Antibody-Secreting Cells Formation of High-Affinity Intraclonal Competition

Intraclonal Competition Inhibits the Formation of High-Affinity Antibody-Secreting Cells

This information is current as Thuc-vy L. Le, Tea Hyun Kim and David D. Chaplin of October 2, 2021. J Immunol 2008; 181:6027-6037; ; doi: 10.4049/jimmunol.181.9.6027 http://www.jimmunol.org/content/181/9/6027 Downloaded from References This article cites 37 articles, 12 of which you can access for free at: http://www.jimmunol.org/content/181/9/6027.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on October 2, 2021 Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2008 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Intraclonal Competition Inhibits the Formation of High-Afﬁnity Antibody-Secreting Cells1

Thuc-vy L. Le,2 Tea Hyun Kim, and David D. Chaplin3

Protective immunity requires a diverse, polyclonal B cell repertoire. We demonstrate that affinity maturation of the humoral response to a hapten is impaired when preexisting clonally restricted cells recognizing the hapten are dominant in the B cell -repertoire. B1-8i؉/؊ mice, which feature a high frequency of B cells with nitrophenyl (NP)-binding specificity, respond to NP haptenated proteins with the production of NP-specific Abs, but affinity maturation is impaired due to insufficient generation of high-affinity Ab-producing cells. We manipulated the frequency of NP-specific B cells by adoptive transfer of B1-8 B cells into naive, wild-type recipients. Remarkably, when 104 B1-8 B cells were transferred, these cells supported efficient affinity maturation and plasma cell differentiation. In contrast, when 106 B1-8 cells were transferred, affinity maturation did not occur. These data indicate that restricting the frequency of clonally related B cells is required to support affinity maturation. The Journal of Immunology, 2008, 181: 6027–6037. Downloaded from

ffinity maturation of the Ab response is a hallmark of curs at the B cell/T cell interface (12). Activated Ag-specific B and adaptive immunity. Affinity maturation is a process by T cells subsequently migrate to the follicles and enter into germi- A which Ab produced in response to immunization show nal center (GC) reactions (8, 13). Within the GC structure, GC B increased binding strength to their target Ag over time. This pro- cells interact intimately with the local cellular microenvironment

cess involves both clonal selection (1) and somatic hypermutation and here it is thought that the GC B cells undergo SHM, afﬁnity- http://www.jimmunol.org/ (SHM)4 (2). A broad range of studies using cellular and molecular based selection, and differentiation into memory B cells and long- approaches are providing an increasingly complete appreciation of lived ASC (8). the molecular mechanisms underlying these processes. When C57BL/6 mice are immunized with (4-hydroxy-3-nitro- CD4ϩ T cells are activated by encounter with proteolytically phenyl)acetyl (NP)-haptenated proteins, B cells expressing the ␭ processed Ag presented on APC (3–5). In contrast, naive B cells VH186.2 gene paired with a 1-L chain (LC) predominate in the can be primed by either soluble (6) or membrane-bound unproc- repertoire of responding cells. The B1-8i mouse strain was created

essed Ag (7). Priming of Ag-specific B and T cells is thought to by targeted insertion of a VH186.2(DFL16.2)JH2 gene into the H occur in the B cell and T cell zones of secondary lymphoid tissues chain (HC) locus of a 129/Ola ES cell (14). B cells that coexpress (8). Within 1–3 days after immunization, IgMϩ Ab-secreting cell this B1-8 HC transgene with an endogenous ␭1-LC express the a by guest on October 2, 2021 (ASC) clusters appear in bridging channels between the red pulp allotypic variant of the HC and confer specificity to the hapten NP. and the periarteriolar lymphoid sheathes of the spleen or in the Because not all ␭-LC-bearing B cells that pair with the B1-8 HC medullary cords of lymph nodes (9). The bridging channel ASC have the same NP-binding affinity, the ␭-LC- bearing B cells in the are thought to arise largely from marginal zone B cells (10). In B1-8i strain do not act as a true monoclonal population. Never- contrast to activated marginal zone B cells, Ag-primed follicular B theless, because the predominant ␭-LC in mice is ␭1, we consid- cells and T cells relocate to and interact at the junction between the ered this NP-binding population to be functionally monoclonal. B cell and T cell zones. There, B cells present processed Ag to In wild-type (WT) mice, as the immune response progresses, ϩ CD4 TH cells (11). Class switch recombination (CSR) also oc- increasing numbers of high-affinity ASC and memory B cells are detected in secondary lymphoid tissues and bone marrow (BM). Clonal proliferation of high-affinity B cells is thought to be the Department of Microbiology, University of Alabama at Birmingham, Birmingham, result of competition for growth signals between high-affinity B AL 35294 cells, low-affinity B cells, and B cells with no affinity to the Ag. Received for publication October 10, 2007. Accepted for publication August 1, 2008. This competition is defined as interclonal and is widely accepted as The costs of publication of this article were defrayed in part by the payment of page a mechanism for selection of high-affinity B cells within the GC. charges. This article must therefore be hereby marked advertisement in accordance We demonstrate that, in addition to interclonal competition be- with 18 U.S.C. Section 1734 solely to indicate this fact. tween B cells that express different BCR, elevated numbers of B 1 This work was supported by National Institute of Allergy and Infectious Diseases— cells expressing similar or identical Ig genes undergo intraclonal National Institutes of Health Training Grant T32 AI007493 (to T.L.L.). competition. When elevated, this form of competition results in the 2 Current address: Department of Microbiology & Immunology, Emory University School of Medicine, 1510 Clifton Road, Atlanta, GA 30322. reduction of the high-affinity Ab response and alters the pathways 3 Address correspondence and reprint requests to Dr. David D. Chaplin, Depart- that lead to generation of high-affinity ASC. ment of Microbiology, University of Alabama at Birmingham, 845 19th Street South, Birmingham, AL 35294. E-mail address: [email protected] Materials and Methods 4 Abbreviations used in this paper: SHM, somatic hypermutation; ASC, Ab-secreting cells; BM, bone marrow; HC, H chain; KLH, keyhole limpet hemocyanin; LC, L Mice chain; NP, (4-hydroxy-3-nitrophenyl)acetyl; CSR, class switch recombination; GC, ␬Ϫ/Ϫ germinal center; WT, wild type; MHC II, MHC class II; NIP, 4-hydroxy-3-iodo-5- C57BL/6, C57BL/6(Ig ), and C57BL/6(CD45.1) mice were purchased nitrophenyl acetyl; R:S, replacement: silent; VLP, virus-like particle. from The Jackson Laboratory and 129/Ola mice were purchased from Har- lan. The B1-8iϩ/Ϫ knock-in mouse strain was produced by Sonoda et al. Copyright © 2008 by The American Association of Immunologists, Inc. 0022-1767/08/$2.00 (14) and was generously provided by F. Alt (Harvard Medical School,

www.jimmunol.org 6028 CLONALLY ABUNDANT B CELLS BLOCK AFFINITY MATURATION

Boston, MA). It was backcrossed to the C57BL/6 strain for at least 10 ELISPOT generations. Previous studies have demonstrated that the B1-8 knock-in o ␮ allele effectively excludes the endogenous allele (14). Therefore, we used MAHA (Millipore) plates were coated overnight at 4 C with 10 g/ml heterozygous mice in all subsequent studies. Mice were immunized i.p. NP24BSA or NP2BSA, then washed, and incubated for1hat37°C with with 50 ␮g of NP-keyhole limpet hemocyanin (KLH; Biosearch Technol- blocking buffer. BM cells were harvested and suspended in RPMI 1640 ogies) precipitated in alum (Sigma-Aldrich) and boosted with 25 ␮gof supplemented with 10% FBS. Cells were added to each well at the con- NP-KLH at day 25. All immunization protocols followed this time line centrations described in the text and cultured for 12 h at 37°C in a humid- and dose unless stated in the text. Mice were housed in a barrier facility ified chamber under 5% CO2. All subsequent blocking, washing, and in- and maintained under protocols approved by the Institutional Animal Care cubation with primary and secondary Abs were performed as described for and Usage Committee at the University of Alabama at Birmingham. ELISA. Bound anti-mouse HRP-conjugated Abs were detected with 3-ami- no-9-ethylcarbazole (Moss). Spots were systematically counted using the CTL ImmunoSpot Reader and analyzed with ImmunoSpot software (Cel- Abs and reagents lular Technologies). Monoclonal mouse anti-IgMa-PE, anti-IgMb-FITC, anti-B220-PerCP, anti- B220-allophycocyanin-Cy7, anti-IgG1a-biotin, anti-IgG1b-biotin, anti- Lymphocyte adoptive transfer CD138-PE, and anti-MHC II-PE, were purchased from BD Biosciences. ϩ Ϫ For adoptive transfer experiments, the B1-8i / mice were crossed with Anti-CD22-PE-Cy5 was purchased from Abcam and anti-CD38-FITC was Ϫ Ϫ Ig␬ / and CD45.1 mice, also on the C57BL/6 background. Recipient purchased from eBioscience. Monoclonal rat anti-mouse IgM-HRP, anti- C57BL/6 mice were prepared for adoptive transfer by exposure to 500 rad IgG1-HRP, and streptavidin-HRP were purchased from Southern Biotech- from a cesium source. Two days after irradiation, the recipients were infused nology Associates. Polyclonal goat anti-mouse IgM, goat anti-mouse IgG, i.v. with a mixture of 1 ϫ 108 spleen cells from C57BL/6 mice and either 104 and mouse IgG1 Abs were purchased from Sigma-Aldrich. NP-allophyco- ϩ Ϫ or 106 purified B cells from either C57BL/6, CD45.1, or B1-8 / ϫ cyanin was made by coupling NP-Osu (Biosearch Technologies) with al- Ϫ Ϫ ␬ / ϫ CD45.1 mice. The purified B cells were prepared by depletion of lophycocyanin (ProZyme) in N,N-dimethylformamide. ϩ CD43 cells with CD43 microbeads followed by magnetic-associated cell Downloaded from sorting with the AutoMACS Separator (Miltenyi Biotec). Two days after Immunofluorescence microscopy adoptive transfer, the mice were immunized i.p. with NP-KLH as described above. Freshly isolated tissues were embedded in Tissue Tek OCT compound (Fisher Scientific) and frozen by floating the tissues on liquid nitrogen- chilled 2-methylbutane. Eight- to 10-␮m sections were cut on a cryostat Nucleotide sequencing (Leica), air dried on Superfrost Plus slides (Fisher Scientific), and fixed for Spleen RNA was harvested using a RNeasy kit (Qiagen). All subsequent Ϫ http://www.jimmunol.org/ 10 min in acetone at 4°C before storage at 20°C until further use. Non- reagents for purification of RNA and sequencing were purchased from specific binding was blocked using a combination of 10% rabbit and 10% Invitrogen unless stated otherwise. cDNA was prepared using SuperScript goat serum together with a biotin blocking kit (Vector Laboratories), fol- ␥ III reverse transcriptase. VH186.2 sequences were amplified from 1HC lowed by staining with primary and secondary Abs or 4-hydroxy-3-iodo- transcripts using Pfu50 polymerase and two rounds of PCR with the fol- 5-nitrophenylacetyl (NIP)-biotin (NIP-biotin; Biosearch Technologies). Ј lowing nested primer pairs: VH186.2 forward, 5 gatggagctgtatcatgctcttctt Signals due to bound NIP were enhanced using streptavidin-HRP Alexa Ј Ј Ј ggcag3 and IgG1 reverse, 5 gctgctcagagtgtagaggtcagactgc-3 ; VH186.2 Fluor 350 Tyramide Signal Amplification System (Invitrogen). forward nest, 5Јatcgatttggcagcaacagctacagg3Ј, and IgG1 reverse nest, 5Јccggcctcaccatggagttagtttgg3Ј. The amplified products were purified us- Flow cytometry ing the ImmunPure PCR cleanup kit, cloned into the PCR-Blunt II TOPO vector using the Zero-Blunt TOPO cloning kit, and transformed into

Single-cell suspensions from spleen tissues were stained with the indicated DH5␣-competent cells. Kanamycin-resistant colonies were picked and by guest on October 2, 2021 Abs or NP-allophycocyanin. For adoptive transfer experiments, purified B plasmids were purified with the Wizard SV 96 plasmid purification system cells were prepared by incubation with MACS anti-CD43 beads and frac- (Promega). Sequence analysis was performed at the Genomics Core Fa- tionation with the AutoMACS (Miltenyi Biotec) according to the manu- cility of the Howell and Elizabeth Heflin Center for Human Genetics of the facturer’s protocols. For intracellular staining, cells were incubated with University of Alabama at Birmingham using the BigDye Terminator Ver- mAb 24. G2 and anti-mouse IgG1 to block Fc receptors and surface IgG1. sion 3.1 Cycle Sequencing Ready Reaction Kit (catalog no. 4336919) and The cells were fixed and permeabilized using CytoFix/Cytoperm (BD Bio- an Applied Biosystems 3730 Capillary Sequencer. All sequences were an- sciences) and then intracellular IgG1 was stained according to the manu- alyzed using the Vector NTI Advance Software (Invitrogen). facturer’s guidelines. Cytometric data were acquired using an LSR II (BD Biosciences) and analyzed using FlowJo software (Tree Star). Statistical analysis

ELISA All ELISA and ELISPOT data were analyzed by performing a one-tailed t test assuming unequal variance. Signiﬁcance was deﬁned as p Ͻ 0.05. Immunlon plates (Thermo Labsystems) were coated with 10 ␮g/ml

NP24BSA or NP2BSA (Biosearch Technologies) overnight at 4°C. The plates were washed with 0.05% Tween 20 in PBS (PBS-Tween20) and Results ϩ Ϫ blocked for1hat37°C with blocking buffer consisting of 1% BSA in NP-reactive B cells in B1-8i / mice participate in the GC PBS-Tween 20. All subsequent incubations were at 37°C for 1 h. Sera were reaction diluted in blocking buffer, added to the NP-BSA-coated plates, and incubated for 1 h. Unbound Abs were removed by washing and bound Abs To identify how frequently mature NP-reactive B cells are found in were detected using anti-mouse IgM-HRP or anti-mouse IgG1-HRP. For the peripheral lymphoid tissues of the B1-8iϩ/Ϫ mouse strain, we allele-specific ELISA, NP-reactive IgG1 Abs that were produced by WT B b analyzed spleen lymphocytes for NP binding by flow cytometry. In cells were detected by anti-IgG1 -biotin, while NP-reactive IgG1 Abs pro- ϩ/Ϫ Ͼ duced by B1-8 B cells were detected with anti-IgG1a-biotin followed by B1-8i mice, B cells that bind NP represent 1.5% of splenic streptavidin-HRP. Unbound Abs were removed by washing and ABTS lymphocytes (Fig. 1a). The specificity of binding with NP-allo- (Roche Applied Sciences) substrate was added. The reaction product was phycocyanin was demonstrated by preincubating spleen cells with detected at OD . Serial dilutions of mouse sera were performed to de- 405 15 ng/ml NP24BSA. This completely blocked subsequent binding termine the working dilutions and dilutions that produced OD that were of NP-allophycocyanin, whereas preincubation with BSA alone within the linear range for the assay were used. In the indicated experi- ␭ ments, the concentrations of Abs in mouse sera were normalized as fol- did not. Approximately 5% of murine B cells express -LC, with lows. ELISA plates were coated with polyclonal goat anti-mouse IgM or approximately one-half of these expressing ␭1. This correlates IgG followed by incubation with mouse IgM or IgG subclass isotype con- with the observed 2.5% of NP-binding splenic B cells in naive trol Abs. Secondary HRP-conjugated detection Abs were added as de- mice (Fig. 1b, right panel). In contrast, NP binding is undetectable scribed above. The concentrations of the standard Abs were converted into arbitrary units. The Ab concentrations in experimental samples were sim- in naive C57BL/6 splenic B cells (Fig. 1b, left panel). The fact that ilarly converted to arbitrary units based on the dilutions that yielded OD NP-reactive B cells were present in substantial numbers in B1- ϩ Ϫ measurements in the linear range. 8i / mice permitted the detection of Ag-induced changes in B The Journal of Immunology 6029

littermates was indistinguishable from the primary response. In contrast, in B1-8iϩ/Ϫ mice, the secondary response resulted in IgM levels that were increased 20-fold above baseline levels. Previous reports have established techniques to use steady-state ELISAs with differentially haptenated target Ags to estimate the relative afﬁnities of NP-reactive IgGs (15–18). Using this ap- proach, we detected total NP-reactive IgG1 Abs using ELISA

plates that were coated with NP24BSA. In contrast, by using plates that were coated with the same concentration of BSA carrying only

two NP residues for every BSA (NP2BSA), we detected IgG1s with only relatively high afﬁnities for the hapten. It should be indicated clearly that this technique actually detects the average avidity of the Ab for the hapten Ag; however, in the case of biva- lent IgG Abs, this measurement closely estimates afﬁnity (16).

Using this strategy, NP2BSA efﬁciently captures Abs that bind with either high avidity or high afﬁnity. Abs with low avidity or

low affinity manifest higher off rates from NP2BSA than from NP24BSA and are not detected using the sparsely haptenated target Ag (15, 17). The serum anti-NP IgG1 titer of B1-8iϩ/Ϫ mice was similar to Downloaded from that of WT littermates (Fig. 3b). Serum anti-NP IgG2b, IgG2c, and IgG3 were also similar between the two strains (data not shown). These data were consistent with previous reports that demonstrated that the increase in NP-specific B cells in B1-8i mice does not correlate directly with the levels of serum anti-NP IgGs (19). In- terestingly, after a booster immunization, B1-8iϩ/Ϫ mice produced http://www.jimmunol.org/ 4-fold less high-affinity anti-NP IgG1 Ab than C57BL/6 mice (Fig. 3c). This defect in affinity maturation persisted even after secondary and tertiary booster immunizations (data not shown). High B cell clonal abundance interferes with affinity maturation FIGURE 1. Enrichment of NP-specific B cells in B1-8iϩ/Ϫ knock-in ϩ/Ϫ mice. Spleen cells were harvested from naive C57BL/6 (WT) and B1-8iϩ/Ϫ Immunization of B1-8i mice with NP-KLH resulted in the re- mice. NP-reactive cells were detected with allophycocyanin-conjugated distribution of NP-reactive cells in a fashion that was similar to ϩ/Ϫ that described previously for adoptively transferred Ag-specific B

NP. a, Top, 1.6% of splenic lymphocytes from B1-8i mice demon- by guest on October 2, 2021 strated binding to NP-allophycocyanin. The numbers of NP-binding cells cells in the lymph nodes (6) and Ag-specific B cells in the spleen in naive C57BL/6 mice were below the levels of detection. Preincubation (12). Furthermore, the fact that specific immunization induced the ϩ/Ϫ of cells from B1-8i mice with 15 ng of NP24BSA completely blocked production of NP-reactive IgM and IgG Abs demonstrated that ϩ the binding of NP-allophycocyanin (bottom right). b, Naive B220 lym- B1-8iϩ/Ϫ mice were capable of responding to a primary and a ϩ/Ϫ phocytes from C57BL/6 and B1-8i mice. NP-specific B cells were be- secondary immunization and that their ability to undergo isotype low the level of detection by FACS in WT mice and enriched to 2.5% of ϩ ϩ Ϫ switching was preserved; however, the dramatic impairment of the B220 gate in B1Ϫ8i / mice. More than 90% of B cells expressed the B1-8 HC transgene (IgMa), and Ͻ10% expressed endogenous HC genes affinity maturation indicated that there is disturbed regulation of (IgMb). All NP-binding B cells express the B1-8 HC transgene. the humoral response in this transgenic model. There is precedent for impaired lymphocyte regulation in mice carrying abundantly expressed Ag receptors. For example, primary immunization of cell localization in the spleen. Following immunization with NP- TCR-transgenic mice such as those of the DO11.10 strain leads to KLH plus alum, NP-reactive B cells which were initially distrib- abnormally high levels of cell death and thymic involution (20). In uted evenly in the B cell zones of the spleen relocalized to form our experiments, we have considered that the aberrant affinity mat- foci at the T cell-B cell interface (Fig. 2a) and subsequently en- uration may have been due to a B cell intrinsic defect. Alterna- tered into GC reactions (Fig. 2c) in a fashion similar to previous tively, the abnormal response may not represent an intrinsic B cell reports (12). These changes in distribution of the NP-reactive cells impairment, but rather may represent the natural expression of were Ag specific since no changes in the distribution of NP-reac- normal immunoregulatory pathways. ϩ Ϫ tive cells were detected when B1-8i / mice were immunized A major distinction between B1-8iϩ/Ϫ mice and WT mice is the with KLH plus alum alone (Fig. 2b, right panels). proportion of recirculating B cells that has specificity for the hap-

ϩ/Ϫ ten NP. We hypothesized that the large numbers of NP-specific Affinity maturation is impaired in B1-8i mice precursor B cells in the B1-8iϩ/Ϫ mice (Fig. 1) may result in in- To address how the presence of increased numbers of NP-specific creased competition for limiting signals that were required for af- B cells affected the serum Ab response, we measured this response finity maturation of the Ab response to the hapten. It is well es- in immunized B1-8iϩ/Ϫ and C57BL/6 mice using an ELISA. Dur- tablished that affinity maturation and isotype switching both ing the primary response to NP-KLH, a 4–8-fold induction of require T cell help (13). We considered that the impaired affinity NP-specific IgM was measured in B1-8iϩ/Ϫ serum. The response maturation we observed in the B1-8iϩ/Ϫ mice might have been due peaked 21 days postimmunization (Fig. 3a). C57BL/6 mice re- to a disproportionately smaller TH cell population compared with sponded similarly; however, the response peaked at 7 days postim- the large number of NP-specific B cells in this strain. However, munization. Thereafter, the serum IgM titers declined. In the sec- when we expanded the available pool of KLH-specific TH cells ondary response, the serum anti-NP IgM response in WT by priming B1-8iϩ/Ϫ mice with KLH before immunization with 6030 CLONALLY ABUNDANT B CELLS BLOCK AFFINITY MATURATION

FIGURE 2. B1-8 NP-binding B cells redistribute to the B cell-T cell interface and enter GC in response to immunization with NP-KLH. a, C57BL/6 and B1-8iϩ/Ϫ mice were immunized with NP-KLH plus alum and sacrificed at the indicated times. The location of NP-specific B cells was demonstrated by binding of biotinylated NIP followed by counter- staining with the streptavidin-HRP Alexa Fluor 350 Tyramide Signal Amplification System. b, Specificity of the anti-NP response. Serial sections prepared from spleens of naive B1-8iϩ/Ϫ mice were preincubated Downloaded from with BSA or NP24BSA before incubation with biotinylated NIP (left panels). NIP binding was blocked by pre- ϩ/Ϫ incubation with NP24BSA. B1-8i mice were immunized with nonhapte- nated KLH plus alum and spleens were harvested on days 3, 5, and 10, and NIP-binding cells were identified http://www.jimmunol.org/ by incubation of sections with biotinylated NIP (right panels). NIP-binding cells showed minimal relocaliza- tion following immunization with KLH. NIP-binding cells were detected as in a. c, Ten days after immunization, GC in both WT and B1- 8iϩ/Ϫ mice contained predominately ϩ ␭-LC cells. by guest on October 2, 2021

NP-KLH, we did not restore affinity maturation (Fig. 4). This in- titer in WT mice (Fig. 3a). Unlike B1-8iϩ/Ϫ mice, the levels of dicated that T cell help was not the critical limiting factor for NP-specific IgM in the sera of mice that received either adoptively ϩ Ϫ affinity maturation in B1-8i / mice when immunized with NP- transferred WT B cells, 104 B1-8iϩ/Ϫ B cells, or 106 B1-8ϩ/Ϫ B haptenated carrier proteins. cells all were similar following immunization with 50 ␮gofNP- To test whether the frequency of Ag-specific B cells affects af- KLH (Fig. 5aii). The higher levels of serum IgM in immunized finity maturation, we used adoptive transfer of different numbers of WT mice compared with the levels in the mice that received adop- ϩ/Ϫ B1-8i B cells into naive C57BL/6 mice to regulate the starting tively transferred WT or B1-8iϩ/Ϫ B cells may have been due to ϫ 8 ϩ Ϫ numbers of NP-reactive cells (Fig. 5ai). Spleen cells (1 10 ) the steady input of naive B1-8i / B cells that occurs in an on- from C57BL/6 mice and varying numbers of either CD43-depleted going fashion in the B1-8iϩ/Ϫ mouse strain. The induced levels of CD45.1ϩ B cells from WT-congenic donors or CD43-depleted B1- ϩ/Ϫ Ϫ/Ϫ ϩ total anti-NP serum IgG1 were similar in all groups of mice despite 8i ␬ CD45.1 B cells were adoptively transferred i.v. into ϩ Ϫ a 100-fold difference in the starting frequencies of B1-8i / B sublethally irradiated recipients. In naive B1-8iϩ/Ϫ mice, ϳ2.5% cells between the mice that received the low and high numbers of of spleen B cells were NP reactive (Fig. 1b). This represents ϳ1 ϫ adoptively transferred NP-reactive cells (Fig. 5aiii). This was con- 106 NP-specific B cells out of an average of ϳ5 ϫ 107 total B cells 6 sistent with the observation that the numbers of NP-specific B cells in the adult mouse spleen. Therefore, adoptive transfer of 10 B1- 8iϩ/Ϫ␬Ϫ/ϪCD45.1ϩ B cells into naive irradiated recipient mice do not correlate with the titers of NP-specific IgG1 (19) (Fig. 3b). produced recipients with NP-specific B cells at a level similar to Strikingly, mice that received the high frequency of adoptively that detected in the B1-8iϩ/Ϫ mice. In contrast, adoptive transfer of transferred NP-specific B cells produced 11-fold less high-affinity 104 donor B1-8iϩ/Ϫ B cells yielded recipients in which NP-reac- anti-NP IgG1 than control mice. In contrast, mice that received the ϩ/Ϫ tive transgenic B cells represented ϳ0.025% of the total B cell low frequency of adoptively transferred B1-8i B cells produced population, a frequency closer to the estimated frequency in WT levels of high-affinity anti-NP IgG1 only 2-fold lower than mice adult mice. reconstituted with WT B cells (Fig. 5aiv). Together, these data B1-8iϩ/Ϫ mice responded to immunization with NP-KLH with demonstrated that total production of Ag-specific serum Ab was an ϳ8-fold higher titer of hapten-specific IgM compared with the independent of the starting numbers of Ag-specific B cells, but that The Journal of Immunology 6031

FIGURE 4. The induction of a high-affinity Ab response is not restored in B1-8iϩ/Ϫ mice by priming with the T cell Ag. C57BL/6 or B1-8iϩ/Ϫ mice were immunized with 50 ␮g of KLH precipitated in alum. Three weeks after priming, the mice were reimmunized i.p. with 50 ␮gofNP- KLH plus alum. Sera were collected 14 days after immunization and serum Downloaded from anti-NP Ab was measured by ELISA. All sera were diluted 1/200. This working concentration was determined to yield measurements within the linear range of the assay for all samples tested. a, Total anti-NP IgG1 titers were similar in WT and B1-8iϩ/Ϫ mice. Despite priming, B1-8iϩ/Ϫ mice showed impaired ability to undergo affinity maturation. b, The ratio of serum high-affinity anti-NP IgG1 to total serum anti-NP IgG1 was reduced ϩ/Ϫ

in B1-8i mice. Each data point represents one mouse. This experiment http://www.jimmunol.org/ was repeated three times with similar results.

cells was suppressed in a B1-8 cell number-dependent manner (Fig. 5b, i and ii). The titer of total anti-NP IgG1a was similar regardless of the numbers of B1-8 donor B cells (Fig. 5biii). When B1-8 B cells were present in high numbers, affinity maturation was inhibited (Fig. 5biv). In contrast, affinity maturation by guest on October 2, 2021 was observed when low numbers of B1-8 B cells were transferred. These data established that B1-8 B cells were proficient to undergo affinity maturation when their population density was low. The inverse correlation between the numbers of B1-8 B cells transferred and the amount of anti-NP Ab produced by the WT B cells represent interclonal competition between the functionally monoclonal B1-8 B cells and the diverse WT B cell repertoire. Although the frequency of B1-8 cells did not have a large affect on either the amount of IgM or the total (high affinity plus low affinity) IgG1 produced, affinity maturation was attenuated when the B1-8 B cell frequency was high. These data indicate that the density of the responding cell population does not affect the overall quantity of the Ab response but does sig- FIGURE 3. Affinity maturation is suppressed in B1-8iϩ/Ϫ mice. Mice nificantly alter the quality of the serum Ab. were immunized i.p. with NP-KLH plus alum on day 0 and boosted at the time indicated by the arrow. a, Measurement of NP-specific IgM by ELISA Clonal selection is modulated by the density of the indicates the anti-NP IgM response is elevated in B1-8iϩ/Ϫ mice. b, In- responding B cell population duction of isotype class switching and the anti-NP IgG1 response occur to ϩ/Ϫ To investigate whether the defect in affinity maturation observed in similar levels in B1-8i mice compared with WT mice. c, The high- ϩ/Ϫ ϩ Ϫ the B1-8i mice was due to aberrant SHM, we sequenced the affinity anti-NP IgG1 response was severely reduced in B1-8i / mice. n V 186.2-␥1 ϩ/Ϫ represents the number of mice in each experimental group. The experiment H transcripts from immunized C57BL/6 and B1-8i was repeated three times with similar results. Error bars represent SEM. mice. Affinity maturation was tracked by the accumulation of the affinity-determining tryptophan to leucine mutation at codon 33 (W33L) in the sequence of individual clones. This single point production of high-affinity Ab was dependent on maintenance of mutation of the germline W to L results in a 10-fold increase in an appropriately low frequency of Ag-specific B cells. affinity of the Ab for NP (21). Seventy to 80% of unique ␥ To determine the contributions of the transgenic and the WT VH186.2- 1 transcripts amplified from spleens of immunized B cells to the NP-specific Ab response, we performed allele- C57BL/6 mice contained the high-affinity W33L mutation in con- specific ELISA. When mice that received adoptive transfer of trast to 50% of transcripts from B1-8iϩ/Ϫ spleens. Furthermore, NP-specific B1-8 B cells were immunized with NP-KLH, the the ratio of replacement:silent (R:S) mutations in the CDR1 and high-affinity anti-NP Ab response from the endogenous WT CDR2 from C57BL/6 mice was 3.2 vs 1.5 from B1-8iϩ/Ϫ mice. 6032 CLONALLY ABUNDANT B CELLS BLOCK AFFINITY MATURATION Downloaded from http://www.jimmunol.org/

FIGURE 5. Attenuation of affinity maturation is a consequence of high clonal abundance in B1-8iϩ/Ϫ mice. a, i, Sera collected 14 days after a booster immunization of mice that had received adoptive transfer of 108 WT spleen lymphocytes with either 106 WT polyclonal B cells (control), 104 (low), or 106 (high) B1-8␬Ϫ/ϪCD45.1ϩ B cells. NP-specific IgM (ii), and total anti-NP IgG1 (iii) were similar in all experimental groups. iv, Mice that received adoptive transfer of 106 B1-8␬Ϫ/ϪCD45.1 B cells produced a 10-fold lower titer of high-affinity anti-NP IgG1 than control mice, whereas mice that received lower numbers (104 cells/mouse) produced a high-affinity anti-NP IgG1 response. b, i, B1-8 B cells suppressed the WT total anti-NP IgG1 response. ii, Affinity maturation of the WT response was also disrupted in a cell density-dependent manner. iii, B1-8 B cells produced total anti-NP IgG1 in a fashion by guest on October 2, 2021 independent of clonal abundance. iv, Induction of a high-affinity NP-IgG1 response by the B1-8 B cells was observed when B1-8 B cells were transferred Values of p comparing control mice and mice that received 104 B1-8 B cells; †, values of p comparing WT mice and ,ء .(at low cell numbers (104/mouse mice that received 106 B1-8 B cells. Four mice were used in each experimental group. The adoptive transfer experiment was performed three times with similar results.

These data indicate that selection of high-afﬁnity Ab-expressing B quences carrying the W33L mutation was 58%, (35–77%; Table I). cells is impaired in immunized B1-8iϩ/Ϫ mice. Sixty-two percent of sequences from B1-8 B cells in the mice that ␥ 4 VH186.2- 1 sequences were also analyzed in the adoptive trans- received adoptive transfer of 10 B1-8 B cells carried the W33L fer model. During the generation of the B1-8iϩ/Ϫ strain, a silent T mutation, ranging from 43 to 76%. In contrast, mice that received to C mutation was engineered into the HC cryptic recombination adoptive transfer of 106 B1-8 B cells manifested a lower distribu- signal sequence at nt 292 to prevent VH gene replacement (14). tion of clones carrying the high-afﬁnity mutation (22–58%, with a This mutation and the DFL16.1-JH2 rearrangement distinguished mean of 41%). Although the accumulation of point mutations was the B1-8 sequences from the nontransgenic sequences. For mice only modestly reduced (data not shown), these data suggest that carrying only WT HC genes, the mean percentage of unique se- when there are abnormally high numbers of B cells with the same

␥ Table I. Summary of VH186.2- 1 Sequences

Total Sequences Unique Group Mouse Determined Sequences % Unique R:S Mean R:S % W33L Mean % W33L

Control 1 20 15 86.06 2.27 2.38 66.67 58.14 2 23 22 2.63 77.27 3 21 17 2.35 35.29 4 15 14 1.66 42.86 Low 104 B1-8 1 28 22 80.49 2.4 2.77 68.18 62.43 2 24 22 3.61 p ϭ 0.1 72.73 3 30 22 2.28 63.64 High 106 B1-8 1 21 17 63.86 1.67 1.34 50 40.92 2 20 12 1.5 p Ͻ 0.0005 58.33 3 21 9 1.33 22.22 4 21 15 0.88 31.25 The Journal of Immunology 6033 antigenic specificity, the frequency of transcripts that accumulate mutations resulting in a high-affinity Ab product is reduced. Inter- estingly, similar to C57BL/6 mice, the mean R:S ratios of mutations in the HC were 2.2 and 2.5, respectively, for the control mice and the mice that received adoptive transfer of the low frequency of B1-8 B cells, indicating a strong selection for replacement mutation within the CDRs. The mean R:S ratio in mice that received the high frequency of B1-8 B cells was 1.4, confirming in the adoptive transfer system that reduced selection is observed when the frequency of clonally related B cells is high. Intraclonal competition suppresses formation of high-affinity ASC Analysis of the serum Abs produced after immunization indicated that production of high-affinity anti-NP IgG1 was dramatically reduced in B1-8iϩ/Ϫ mice and mice that received 106 B1-8 B cells. In contrast, DNA sequencing indicated that approximately one- a ␥ half of the VH186.2- 1 transcripts amplified from spleens of these mice were from cells that potentially could produce high-affinity Ab because the included the high-affinity encoding W33L muta- Downloaded from tion. Because each of the sequences analyzed displayed a unique nucleotide sequence, the possibility that the discrepancy between the measurements of hapten-binding avidity and the presence of the high-affinity-determining codon 33 mutation might be due to the PCR artifact representing contamination of the samples with plasmid FIGURE 6. The ASC repertoire is largely limited to low-affinity Ab http://www.jimmunol.org/ DNA is very unlikely. A more likely alternative explanation for the when B1-8 B cells are abundant. a, ELISPOT assays were performed using discrepancy between the sequence data and the ELISA data is that C57BL/6 or B1-8 BM to test for the presence of long-lived ASC. a, B1- the sequenced transcripts were from a pool of total spleen cells, 8iϩ/Ϫ mice produced ϳ4-fold more NP-specific IgM-secreting cells than ␥ ϳ representing the entire expressed VH186.2- 1 repertoire including WT mice, 1.8-fold less total NP-specific IgG1 ASC, and 3-fold less transcripts for secreted Ab from ASC and membrane-bound Ab high-affinity NP-specific IgG1 ASC. b and c, BM from mice adoptively from memory or activated B cells. In contrast, the serum Abs mea- transferred with WT B cells or B1-8 B cells. b, The frequency of long-lived sured by ELISA derive from cells driven to produce secreted Ab. total NP IgG1 ASC is not affected by B1-8 population density. c, B1-8 B ϩ Ϫ cells from mice adoptively transferred with the high frequency of B cells To determine whether cells in the B1-8i / mice were compe- do not produce high-affinity IgG1 ASC. Each data point represents one tent to produce NP-specific Abs, particularly high-affinity Abs, we mouse. The experiment was repeated three times with similar results. by guest on October 2, 2021 used ELISPOT assays to identify high-affinity and total anti-NP IgG1-producing cells in BM of immunized mice. There were 3-fold greater numbers of high-affinity IgG1 ASC in BM of C57BL/6 mice compared with BM of B1-8iϩ/Ϫ mice (Fig. 6a). of adoptively transferred cells. Although there was an increase Interestingly, when WT mice were sublethally irradiated and re- in the numbers of CD138high ASC in the mice that received 106 constituted with WT splenocytes plus 106 B1-8 B cells, then im- adoptively transferred cells, the percentage of long-lived ASC munized, the numbers of high-affinity anti-NP IgG1 ASC in BM in these animals was significantly reduced compared with that were also reduced (Fig. 6, b and c). In contrast, reconstitution of in WT mice and mice that had received 104 adoptively trans- mice with WT splenocytes plus 104 B1-8 B cells before immuni- ferred cells (Fig. 7a). zation showed a near normal high-affinity ASC response. This sug- We have demonstrated that the development of long-lived ASC gested that the transcripts carrying the high-affinity-determining can be influenced and the numbers of ASC can be suppressed W33L mutation that we observed in the immunized B1-8iϩ/Ϫ mice when unusually high numbers of Ag-specific B cells are present. and in WT mice that had received 106 adoptively transferred B1-8 Previous studies from Marzo et al. (23) have established that fol- B cells were unlikely to contribute to the long-lived ASC pool and lowing immunization, large numbers of clonally restricted, Ag- circulating Ab. specific CD8 T cells revert from an effector memory CD8ϩ T cell Under normal conditions, the pool of long-lived ASC elicited in phenotype into a central memory CD8ϩ T cell phenotype. We response to a systemic infection resides primarily in the BM and is hypothesize that the large numbers of NP-specific B cells present composed largely of high-affinity ASC while the pool of memory in the B1-8iϩ/Ϫ strain in similar fashion may be impaired for dif- B cells in the BM is thought to be composed of a mixture of high ferentiation into the high-affinity long-lived ASC compartment. and low or primarily low-affinity cells (22). Our observation that This impairment may not be restricted to the long-lived ASC com- the numbers of high-affinity ASC are reduced in the BM of B1- partment alone, but may be observed in the memory compartment 8iϩ/Ϫ mice indicated that the mechanisms governing ASC and/or as well. By definition, memory B cells have previously encoun- memory B cell development are disrupted in these animals. To tered Ag and most have class switched. Functionally, memory B address whether the effector ASC and memory B cell populations cells are thought to be more potent APC compared with naive B are altered when B cells are subjected to intraclonal competition, cells. Following Ag restimulation, they up-regulate expression of we examined BM from control mice and from mice that had re- MHC II molecules and other activation molecules more rapidly ceived adoptive transfer of either 104 or 106 B1-8 B cells before than naive cells. In the CD43Ϫ cell population, IgG1ϩNP-binding primary and booster immunization. Long-lived intracellular cells from both WT mice and mice that had received 104 B1-8 B IgGϩCD138ϩ ASC make up 1.7% of the CD43ϩ fraction in WT cells were found to express surface MHC II at similar levels. In mice compared with 2.1% in mice that received the low numbers contrast, IgG1ϩNP-binding cells from mice that had received 106 6034 CLONALLY ABUNDANT B CELLS BLOCK AFFINITY MATURATION Downloaded from

wt B18

150 µg http://www.jimmunol.org/ FIGURE 8. Additional immunizing Ag restores affinity maturation in B1-8iϩ/Ϫ mice. C57BL/6 mice and B1-8iϩ/Ϫ mice were immunized with the indicated dose of Ag precipitated with a constant amount of alum and boosted 25 days after the initial immunization. Sera were collected and analyzed for high-affinity anti-NP IgG1 at the designated time points. a, Induction of total anti-NP IgG1 was similar in all mice irrespective of Ag dose. b, Induction of high- affinity Ab was restored in B1-8iϩ/Ϫ mice immunized with 150 ␮g of NP-KLH plus alum. The experiment was repeated three times with similar results.

FIGURE 7. The formation of long-lived ASC is suppressed when B1-8 by guest on October 2, 2021 B cells are overrepresented. BM cells were recovered from adoptive transfer mice 14 days after the booster immunization. a, CD43ϩ MACS-sorted cells from control mice or from mice that received 104 cells or 106 cells Ag-specific B cells along with T cell-dependent signals that pro- were stained for CD138 and intracellular IgG1 to detect IgG1 ASC. b, vide B cell help. Since our initial studies showed that preactivation Ϫ CD43 BM cells were gated for NP binding and intracellular IgG1 staining of carrier-specific TH cells by priming with the carrier protein (top) and analyzed for B220 (left), CD22 and CD38 (center) and MHC II failed to restore affinity maturation, we suspected that T cell help, ϩ expression (right). Numbers to the left indicate the percentage of IgG1 per se, was not limiting for the affinity maturation response. We NP binding in CD43Ϫ cell fractions. Numbers inside the gates indicate the low/Ϫ high therefore tested whether Ag-dependent signals were limiting by percentage of gated cells that are B220 (red) and B220 (green). ␮ low/Ϫ immunizing mice with an increased dose of the Ag (150 gof B220 plasma and B cell blasts (red) down-regulate expression of ␮ CD22. B220high (green) up-regulate CD38 and CD22 expression (center). NP-KLH, rather than the 50- g dose used originally). Although Control mice (gray histogram) and mice transferred with 104 B1-8 B cells the titer of total NP-specific IgG1 remained unchanged in WT and ϩ/Ϫ (red) expressed equal levels of MHC II while IgG1ϩ NP-binding cells from B1-8i mice, immunization with the higher Ag dose induced ϩ/Ϫ mice transferred with 106 B1-8 B cells (blue dashes) up-regulate expres- high titers of high-affinity NP-specific IgG1 in the B1-8i mice sion of MHC II. (Fig. 8). To test whether immunization with high-dose Ag restored the adoptively transferred B1-8 B cells showed substantially up-reg- production of high-affinity Abs in the adoptive transfer model, we ϩ ulated expression of class II MHC (Fig. 7b). IgG1 low NP-bind- analyzed mice that received 106 WT B cells or 106 B1-8 ϫ␬Ϫ/Ϫ B Ϫ ing cells were predominately B220low/ . This population consisted cells by adoptive transfer before immunization with either 50 ␮g ϩ of a mixture of low-affinity ASC and B cell blasts. The IgG1 /high or 150 ␮g of NP-KLH plus alum. Mice treated in this fashion NP-binding population expressed high levels of B220 and was accumulated similar levels of total serum anti-NP IgG1 when im- composed of high-affinity, activated, and memory B cells that were munized and boosted with either the low or the high dose of Ag CD38highCD22int/high. Although mice that received adoptive trans- (Fig. 9a, f). In contrast, raising the dose of immunizing Ag from fer of 106 B1-8 B cells had reduced numbers of B cell blasts and 50 ␮gto150␮g increased the high-affinity IgG1 response nearly ASC, these mice did show enrichment of the high-affinity 3-fold (Fig. 9a, Ⅺ). To determine whether the B1-8 B cells con- B220highCD38highCD22int/high activated and memory phenotype B tributed to the high-affinity Ab production, we performed allele- cells (Fig. 7b). specific ELISA. As previously shown, when B1-8 B cells are 6 ϩ/Ϫ present at high cell density (following adoptive transfer of 10 High Ag dose restores affinity maturation in B1-8i mice cells), immunization with 50 ␮g of NP-KLH plus alum resulted in The process of affinity maturation of the Ab response requires suppression of the production of high-affinity Ab (Fig. 5biv) and delivery of a combination of BCR-dependent signals that activate also suppression of the generation of high-affinity ASC (Fig. 6c). The Journal of Immunology 6035

uration to permit the immune response to eliminate the Ag even as its concentration becomes very limited. In our current study, we have observed in the B1-8iϩ/Ϫ Ig HC-transgenic mouse strain that affinity maturation is impaired when the number of Ag-specific naive B cells is inappropriately large. We saw no evidence of failure to activate the SHM process. Rather, we observed that there was failure to select high-affinity B cell variants as a likely cause of the impaired generation of high-affinity ASC. This impairment of affinity maturation in the B1-8iϩ/Ϫ mice was not a B cell intrinsic defect. Affinity maturation could be reestablished either by reducing the number of Ag-specific naive B cells by adoptive transfer into a nontransgenic host or by immunizing with an increased amount of Ag. a In addition to having large numbers of B cells that are poised to respond to immunization with NP-haptenated Ag, the B1-8iϩ/Ϫ strain differs from WT mice in that it has elevated levels of circulating NP-binding serum Abs that are present before deliberate immunization with NP-haptenated Ag (data not shown). We have considered that the elevated levels of anti-NP IgM in naive B1- 8iϩ/Ϫ mice might affect the quality of the Ab response following Downloaded from immunization. Previous studies have demonstrated that infusion of Ag-specific Ab into naive mice can alter the quality of the humoral response following systemic immunization. Several studies (24– 26) found that injection of Ag-specific IgM before immunization enhanced the endogenous Ab response to subsequent immunization. This enhancement of the humoral response was dependent on http://www.jimmunol.org/ the formation of immune complexes by the passively transferred IgM (27). Arguing against a role for elevated levels of anti-NP ϩ Ϫ FIGURE 9. Immunization with an increased dose of Ag induces the IgM in the attenuated affinity maturation in the B1-8i / mice is production of high-affinity Ab in mice that received adoptive transfer of our observation that affinity maturation was attenuated both in B1- high numbers of B1-8 B cells. C57BL/6 mice were conditioned by adoptive 8iϩ/Ϫ mice and in WT mice that received adoptive transfer of 106 6 6 transfer of either 10 WT or 10 B1-8 B cells, then were immunized with B1-8 B cells (Fig. 5). Before immunization, the mice that received ␮ ␮ either 50 gor150 g of NP-KLH plus alum and boosted 25 days after the adoptive transfer of B1-8 B cells had levels of anti-NP IgM and primary immunization. a, Sera were collected on day 35 and analyzed for IgG equivalent to the levels observed in naive WT mice that are either total (f) or high-affinity (Ⅺ) NP-reactive IgG1 by ELISA. b, Total by guest on October 2, 2021 (f) and high-affinity (Ⅺ) anti-NP IgG1a Abs derived from the adoptively competent to exhibit normal levels of affinity maturation. transferred B1-8 cells were measured by ELISA. Data represent the means Studies by Freitas et al. (28) showed that when WT mice were of two independent experiments with three mice per group in each exper- reconstituted with a mixture of WT and BCR-transgenic BM, the iment. Error bars represent SEM. initial growth curves of WT and transgenic B cells were similar. Once steady state was reached, the WT B cells became the predominant population (28). The B cell population could be shifted Although the B1-8 B cells produced detectable high-affinity NP- toward greater representation of the transgenic B cells if the mice reactive IgG1a (Fig. 9b, Ⅺ) when mice were immunized with 50 were immunized with the Ag specific to the transgenic BCR (29). ␮g of NP-KLH, this represented only 30% of the total NP-reactive In other experiments, mice with conditional ablation of surface Ig Ab present in the blood. In contrast, when mice that had received had diminishing numbers of peripheral B cells, indicating that adoptive transfer of 106 B1-8 B cells were immunized with 150 ␮g BCR signaling is necessary for B cell survival in the periphery of NP-KLH, a robust affinity-matured Ab response was observed (30). (Fig. 9b, Ⅺ). Thus, in the adoptive transfer model, as well as in the In our adoptive B cell transfer experiments, we anticipated that ϩ Ϫ B1-8i / mouse strain itself, immunization with a high dose of Ag irradiation of the recipient mice created niches that support the restored the high-affinity Ab response. engraftment of all donor cells equivalently, irrespective of their These data have identified Ag-dependent signaling as a key re- clonal specificity (Ref. 28 and T. L. Le, unpublished data). Fur- source that B cells compete for in the induction of a high-affinity thermore, we anticipated that immunization with NP-haptenated Ab response. When the numbers of naive, clonally related, Ag- carrier proteins should favor selection of the NP-specific trans- specific cells are high and when the relative amount of the Ag is genic cells over the WT donor cell population. Our data indicate limited, then affinity maturation is attenuated. This is in sharp con- that when the numbers of cells from an individual B cell clone are trast to the situation in mice with a diverse polyclonal B cell rep- high, this leads to failure to select cells carrying somatically mu- ertoire in which cells compete for limiting Ag and clones with tated high-affinity Ig HC. Our data support the concept that inter- high-affinity BCRs expand. clonal competition selected for the NP-specific B1-8 B cells over the predominately Ag-nonspecific WT polyclonal B cells. Com- Discussion petition for limited resources was high because of the large rep- The normal T cell-dependent humoral immune response is char- resentation of NP-specific B1-8 B cells in the total B cell pool. acterized by activation of Ag-specific B cells and T cells, expan- Cells compete for limiting resources that are needed for growth sion of the pool of responding B cells to permit production of or survival, including cytokines, mitogens, or local cellular niches substantial amounts of circulating Abs, switching to isotypes ap- such as the GC. In our studies, increasing the dose of immunizing propriate for elimination of the immunizing Ag, and affinity mat- Ag was able to restore affinity maturation in B1-8iϩ/Ϫ mice in a 6036 CLONALLY ABUNDANT B CELLS BLOCK AFFINITY MATURATION fashion similar to reducing the B1-8 B cell clonal abundance by fer of large numbers of B1-8 B cells and because the amplified adoptive transfer into WT mice. Batista et al. (7) have demon- transcripts were from B cells that were activated and class strated that B cells are capable of acquiring membrane-bound Ag, switched, we conclude, rather, that recovery of multiple identical internalizing the Ag, and subsequently presenting the processed HC sequences was the result of there being a limited number of

Ag to TH cells. This ability to acquire membrane-bound Ag dis- mutations in the VH genes of the pool of activated B cells. This played on target cells provides a possible mechanism for affinity- indicated that in the setting of high intraclonal competition there based selection of B cells. When B1-8 B cells or NP-specific B was reduced efficiency of SHM. This was in the context of the total cells are rare and infrequent, the cells are able to test their affinity level of NP-specific serum IgG being similar in the B1-8iϩ/Ϫ mice and undergo affinity-based selection efficiently. B cells with low and the WT mice, as were the numbers of total NP-specific IgG- affinity have the potential to interact with the membrane-bound producing ASC. Although CSR and SHM are mechanistically Ag, but are at a disadvantage and unable to acquire and internalize linked by activation-induced cytidine deaminase (33), studies from the Ag as efficiently as high-affinity cells. These cells likely receive the laboratories of Honjo and colleagues (34) and Nussenzweig very low-level stimulation, are negatively selected in the GC, and and colleagues (35) have independently demonstrated that these undergo apoptosis if the Ab is nonspecific or autoreactive. How- processes can occur separately. Our data further underscore the ever, when B1-8 B cells are abundant, a fraction of the cells may fact that CSR and affinity SHM are not obligately linked. capture and internalize the Ag, preventing the remaining B1-8 B The mechanisms governing the formation of the memory and cells from contacting the now more limited amounts of Ag. Thus, long-lived ASC compartments remain controversial. The stochas- in the setting of high numbers of Ag-specific cells, the acquisition tic model suggests that BCR-independent signals such as IL-4 and of Ag by a minority population of cells may create an environment in which the majority of cells (which have not encountered the Ag) IL-5 drive ASC differentiation (36). In contrast, studies comparing Downloaded from are at a selective disadvantage, resulting in an overall reduced GC B cells with high-affinity BCR to cells with low-affinity BCR high-affinity response. demonstrated that high-affinity cells expressed Blimp-1, whereas B cell activation requires BCR cross-linking. This process is low-affinity cells did not (37). Also, analysis of VH gene sequences enhanced when the antigenic particle is displayed at high density. supported the hypothesis that the memory B cell pool expresses Immunization with virus-like particles (VLP) that displayed low, relatively low-affinity BCR while the ASC compartment in the BM medium, or high densities of an antigenic peptide induced similar produces Ab with relatively high affinity (22). These data support http://www.jimmunol.org/ IgM responses to the peptide and was not dependent on epitope a differential signaling model for determination of memory cell density. Alternatively, the IgG response to the peptide was se- and ASC formation. verely reduced when VLP with low-density peptide were used to In the setting of high intraclonal competition, we found that the immunize mice. Antipeptide IgG was induced only when high- numbers of high-affinity ASC were reduced and a large fraction of density peptide VLP were used to immunize mice (31). This in- the high NP-binding B220ϩ cells displayed a memory cell-like dicated that the induction of an IgG response requires multiple phenotype with enhanced expression of CD22 and MHC II. The Ag-capturing events and cross-linking of the BCR. Increasing the relative frequency of these NP-specific memory-like cells com- numbers of Ag-specific B cells skews the ratio of B cell:Ag, which ϩ/Ϫ pared with ASC in immunized B1-8i mice exceeded that in by guest on October 2, 2021 can result in insufficient activation or selection via the reduction of immunized WT mice and also in immunized mice that were re- cross-linking events. The lack of Ag-based selection may result in a cipients of 104 adoptively transferred NP-binding B1-8 B cells. loss of apoptotic signals, further altering the balance of high-affinity Additional studies will determine whether these phenotypically and low-affinity B cells and the pathway to memory or ASC dif- memory-like cells express functional memory. Nonetheless, the ferentiation. Mice that are reimmunized with Ag shortly after the data indicate that a low Ag-specific precursor B cell population is primary immunization have increased numbers of apoptotic bodies optimal for establishment of a high-affinity Ab response. The re- within the GC (32). This observation is in keeping with our finding duction of high-affinity ASC correlates with a shift toward gener- that increasing the amount of Ag can rescue affinity maturation by ation of memory-like B cells albeit with a reduction in the numbers enhancing Ag-dependent selection in the GC. Our data show that of total NP ASC and memory cells. Together, these data indicate isotype switching can occur in the context of failed selection for that there were insufficient signals or growth factors to support the production of high-affinity Abs. Specifically, we find that high in- development of high-affinity ASC. These signals most likely orig- traclonal competition results in failure of somatically mutated cells inate from cells that reside within the GC structure. to progress to ASC. Our observation that immunization with in- Our data demonstrate clearly that the normal differentiation creased amounts of Ag can restore affinity maturation suggests that pathway that guides isotype-switched B cells through the process this intraclonal competition is based at least in part on Ag being a of SHM, selection of high-affinity variants, and progression either limiting resource for the subsequent selection of high-affinity B to ASC formation or B memory cell formation is dependent on cell clones to generate high-affinity ASC when B1-8 B cells are present in large numbers. maintaining limited intraclonal competition. Our studies have been facilitated by the well-characterized B1-8iϩ/Ϫ BCR-trans- When the VH186.2 sequences of IgG1 transcripts were compared between NP-KLH-immunized and boosted control mice and genic mouse strain (14) in which the precursor frequency of mice that received 106 B1-8 B cells, we observed that there were naive B cells specific for the small hapten NP can be manipu- small numbers of unique sequences and a high frequency of se- lated over a broad range. The extent to which the mechanisms quences that were identical in the mice that received transgenic B identified here will dramatically impact the humoral responses cells (Table I). These data suggested that although B cell activation to complex, multi-epitope protein Ags remains to be tested. With this caveat in mind, our data highlight the need to be and class switching occurred, the repertoire of VH gene diversity generated by SHM was very limited. It is formally possible that aware of the stoichiometry under which B cells are activated in recovery of a large number of identical sequences was the result of vivo. Furthermore, our data establish the presence of previously selective PCR amplification of a limited subset of the expressed VH unrecognized control points that govern important B cell dif- sequences; however, because recovery of large numbers of iden- ferentiation processes that follow class switching, leading to the tical sequences was restricted to mice that received adoptive trans- expression of an affinity-selected Ab response. The Journal of Immunology 6037

Acknowledgments 19. Shih, T. A., E. Meffre, M. Roederer, and M. C. Nussenzweig. 2002. Role of BCR affinity in T cell dependent antibody responses in vivo. Nat. Immunol. 3: We thank E. Lefkowitz for helpful discussions regarding analysis of so- 570–575. matic mutation and clonal selection, K. Rajewsky for permitting access to 20. Murphy, K. M., A. B. Heimberger, and D. Y. Loh. 1990. Induction by antigen of the B1-8iϩ/Ϫ mouse strain, and C. Zindl, M. Cooper, P. Burrows, R. Carter, intrathymic apoptosis of CD4ϩCD8ϩTCRlow thymocytes in vivo. Science 250: J. Kearney, and C. T. Weaver for helpful discussions and for critique of this 1720–1723. 21. Allen, D., T. Simon, F. Sablitzky, K. Rajewsky, and A. Cumano. 1988. Antibody manuscript. engineering for the analysis of affinity maturation of an anti-hapten response. EMBO J. 7: 1995–2001. Disclosures 22. Smith, K. G., A. Light, G. J. Nossal, and D. M. Tarlinton. 1997. The extent of affinity maturation differs between the memory and antibody-forming cell com- The authors have no financial conflict of interest. partments in the primary immune response. EMBO J. 16: 2996–3006. 23. Marzo, A. L., K. D. Klonowski, A. Le Bon, P. Borrow, D. F. Tough, and References L. Lefrancois. 2005. Initial T cell frequency dictates memory CD8ϩ T cell lin- eage commitment. Nat. Immunol. 6: 793–799. 1. Nossal, G. J., and J. Lederberg. 1958. Antibody production by single cells. Nature 181: 1419–1420. 24. Harte, P. G., A. Cooke, and J. H. Playfair. 1983. Specific monoclonal IgM is a 2. Li, Z., C. J. Woo, M. D. Iglesias-Ussel, D. Ronai, and M. D. Scharff. 2004. The potent adjuvant in murine malaria vaccination. Nature 302: 256–258. generation of antibody diversity through somatic hypermutation and class switch 25. Lehner, P., P. Hutchings, P. M. Lydyard, and A. Cooke. 1983. II. IgM-mediated recombination. Genes Dev. 18: 1–11. enhancement: dependency on antigen dose, T-cell requirement and lack of evi- 3. Banchereau, J., and R. M. Steinman. 1998. Dendritic cells and the control of dence for an idiotype-related mechanism. Immunology 50: 503–509. immunity. Nature 392: 245–252. 26. Powell, R., P. Hutchings, A. Cooke, and P. M. Lydyard. 1982. Antibody medi- 4. Itano, A. A., and M. K. Jenkins. 2003. Antigen presentation to naive CD4 T cells ated regulation of immune responses: I. Enhancement of specific antibody re- in the lymph node. Nat. Immunol. 4: 733–739. sponses through IgM antibodies. Immunol. Lett. 4: 253–258. 5. von Andrian, U. H., and C. R. Mackay. 2000. T-cell function and migration: two 27. Morgan, E. L., and W. O. Weigle. 1983. Polyclonal activation of murine B lym- sides of the same coin. N. Engl. J. Med. 343: 1020–1034. phocytes by immune complexes. J. Immunol. 130: 1066–1070. 6. Pape, K. A., D. M. Catron, A. A. Itano, and M. K. Jenkins. 2007. The humoral 28. Freitas, A. A., M. M. Rosado, A. C. Viale, and A. Grandien. 1995. The role of Downloaded from immune response is initiated in lymph nodes by B cells that acquire soluble cellular competition in B cell survival and selection of B cell repertoires. Eur. antigen directly in the follicles. Immunity 26: 491–502. J. Immunol. 25: 1729–1738. 7. Batista, F. D., D. Iber, and M. S. Neuberger. 2001. B cells acquire antigen from 29. McLean, A. R., M. M. Rosado, F. Agenes, R. Vasconcellos, and A. A. Freitas. target cells after synapse formation. Nature 411: 489–494. 1997. Resource competition as a mechanism for B cell homeostasis. Proc. Natl. 8. McHeyzer-Williams, L. J., and M. G. McHeyzer-Williams. 2005. Antigen-spe- Acad. Sci. USA 94: 5792–5797. cific memory B cell development. Annu. Rev. Immunol. 23: 487–513. 30. Lam, K. P., R. Kuhn, and K. Rajewsky. 1997. In vivo ablation of surface im- 9. Jacob, J., and G. Kelsoe. 1992. In situ studies of the primary immune response to munoglobulin on mature B cells by inducible gene targeting results in rapid cell

(4-hydroxy-3-nitrophenyl)acetyl: II. A common clonal origin for periarteriolar death. Cell 90: 1073–1083. http://www.jimmunol.org/ lymphoid sheath-associated foci and germinal centers. J. Exp. Med. 176: 31. Jegerlehner, A., T. Storni, G. Lipowsky, M. Schmid, P. Pumpens, and 679–687. high high M. F. Bachmann. 2002. Regulation of IgG antibody responses by epitope density 10. Oliver, A. M., F. Martin, and J. F. Kearney. 1999. IgM CD21 lymphocytes and CD21-mediated costimulation. Eur. J. Immunol. 32: 3305–3314. enriched in the splenic marginal zone generate effector cells more rapidly than the 32. Han, S., B. Zheng, J. Dal Porto, and G. Kelsoe. 1995. In situ studies of bulk of follicular B cells. J. Immunol. 162: 7198–7207. the primary immune response to (4-hydroxy-3-nitrophenyl)acetyl: 11. Lanzavecchia, A. 1985. Antigen-specific interaction between T and B cells. Na- IV. Affinity-dependent, antigen-driven B cell apoptosis in germinal centers as a ture 314: 537–539. mechanism for maintaining self-tolerance. J. Exp. Med. 182: 1635–1644. 12. Pape, K. A., V. Kouskoff, D. Nemazee, H. L. Tang, J. G. Cyster, L. E. Tze, K. L. Hippen, T. W. Behrens, and M. K. Jenkins. 2003. Visualization of the 33. Muramatsu, M., K. Kinoshita, S. Fagarasan, S. Yamada, Y. Shinkai, and genesis and fate of isotype-switched B cells during a primary immune response. T. Honjo. 2000. Class switch recombination and hypermutation require activa- J. Exp. Med. 197: 1677–1687. tion-induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell 102: 553–563. 13. Tarlinton, D. 2006. B-cell memory: are subsets necessary? Nat. Rev. Immunol. 6: by guest on October 2, 2021 785–790. 34. Ta, V. T., H. Nagaoka, N. Catalan, A. Durandy, A. Fischer, K. Imai, 14. Sonoda, E., Y. Pewzner-Jung, S. Schwers, S. Taki, S. Jung, D. Eilat, and S. Nonoyama, J. Tashiro, M. Ikegawa, S. Ito, et al. 2003. AID mutant analyses K. Rajewsky. 1997. B cell development under the condition of allelic inclusion. indicate requirement for class-switch-specific cofactors. Nat. Immunol. 4: Immunity 6: 225–233. 843–848. 15. Chen, Z., S. B. Koralov, M. Gendelman, M. C. Carroll, and G. Kelsoe. 2000. 35. Barreto, V., B. Reina-San-Martin, A. R. Ramiro, K. M. McBride, and Humoral immune responses in Cr2Ϫ/Ϫ mice: enhanced affinity maturation but M. C. Nussenzweig. 2003. C-terminal deletion of AID uncouples class switch impaired antibody persistence. J. Immunol. 164: 4522–4532. recombination from somatic hypermutation and gene conversion. Mol. Cell 12: 16. Herzenberg, L. A., S. J. Black, T. Tokuhisa, and L. A. Herzenberg. 1980. Mem- 501–508. ory B cells at successive stages of differentiation: affinity maturation and the role 36. Hasbold, J., L. M. Corcoran, D. M. Tarlinton, S. G. Tangye, and P. D. Hodgkin. of IgD receptors. J. Exp. Med. 151: 1071–1087. 2004. Evidence from the generation of immunoglobulin G-secreting cells that 17. Shimizu, T., M. Oda, and T. Azuma. 2003. Estimation of the relative affinity of stochastic mechanisms regulate lymphocyte differentiation. Nat. Immunol. 5: B cell receptor by flow cytometry. J. Immunol. Methods 276: 33–44. 55–63. 18. Smith, K. G., U. Weiss, K. Rajewsky, G. J. Nossal, and D. M. Tarlinton. 1994. 37. Phan, T. G., D. Paus, T. D. Chan, M. L. Turner, S. L. Nutt, A. Basten, and Bcl-2 increases memory B cell recruitment but does not perturb selection in R. Brink. 2006. High affinity germinal center B cells are actively selected into the germinal centers. Immunity 1: 803–813. plasma cell compartment. J. Exp. Med. 203: 2419–2424.