further characterize B cells in the ppc1-5H sd-tg mice during memory anti-PC-KLH response, hybridomas were generated from the spleen of ppc1-5H mice 3 days after the third PC-KLH immunization. Three hundred hybridomas were analyzed: 242 (81%) were IgM and 58 (19%) were IgG (Fig. 3). More than 30% (75 of 242) of the IgM hybridomas expressed ␭1. In contrast, ⬍10% (5 of 58) of the IgG hybridomas were ␭1. To determine whether the ␭1 hybridomas represent ppc1-5H/␭1 NAA B cells, the presence of ppc1-5H sd-tg was analyzed by PCR. Thirty-nine (74%) of 53 IgM-␭1 hybridomas analyzed retained the ppc1-5H tg, but none of the five IgG-␭1 hybridomas did. Sequence analysis of some of the ppc1-5H/␭1 IgM hybridomas revealed few or no mutations in the VH and V␭ genes (data not shown), indicating that they represent newly activated naive B cells rather than post-GC memory B cells. These IgM ppc1-5/␭1 hybridomas showed an Ag-binding pattern similar to that of the original ppc1-5 mAb, including binding to PC-KLH and DNA (data not shown) (27). The lack of ppc1-5H/␭1 IgG B cells in the memory response could be due to competition with ␬ B cells. We therefore crossed the ppc1-5H sd-tg to J␬del mice, which do not express the ␬ L chain. The mature B cell repertoire of the ppc1-5H/J␬del mice comprised ⬎95% of tg⫹/␭ B cells and ⬍5% of non-tg/␭ B cells (data not shown). Hybridomas were generated from two ppc1-5H/J␬del mice after PC-KLH immunization. A total of 209 hybridoma clones was screened by ELISA: 180 (86%) were IgM and 29 (14%) were IgG. Nineteen stable IgG hybridomas were established for further analysis. To our surprise, only two of the 19 hybridomas retained the ppc1-5H sd-tg and both of these were inactivated by nucleotide deletion (see below). All 19 hybridomas expressed a functional endogenous H chain (Table I). Somatic mutations were identified in all 21 VH sequences and 18 of the 19 V␭ sequences, indicating that these were GC-derived B cells. A variety of endogenous VH genes were used by these hybridomas, including members from the Q52, J558, 7183, and VGam families. Interestingly, such diverse VH gene usage has also been observed in the memory anti-PC-KLH response in non-tg mice (34 –36), although the individual VH genes used are different. It appears that diversification of the Ab repertoire during memory anti-PC-KLH response is a feature common to both ppc1-5H and non-tg mice.

7638

EXCLUSION OF NATURAL Ab FROM IgG RESPONSE

Table I. Characterization of hybridomas from ppc1-5H/J␬del mice following PC-KLH immunization Mutationd Hybrida

Isotype

tgb

ppc1-5 F2-6

␥2b/␭1 ␥1/␭1

⫹ ⫹

F2-12 F1-5 F2-3 F2-4 F2-5 F1-6 F2-2 F1-3 F1-8 F2-9 F2-8 F1-7 F1-4 F1-9 F2-1 F2-10 F2-11 F2-7

VHc

7183.14 DQ52.2.4 7183.14 ␥1/␭1 ⫹ DQ52.2.4 7183.14 ␥3/␭1 ⫺ SM7.4.63 ␥1/␭1 ⫺ J558.15 ␥1/␭1 ⫺ J558.67.166 ␥1/␭1 ⫺ J558.83.189 ␥2b/␭1 ⫺ J558.6.96 ␥1/␭1 ⫺ 7183.2.3 ␥3/␭1 ⫺ Vgam3.8 ␥2b/␭1 ⫺ 7183-20.37 ␥1/␭1 ⫺ J558.83.189 ␥1/␭1 ⫺ VH37.1 ␥2a/␭1 ⫺ 7183.20.37 ␥3/␭1 ⫺ J558.78.182 ␥2b/␭1 ⫺ 7183.20.37 ␥1/␭1 ⫺ VH61-1p ␥1/␭1 ⫺ 7183.20.37 ␥2a/␭1 ⫺ J558.83.189 ␥2a/␭1 ⫺ 7183.20.37 Average mutational frequency

Ag-Binding Activitye

JH

VH

V␭

PC-His

PC-KLH

ssDNA

3 3 3 3 3 3 3 4 3 3 3 2 4 3 3 1 1 1 3 1 3 1

0 2 5 ⫹ nt del 2 6 ⫹ nt del 4 5 6 8 3 5 3 5 10 10 1 12 1 3 1 8 11 5.3/VH

0 3

100 33

100 37

100 31

2

40

29

16

6 8 1 2 3 6 0 2 2 3 7 5 3 2 2 3 1 3.3/V␭

194f 14 43 28 29 49 26 75 42 28 28 21 22 20 16 16 11

120 400 374 360 133 170 146 18 17 39 19 84 77 34 22 15 19

⬍5 ⬍5 ⬍5 8 19 50 16 ⬍5 10 18 ⬍5 19 ⬍5 28 ⬍5 ⬍5 ⬍5

Hybridomas were generated from splenic B cells of two ppc1-5H/J␬del sd-tg mice following the third immunization with PC-KLH (see Materials and Methods). Presence or absence of ppc1-5H sd-tg was determined by PCR. c VH and JH gene sequences were determined by RT-PCR sequencing. Individual genes were assigned based on the NCBI BLAST Ig Germline Gene Database. The 7183.20.37 and J558.83.189 VH genes were used by more than one hybridoma, but none of the hybridomas in these two VH gene groups were clonally related because they had different CDR3 junctions (data not shown). d Number of mutations per V gene. e Ag binding was determined by solid-phase ELISA as described in Materials and Methods. Relative binding activity was calculated based on OD at 2 h in ELISA and expressed as percentage of original ppc1-5 Ab OD. All samples were tested at the same time using the same concentration (2 ␮g/ml). f Bold indicates binding avidity higher than ppcl-5 Ab. a b

We compared the relative binding avidity of the IgG-␭1 hybridomas to that of an unmutated ppc1-5H/␭1 IgG Ab (from a transfectoma). Only 1 of the 19 hybdridomas had increased binding for PC-histone but 7 had improved binding for PC-KLH, indicating a selection toward the carrier protein, a phenomenon described previously in the memory anti-PC-KLH response in non-tg mice (37). Notably, all 19 hybridomas had much reduced or undetectable binding for ssDNA.

FIGURE 4. Somatic mutations in VH and V␭ genes of F2-6 and F2-12 hybridomas. The top line shows portions of the germline sequences of ppc1-5, Q52.2.4, or V␭2. Sequence identities to germline are shown as dashes and nucleotide deletions as asterisks. Sequences are numbered according to Kabat et al. (71). The codon numbers may not be in consecutive order (such as that of ppc1-5 FWR2) due to space limits.

Two hybridomas (F2-6 and F2-12) are of particular interest. They kept the ppc1-5H sd-tg yet expressed an endogenous IgH, VHQ52.2.4. The sd-tg in both hybridomas was inactivated by an eight-nucleotide deletion in the CDR2 resulting in stop codons (Fig. 4). These two hybridomas were clonally related because they had the same eight-nucleotide deletion, expressed the same endogenous VDJ, and shared most, but not all, mutations in VH and V␭. Remarkably, the defunct ppc1-5 VH in both hybridomas

The Journal of Immunology

7639

FIGURE 5. VH replacement of ppc1-5H sd-tg. A, VDJ junction sequences of VH genes that have undergone VH replacement. The top line shows the ppc1-5H VDJ sequence. The 3⬘ end heptamer is shown in the box, the junction in bold, the DH underlined, the N additions in italic, and the P nucleotide in capital letters. B, Schematic representation of VH gene replacement at the ppc1-5H sd-tg locus. The initial rearrangement involves D3 VDJH invasion and the subsequent VH3 D-VDJH rearrangement results in a new functional VH gene. The excised ppc1-5 VH along with the intervening D segments are shown. The positions of the BW-1H linker, PCR primers, and hybridization probe are indicated. C, LM-PCR detection of dsDNA breaks at the ppc1-5 VH recombination signal end. Bone marrow and splenic DNA samples from the naive and immunized mice were subjected to LM-PCR followed by Southern blotting for detection of VH signal breaks. The actin gene was amplified as a control for DNA input. D, Detection of RAG1 and RAG2 transcripts by RT-PCR. PCR products were detected by ethidium bromide staining. Lane 1, Negative control (no transcriptase); lane 2, ppc1-5 bone marrow cells; lane 3, ppc1-5 tail RNA; lane 4, spleen cells from nonimmunized ppc1-5H mice; and lane 5, spleen cells from immunized ppc1-5H mice. ␤-Actin was used as a control for mRNA input. C and D, Each sample is representative of two experiments, each with a mixture of cells from three mice of the indicated group.

accumulated three times as many mutations as the VHQ52.2.4 gene. Assuming that the mutational rate is similar in all rearranged VH genes in the same B cell, one can then conclude that SHM must have started earlier in ppc1-5 VH than in VHQ52.2.4, suggesting that the rearrangement of the endogenous VH may have occurred in the GC after initiation of SHM (38). The V␭ gene is also unusual in these two hybridomas: it involves rearrangement of V␭2 to J␭1-C␭1. Such a combination is rare: of 44 V␭2 sequences we have identified in the NCBI Ig Blast Database, only 2 are joined to J␭1 and the remaining 42 are joined to J␭2. This is probably due to the fact that the V␭2 gene is further away from the J␭1-C␭1 cassette on the mouse ␭ locus. This unusual combination could be the result of receptor editing at the ␭ allele (30). The observation

of fewer mutations in V␭2 than in VHppc1--5 in these two hybridomas is consistent with this prediction. Aside from the two hybridomas that inactivated the ppc1-5H tg by nucleotide deletion, the rest of the IgG-␭ hybridomas lost the tg entirely either by deletion of the tg locus (7) or by VH replacement. The latter is a process by which the targeted VH gene is replaced by an upstream VH via secondary recombination mediated by a heptamer embedded at the 3⬘ end of most of the VH genes (39). By examining the VH sequences, we found five clones, three hybridomas (F2-1, F2-2, and F2-9) and two cDNA clones (BG3-3 and CG1-2) (see below) that showed evidence of VH replacement (Fig. 5A). These clones retained the unique ppc1-5 VDJH junction sequence (or footprint) as part of the new VHCDR3 region. Four of

7640 the five clones appeared to have an initial DH to VDJppc1-5 rearrangement followed by a VH to DH-VDJppc1-5 recombination (Fig. 5B) similar to what we described previously (39). One clone (F2-2) did not contain an identifiable DH sequence and therefore could have resulted from direct VH to VDJppc1-5 recombination. Two clones (F2-1 and F2-2) had multiple non-templated (N) nucleotide additions at the new VD and/or DJ junctions, and therefore they must have undergone VH replacement in the bone marrow at the pro-B stage because the enzyme (TdT) responsible for N addition is thought to be active mainly in the pro-B cells (32, 40). Clones F2-9 and CG1-2 had no apparent N additions except for P (palindromic) nucleotide addition, which does not require TdT (41). These clones could have undergone VH replacement at a later stage of B cell development. To formally prove that VH replacement has occurred at the ppc1-5H sd-tg site and to test whether VH replacement can take place in the bone marrow and spleen, we performed a linker-mediated PCR (LM-PCR) assay, which detects dsDNA breaks at the signal ends generated during Ig gene rearrangement (31). Specific primers were designed to identify the signal breaks at the ppc1-5 VH replacement site (Fig. 5B). A 300-bp DNA amplicon, corresponding to the ppc1-5 VH replacement product, was identified in the bone marrow samples from the ppc1-5H sd-tg mice (Fig. 5C). The tail DNA from ppc1-5H mice and the bone marrow and splenic DNA from non-tg mice were used as negative controls and they did not show amplification products, demonstrating that the assay is specific for the ppc1-5 VH signal breaks. The splenic B cells from the immunized ppc1-5H mice (12 days after secondary immunization) also gave rise to a weak but distinct PCR amplicon. We then enriched the splenic B cells from the immunized ppc1-5H mice for GL-7⫹ cells and found abundant signal breaks in the GL-7⫹ B cell population (Fig. 5C). Since VH replacement is thought to be a RAG-dependent process (42), we sought to confirm RAG expression by RT-PCR. As shown in Fig. 5D, RAG1 and RAG2 transcripts were detected in the bone marrow and spleen of the immunized ppc1-5H mice. Together, these results indicate that the ppc1-5 NAA B cells can undergo VH replacement at the sd-tg IgH allele in the bone marrow. The detection of dsDNA breaks and RAG activity in the spleen of the immunized mice suggests that VH replacement could also take place in the peripheral lymphoid tissue during an immune response. However, it is not clear whether the splenic B cells that are undergoing VH replacement represent immature B cells or mature GC B cells (see Discussion). Analysis of IgH cDNA clones from the immunized ppc1-5H mice To characterize the rare IgG ppc1-5H/␭1 B cells, we used another strategy to retrieve the VH genes directly from IgG ␭1 B cells. CD19⫹␭1⫹ splenic B cells were sorted from the ppc1-5H mice 3 days after the memory PC-KLH boost and RNA was extracted from the sorted cells, from which IgG cDNA clones were established using primers specific for C␥ and the 7183 VH family to which the ppc1-5 VH gene belongs. Of 110 nonidentical IgG VH cDNA clones from ␭1⫹ B cells, 21 (19%) were ppc1-5H. Given that the 7183 primer can amplify VH genes from the 7183, S107, 3609N, and J606 families and that these families represent ⬃17% of the total VH repertoire (43), the ppc1-5 VH is therefore expressed by ⬃3% of the ␭1 IgG B cells and ⬍1% of the total IgG B cells. This explains why such B cells were not recovered as hybridomas. Analysis of somatic mutation patterns in ppc1-5H⫹ and ppc1-5H⫺ (endogenous) VH clones revealed that ⬃50% of the ppc1-5 VH sequences were germline, suggesting that they may represent pre-GC B cells. The remaining ppc1-5H sequences contained 1– 6 mutations with an average mutational frequency of 2.9 per VH. The ratio of replacement:silent mutations for ppc1-5 CDRs

EXCLUSION OF NATURAL Ab FROM IgG RESPONSE was 3.5:1 and was 4.3:1 for framework regions (FWRs), both of which were similar to the ratios expected for random mutation (4.2:1 for ppc1-5 CDRs and 3.2:1 for FWRs). On the other hand, 83% of the endogenous VH had mutations and the mutational frequencies were relatively high (ranging from 1 to 24 and averaging 5.7 per VH). The endogenous VH had a higher replacement:silent ratio in CDRs (4.8:1) than in FWRs (1.6:1), suggesting Ag selection. Collectively, our data from hybridoma and VH cDNA clone analyses demonstrate that B cells expressing IgG ppc1-5H/␭1 BCR were extremely rare in the memory anti-PC-KLH response and that the few surviving ones had relatively few mutations and did not show evidence of Ag selection.

Discussion Upon TD Ag immunization, Ag-specific B cells follow one of two pathways: they either migrate to the extrafollicular (EF) region and become short-lived plasma cells or enter GC and develop into memory B cells or long-lived plasma cells (44). During GC development, the Ab genes characteristically undergo switch recombination and SHM, leading to affinity maturation (45– 48). The EF plasma cells give rise to the early Ab response, whereas the GCderived plasma cells sustain the Ab levels in the later phase of the immune response. Our findings of the rapid increase in serum antiPC IgMa-␭1 Ab levels and the early appearance of IgMa-␭1 AFCs in the EF region of the spleen after PC-KLH immunization indicate that the ppc1-5H/␭1 NAA B cells can enter the EF pathway and differentiate to short-lived plasma cells upon Ag encounter. This demonstrates for the first time that a typical polyreactive and autoreactive NAA B cell can participate in the early phase of a TD-immune response and reinforces the importance of NAAs not only as a first line of natural defense but also as a bridge between innate and adaptive immunity. The ppc1-5 NAA B cells can also enter GCs, but they rarely differentiate to IgG-producing cells in the late primary or memory responses as shown by serum Ab studies, immunohistochemical staining, and extensive hybridoma and VH clone analyses. This indicates a blockade or a counterselection of the NAA B cells in GC development. In contrast, the non-NAA B cells (␬⫹ or tg⫺ B cells) in ppc1-5H sd-tg mice are able to differentiate to IgG-producing cells and dominate the memory response. Fig. 6 summarizes the proposed developmental pathways of NAA and non-NAA B cells in ppc1-5H sd-tg mice following PC-KLH immunization. The NAA B cells, although they do not directly participate in the memory response, can serve as precursors of the memory IgG non-NAA B cells via receptor editing. This is particularly evident in the ppc1-5H/J␬del mice, where almost all of the B cells start as ppc1-5H/␭ NAA B cells and the non-NAA B cells would have to be created by H chain editing. Indeed, all of the IgG hybridomas isolated from the immunized ppc1-5H/J␬del mice have undergone H editing, either by VH replacement at the targeted allele or by inactivation of the tg followed by rearrangement of the nontargeted allele. Three of the 19 IgG hybridomas showed evidence of VH replacement as assessed by the residual ppc1-5 VDJ junction sequence. Such a “footprint” however can be lost by nucleotide deletion and/or N addition associated with VDJ recombination. It is particularly true if the recipient VDJ junction is very short, such as that of the ppc1-5 VH. It has been reported that only ⬃20% of the VH replacement events would leave identifiable footprints (49). Therefore, the frequency of VH replacement in ppc1-5 B cells is likely to be much higher. In support of this prediction is the increased usage of JH3 among IgG hybridomas (12 of 19; Table I) because JH3 is part of the ppc1-5 VDJH gene and would be left behind after the replacement of the ppc1-5 VH (Fig. 5B).

The Journal of Immunology

7641

FIGURE 6. Proposed developmental pathways of NAA and non-NAA B cells in ppc1-5H sd-tg mice during anti-PCKLH-immune response. In the preimmune mature B cell repertoire, the NAA B cells (ppc1-5H/␭) constitute ⬃30% of the B cells and the remaining 70% are non-NAA B cells (ppc15H/␬ and non-tg). After the first immunization with PC-KLH, the NAA B cells quickly differentiate to EF AFCs and mount an early IgM response. Some of the non-NAA B cells that recognize PC-KLH also participate in the early Ab response. Both types of B cells can enter GCs, but they have different fates: the non-NAA B cells are able to differentiate to IgG memory B cells and post-GC AFCs, whereas the NAA B cells are arrested during their GC development, leading to the short-lived NAA Ab response. After the secondary immunization, the non-NAA memory B cells undergo extensive proliferation and produce large amounts of IgG Abs. In contrast, the NAA B cells do not proceed to memory response.

Indeed, ongoing VH replacement at the ppc1-5H locus is readily detected in the bone marrow of ppc1-5 sd-tg mice. To our surprise, VH replacement signal ends and RAG gene expression were also identified in the spleen of the immunized ppc1-5 mice, especially in GL7⫹ B cells. These findings, at face value, appear to suggest ongoing VH replacement in the GC B cells. However, such data have to be interpreted with caution since the identity of peripheral RAG-expressing B cells is controversial. In 1996 and 1997, several groups reported reexpression of RAGs and the presence of dsDNA breaks at the ␬ gene recombination signal ends in B220lowGL7⫹ GC B cells (33, 50 –52). Peripheral VH replacement has also been observed in mutating human tonsillar B cells (53) and synovial B cells of rheumatoid arthritis patients (54). These have been taken to mean that secondary Ig gene rearrangement or receptor editing can be reinduced in mature GC or GC-like B cells upon Ag stimulation, a process termed “receptor revision” (55). However, subsequent studies with RAG-GFP reporter mice have found that peripheral B cells expressing RAG had a phenotype of immature B cells (56 –58). In addition, it has been shown that most of the RAG-expressing cells that accumulate in the spleen after immunization do not proliferate and that the accumulation of such cells is not Ag dependent since adjuvant alone can induce similar changes (59). It is therefore concluded that the RAG-expressing B cells found in the spleen are newly emigrated immature bone marrow B cells rather than mature GC B cells. Another source of immature B cells in the spleen could be extramedullary hematopoiesis. Furthermore, SHM can result in dsDNA breaks in VHCDR3, which may be detected by LM-PCR. Taken together, our data indicate that the ppc1-5 B cells can undergo VH replacement and use it as a mechanism to alter the specificity and/or affinity of their BCR. However, it remains unclear at what developmental stage(s) VH replacement takes place in these B cells and whether such a process can be induced in the GC B cells upon Ag immunization. Additional experiments are required to clarify these issues. Given that the ppc1-5H/␭1 NAA B cells are positively selected in primary development (27) and that they can participate in early anti-PC-KLH response and enter GCs, it is surprising that they are absent in the memory IgG response. One explanation could be that the ppc1-5H/␭1 NAAs are unable to improve their Ag binding sites via somatic mutation; alternatively somatic mutation of the ppc1-5 VH gene would result in reduced Ag binding or defective secretion (60, 61). Either scenario would lead to the failure of ppc1-5/␭1 NAAs to compete with non-NAA B cells that do obtain better binding. The finding that some of the mutated non-ppc1-5 Abs had higher avidity for PC-KLH than the germline ppc1-5 Ab

suggests that this may in part account for the disappearance of ppc1-5 NAA in memory response. However, we do not have direct evidence indicating that the ppc1-5 VH is particularly susceptible to deleterious mutations because we were not able to obtain hybridomas that express a mutated yet functional ppc1-5 VH. We are currently investigating this possibility by generating ppc1-5 VH mutants in vitro. Several lines of evidence however suggest that affinity-based competition may not be the only reason for the absence of IgG ppc1-5/␭1 B cells in the memory response. First, most of the non-tg Abs derived from the memory response had lower affinity for PC-KLH than the ppc1-5 NAA. Second, as mentioned above, even in the absence of ␬ B cell competition, we were not able to retrieve IgG clones bearing ppc1-5H/␭1 BCR. The complete absence of ppc1-5/␭1 IgG clones of 509 hybridomas analyzed (300 from ppc1-5H mice and 209 from ppc1-5/Jkdel mice) suggests powerful pressure to purge ppc1-5/␭1 NAA from the IgG Ab repertoire. Third, all memory IgG-␭ Abs had lost anti-DNA activity, suggesting a selection against DNA binding. Comparison of the VDJ sequences of the VHppc1-5 and VH7183.20.37, the most frequently used non-tg VH by the IgG hybridomas reveals that although the two sequences share a high degree of homology, the ppc1-5 VDJ has three more arginine residues, one of which is in CDR3. It is known that arginines, particularly in VH CDR3, contribute significantly to DNA binding (62). Indeed, none of the five VH7183.20.37 Abs bound DNA and they only had modest avidity for PC-KLH (Table I). Therefore, the repeated use of VH7183.20.37 by the memory B cells may not be entirely due to selection for its high Ag-binding affinity; but rather selection against autoreactivity may play a role. These findings are consistent with a previous study in which dual specific B cells that bind the hapten p-azophenylarsonate and nuclear autoantigens were able to enter GCs, but they were absent in the memory compartment following p-azophenylarsonate immunization (63). Forced expression of Bcl-2 could rescue these B cells, but the rescued B cells all had reduced autoreactivity as a result of somatic mutation. It was therefore postulated that negative selection against autoreactivity took precedence over positive selection for foreign Ag binding in multireactive B cells. Collectively, our results indicate that both the need for better Ag binding sites and the need for eliminating autoreactivity may be driving the counterselection of IgG ppc1-5 BCR. In addition to GC regulation, other mechanisms may also control the differentiation of NAA B cells during an immune response. The finding that very few IgG ppc1-5H/␭1 B cells are present in both primary and memory responses points to a checkpoint linked

7642 to CSR. This regulation could be independent of GC reaction because CSR can occur in the EF without somatic mutation (64). Comparison of signal transduction between IgM and IgG BCRs has revealed that the latter has a distinct signal transduction profile that promotes plasma cells differentiation (65– 67). It would therefore not be surprising for the immune system to include a regulatory process at this “switch” point to prevent indiscriminate plasma cell differentiation and Ab production from self-reactive IgG B cells. In summary, our data demonstrate that the ppc1-5 NAA-producing B cells are excluded from the IgG memory B cell compartment. The apparent “innocuous” natural autoimmunity is therefore not “ignored” by the tolerance mechanisms in nonautoimmune mice, but instead it is kept in check during an immune response to prevent the development of high-affinity anti-self-IgG Abs. Similar regulation of NAA has also been observed in humans. For example, natural autoreactive B cells expressing the VH4-34 cold-agglutinin Ab are positively selected in the naive repertoire, but are excluded from the memory B cell compartment and from differentiation to plasma cells in healthy individuals (68). However, in systemic lupus erythematosus patients, these B cells can develop to IgG plasma cells, indicating a defect in the control of NAA B cells as a cause of autoimmunity. Similarly, it has been reported that patients with systemic lupus erythematosus and rheumatoid arthritis have increased frequency of autoreactive/polyreactive B cells (69, 70). It will be of great interest to determine whether the tolerance mechanism for ppc1-5 NAA is defective in autoimmune-prone animals and whether, in other circumstances, NAA can ameliorate autoimmunity.

Acknowledgments We thank Drs. Jan Cerny and Martin F. Flajnik for insightful discussion and critical reading of this manuscript and Dr. Marvin B. Rittenberg for proving PC-KLH.

Disclosures The authors have no financial conflict of interest.

References 1. Nemazee, D., D. Russell, B. Arnold, G. Haemmerling, J. Allison, J. F. A. P. Miller, G. Morahan, and K. Beurki. 1991. Clonal deletion of autospecific B lymphocytes. Immunol. Rev. 122: 117–132. 2. Goodnow, C. C., J. Crosbie, H. Jorgensen, R. A. Brink, and A. Basten. 1989. Induction of self-tolerance in mature peripheral B lymphocytes. Nature 342: 385–391. 3. Erikson, J., M. Z. Radic, S. A. Camper, R. R. Hardy, and M. Weigert. 1991. Expression of anti-DNA immunoglobulin transgenes in non-autoimmune mice. Nature 349: 331–334. 4. Tiegs, S. L., D. M. Russell, and D. Nemazee. 1993. Receptor editing in selfreactive bone marrow B cells. J. Exp. Med. 177: 1009 –1020. 5. Radic, M. Z., J. Erikson, S. Litwin, and M. Weigert. 1993. B lymphocytes may escape tolerance by revising their antigen receptors. J. Exp. Med. 177: 1165–1173. 6. Chen, C., Z. Nagy, M. Z. Radic, R. R. Hardy, D. Huszar, S. A. Camper, and M. Weigert. 1995. The site and stage of anti-DNA B-cell deletion. Nature 373: 252–255. 7. Chen, C., M. Z. Radic, J. Erikson, S. A. Camper, S. Litwin, R. R. Hardy, and M. Weigert. 1994. Deletion and editing of B cells that express antibodies to DNA. J. Immunol. 152: 1970 –1982. 8. Guilbert, B., G. Dighiero, and S. Avrameas. 1982. Naturally occurring antibodies against nine common antigens in human sera: I. Detection, isolation and characterization. J. Immunol. 128: 2779 –2787. 9. Dighiero, G., P. Lymberi, B. Guilbert, T. Ternynck, and S. Avrameas. 1986. Natural autoantibodies constitute a substantial part of normal circulating immunoglobulin. Ann. NY Acad. Sci. 475: 135–145. 10. Hartman, A. B., C. P. Mallett, J. Srinivasappa, B. S. Prabhakar, L. Notkins, and S. J. Smith-Gill. 1989. Organ reactive autoantibodies from non-immunized adult BALB/c mice are polyreactive and express non-biased VH gene usage. Mol. Immunol. 26: 359 –370. 11. Gonzalez, R., J. Charlemagne, W. Mahana, and S. Avrameas. 1988. Specificity of natural serum antibodies present in phylogenetically distinct fish species. Immunology 63: 31–36. 12. Prabhakar, B., J. Saegusa, T. Onodera, and A. Notkins. 1984. Lymphocytes capable of making monoclonal autoantibodies that react with multiple organs are a common feature of the normal B cell repertoire. J. Immunol. 133: 2815–2817.

EXCLUSION OF NATURAL Ab FROM IgG RESPONSE 13. Coutinho, A., M. D. Kazatchkine, and S. Avrameas. 1995. Natural autoantibodies. Curr. Opin. Immunol. 7: 812– 818. 14. Flajnik, M., and L. Rumfelt. 2000. Early and natural antibodies in non-mammalian vertebrates. Curr. Top. Microbiol. Immunol. 252: 233–2340. 15. Casali, P., and A. Notkins. 1989. CD5⫹ B lymphocytes, polyreactive antibodies and the human B-cell repertoire. Immunol. Today 10: 364 –368. 16. Chen, Z. J., C. J. Wheeler, W. Shi, A. J. Wu, C. H. Yarboro, M. Gallagher, and A. L. Notkins. 1998. Polyreactive antigen-binding B cells are the predominant cell type in the newborn B cell repertoire. Eur. J. Immunol. 28: 989 –994. 17. Merbl, Y., M. Zucker-Toledano, F. Quintana, and I. Cohen. 2007. Newborn humans manifest autoantibodies to defined self molecules detected by antigen microarray informatics. J. Clin. Invest. 117: 712–718. 18. Dighiero, G., P. Lymberi, D. Holmberg, I. Lundquist, A. Coutinho, and S. Avrameas. 1985. High frequency of natural autoantibodies in normal newborn mice. J. Immunol. 134: 765–771. 19. Hooijkaas, H., R. Benner, J. R. Pleasants, and B. S. Wostmann. 1984. Isotypes and specificities of immunoglobulins produced by germ-free mice fed chemically defined ultrafiltered “antigen-free” diet. Eur. J. Immunol. 14: 1127–1130. 20. Chen, C., M. P. Stenzel-Poore, and M. B. Rittenberg. 1991. Natural auto- and polyreactive antibodies differing from antigen-induced antibodies in the H chain CDR3. J. Immunol. 147: 2359 –2367. 21. Casali, P., and E. W. Schettino. 1996. Structure and function of natural antibodies. Curr. Top. Microbiol. Immunol. 210: 167–179. 22. Ochsenbein, A. F., T. Fehr, C. Lutz, M. Suter, F. Brombacher, H. Hengartner, and R. M. Zinkenagel. 1999. Control of early viral and bacterial distribution and disease by natural antibodies. Science 286: 2156 –2159. 23. Baumgarth, N., J. W. Tung, and L. A. Herzenberg. 2005. Inherent specificities in natural antibodies: a key to immune defense against pathogen invasion. Springer Semin. Immunopathol. 26: 347–362. 24. Boes, M., A. P. Prodeus, T. Schmidt, M. C. Carroll, and J. Chen. 1998. A critical role of natural immunoglobulin M in immediate defense against systemic bacterial infection. J. Exp. Med. 188: 2380 –2386. 25. Zhou, Z., A. Tzioufas, and A. Notkins. 2007. Properties and function of polyreactive antibodies and polyreactive antigen-binding B cells. J. Autoimmun. 29: 219 –228. 26. Peng, Y., R. Kowalewski, S. Kim, and K. Elkon. 2005. The role of IgM antibodies in the recognition and clearance of apoptotic cells. Mol. Immunol. 42: 781–787. 27. Tian, Q., M. Beardall, Y. Xu, J. Li, D. Parker, N. Casanova, A. Bakke, and C. Chen. 2006. B cells expressing a natural polyreactive autoantibody have a distinct phenotype and are overrepresented in immunoglobulin heavy chain transgenic mice. J. Immunol. 177: 2412–2422. 28. Hayakawa, K., and R. Hardy. 2000. Development and function of B-1 cells. Curr. Opin. Immunol. 12: 346 –353. 29. Chesebro, B., and H. Metzger. 1972. Affinity labeling of a phosphorylcholine binding mouse myeloma protein. Biochemistry 11: 766 –771. 30. Luning-Prak, E., M. Trounstine, D. Huszar, and M. Weigert. 1994. Light chain editing in ␬ deficient animals: a potential mechanism of B cell tolerance. J. Exp. Med. 180: 1805–1815. 31. Schlissel, M., A. Constantinescu, T. Morrow, M. Baxter, and A. Peng. 1993. Double-strand signal sequence breaks in V(D)J recombination are blunt, 5⬘-phosphorylated, RAG-dependent, and cell cycle regulated. Genes Dev. 7: 2520 –7532. 32. Li, Y.-S., K. Hayakawa, and R. R. Hardy. 1993. The regulated expression of B lineage associated genes during B cell differentiation in bone marrow and fetal liver. J. Exp. Med. 178: 951–960. 33. Han, S., B. Zheng, D. G. Schatz, E. Spanopoulou, and G. Kelsoe. 1996. Neoteny in lymphocytes: Rag1 and Rag2 expression in germinal center B cells. Science 274: 2094 –2097. 34. Lotscher, M., H. Amstutz, C. H. Heusser, and K. Blaser. 1990. Fine specificity and heavy chain V gene usage of antibodies to the phosphorylcholine hapten. Mol. Immunol. 27: 369 –378. 35. Stenzel-Poore, M. P., U. Bruderer, and M. B. Rittenberg. 1988. The adaptive potential of the memory response: clonal recruitment and epitope recognition. Immunol. Rev. 105: 113–136. 36. Stenzel-Poore, M. P., and M. B. Rittenberg. 1989. Clonal diversity, somatic mutation, and immune memory to phosphocholine-keyhole limpet hemocyanin. J. Immunol. 143: 4123– 4133. 37. Brown, M., M. Schumacher, G. Wiens, R. Brennan, and M. Rittenberg. 2000. The structural basis of repertoire shift in an immune response to phosphocholine. J. Exp. Med. 191: 2101–2112. 38. Brard, F., M. Shannon, E. Luning Prak, S. Litwin, and M. Weigert. 1999. Somatic mutation and light chain rearrangement generate autoimmunity in anti-singlestranded DNA transgenic MRL/lpr mice. J. Exp. Med. 190: 691–704. 39. Chen, C., Z. Nagy, E. Luning Prak, and M. Weigert. 1995. Immunoglobulin heavy chain gene replacement: a mechanism of receptor editing. Immunity 3: 747–755. 40. Desiderio, S. V., G. D. Yancopoulos, M. Paskind, E. Thomas, M. A. Boss, N. Landau, F. W. Alt, and D. Baltimore. 1984. Insertion of N regions into heavychain genes is correlated with expression of terminal deoxytransferase in B cells. Nature 311: 752–755. 41. Komori, T., A. Okada, V. Stewart, and F. W. Alt. 1993. Lack of N regions in antigen receptor variable region genes of TdT-deficient lymphocytes. Science 261: 1171–1175. 42. Lutz, J., W. Muller, and H. M. Jack. 2006. VH replacement rescues progenitor B cells with two nonproductive VDJ alleles. J. Immunol. 177: 7007–7014.

The Journal of Immunology 43. Johnston, C., A. Wood, D. Bolland, and A. Corcoran. 2006. Complete sequence assembly and characterization of the C57BL/6 mouse Ig heavy chain V region. J. Immunol. 176: 4221– 4234. 44. Brink, R., T. Phan, D. Paus, and T. Chan. 2008. Visualizing the effects of antigen affinity on T-dependent B-cell differentiation. Immunol. Cell Biol. 86: 31–39. 45. Jacob, J., G. Kelsoe, K. Rajewsky, and U. Weiss. 1991. Intraclonal generation of antibody mutants in germinal centres. Nature 354: 389 –392. 46. Berek, C., A. Berger, and M. Apel. 1991. Maturation of the immune response in germinal centers. Cell 67: 1121–1129. 47. Schwickert, T., R. Lindquist, G. Shakhar, G. Livshits, D. Skokos, M. Kosco-Vilbois, M. Dustin, and M. Nussenzweig. 2007. In vivo imaging of germinal centres reveals a dynamic open structure. Nature 446: 83– 87. 48. Tarlinton, D. 2008. Evolution in miniature: selection, survival and distribution of antigen reactive cells in the germinal centre. Immunol. Cell Biol. 86: 133–138. 49. Koralov, S. B., T. I. Novobrantseva, J. Konigsmann, A. Ehlich, and K. Rajewsky. 2006. Antibody repertoires generated by VH replacement and direct VH to JH joining. Immunity 25: 43–53. 50. Hikida, M., M. Mori, T. Takai, K. Tomochika, K. Hamatani, and H. Ohmori. 1996. Reexpression of RAG-1 and RAG-2 genes in activated mature mouse B cells. Science 274: 2092–2094. 51. Han, S., S. R. Dillon, B. Zheng, M. Shimoda, M. S. Schlissel, and G. Kelsoe. 1997. V(D)J recombinase activity in a subset of germinal center B lymphocytes. Science 278: 301–305. 52. Papavasiliou, F., R. Casellas, H. Suh, X. F. Qin, E. Besmer, R. Pelanda, D. Nemazee, K. Rajewsky, and M. C. Nussenzweig. 1997. V(D)J recombination in mature B cells: a mechanism for altering antibody responses. Science 278: 298 –301. 53. Wilson, P. C., K. Wilson, Y.-J. Liu, J. Banchereau, V. Pascual, and J. D. Capra. 2000. Receptor revision of immunoglobulin heavy chain variable region genes in normal human B lymphocytes. J. Exp. Med. 191: 1881–1894. 54. Itoh, K., E. Meffre, E. Albesiano, A. Farber, D. Dines, P. Stain, S. E. Asnis, R. A. Furie, R. I. Jain, and N. Chiorazzi. 2000. Immunoglobulin heavy chain variable region gene replacement as a mechanism for receptor revision in rheumatoid arthritis synovial tissue B lymphocytes. J. Exp. Med. 192: 1151–1164. 55. Nemazee, D., and M. Weigert. 2000. Revising B cell receptors. J. Exp. Med. 191: 1813–1817. 56. Yu, W., H. Nagaoka, M. Jankovic, Z. Misulovin, H. Suh, A. Rolink, F. Melchers, E. Meffre, and M. C. Nussenzweig. 1999. Continued RAG expression in late stages of B cell development and no apparent re-induction after immunization. Nature 400: 682– 687. 57. Monroe, R. J., K. J. Seidl, F. Gaertner, S. Han, F. Chen, J. Sekiguchi, J. Wang, R. Ferrini, L. Davidson, G. Kelsoe, and F. W. Alt. 1999. RAG2:GFP knockin mice reveal novel aspects of RAG2 expression in primary and peripheral lymphoid tissues. Immunity 11: 201–212.

7643 58. Kuwata, N., H. Igarashi, T. Ohmura, S. Aizawa, and N. Sakaguchi. 1999. Cutting edge: absence of expression of RAG1 in peritoneal B-1 cells detected by knocking into RAG1 locus with green fluorescent protein gene. J. Immunol. 163: 6355– 6359. 59. Nagaoka, H., G. Gonzalez-Aseguinolaza, M. Tsuji, and M. C. Nussenzweig. 2000. Immunization and infection change the number of recombination activating gene (RAG)-expressing B cells in the periphery by altering immature lymphocyte production. J. Exp. Med. 191: 2113–2120. 60. Chen, C., V. Roberts, and M. Rittenberg. 1992. Generation and analysis of random point mutations in an antibody CDR2 sequence: many mutated antibodies lose their ability to bind antigen. J. Exp. Med. 176: 855– 866. 61. Chen, C., T. Martin, S. Stevens, and M. Rittenberg. 1994. Defective secretion of an immunoglobulin caused by mutations in the heavy chain complementarity determining region 2. J. Exp. Med. 180: 577–586. 62. Radic, M. Z., and M. Weigert. 1994. Genetic and structural evidence for antigen selection of anti-DNA antibodies. Annu. Rev. Immunol. 12: 487–520. 63. Notidis, E., L. Heltemes, and T. Manser. 2002. Dominant, hierarchical induction of peripheral tolerance during foreign antigen-driven B cell development. Immunity 17: 317–327. 64. Jacob, J., R. Kassir, and G. Kelsoe. 1991. In situ studies of the primary immune response to (4-hydroxy-3-nitrophenyl)acetyl: I. The architecture and dynamics of responding cell populations. J. Exp. Med. 173: 1165–1175. 65. Wakabayashi, C., T. Adachi, J. Wienands, and T. Tsubata. 2002. A distinct signaling pathway used by the IgG-containing B cell antigen receptor. Science 298: 2392–2395. 66. Pogue, S., and C. Goodnow. 2000. Gene dose-dependent maturation and receptor editing of B cells expressing immunoglobulin (Ig)G1 or IgM/IgG1 tail antigen receptors. J. Exp. Med. 191: 1031–1044. 67. Horikawa, K., S. Martin, S. Pogue, K. Silver, K. Peng, K. Takatsu, and C. Goodnow. 2007. Enhancement and suppression of signaling by the conserved tail of IgG memory-type B cell antigen receptors. J. Exp. Med. 204: 759 –769. 68. Pugh-Bernard, A., G. Silverman, A. Cappione, M. Villano, D. Ryan, R. Insel, and I. Sanz. 2001. Regulation of inherently autoreactive VH4-34 B cells in the maintenance of human B cell tolerance. J. Clin. Invest. 108: 1061–1070. 69. Yurasov, S., H. Wardemann, J. Hammersen, M. Tsuiji, E. Meffre, V. Pascual, and M. Nussenzweig. 2005. Defective B cell tolerance checkpoints in systemic lupus erythematosus. J. Exp. Med. 201: 703–711. 70. Samuels, J., Y. Ng, C. Coupillaud, D. Paget, and E. Meffre. 2005. Impaired early B cell tolerance in patients with rheumatoid arthritis. J. Exp. Med. 201: 1659 –1667. 71. Kabat, E. A., T. T. Wu, H. M. Perry, K. S. Gottesman, and C. Foeller. 1991. Sequences of Proteins of Immunological Interest. National Institutes of Health, Bethesda, MD.