Ig Allotype Chimeric Mice to Intestinal Iga, Production in Gnotobiotic B1 Cells Contribute to Serum Igm, But

B1 Cells Contribute to Serum IgM, But Not to Intestinal IgA, Production in Gnotobiotic Ig Allotype Chimeric Mice

This information is current as M. Christine Thurnheer, Adrian W. Zuercher, John J. Cebra of September 24, 2021. and Nicolaas A. Bos J Immunol 2003; 170:4564-4571; ; doi: 10.4049/jimmunol.170.9.4564 http://www.jimmunol.org/content/170/9/4564 Downloaded from

References This article cites 33 articles, 19 of which you can access for free at: http://www.jimmunol.org/content/170/9/4564.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

• 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

by guest on September 24, 2021 *average

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 © 2003 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

B1 Cells Contribute to Serum IgM, But Not to Intestinal IgA, Production in Gnotobiotic Ig Allotype Chimeric Mice1

M. Christine Thurnheer,* Adrian W. Zuercher,* John J. Cebra,* and Nicolaas A. Bos2†

B1 cells are a significant source of natural serum IgM, thereby serving as a first line of defense against systemic bacterial and viral infections. They can migrate to the intestinal lamina propria and differentiate into IgA-producing plasma cells and thus might play a similar role in mucosal immunity. To investigate the contribution of B1 cells to the intestinal IgA response induced by the commensal flora in immunocompetent animals, we generated gnotobiotic and conventionally reared Ig allotype chimeric mice. In this system B1- and B2-derived Abs can be distinguished based on different allotypes. FACS analysis of peritoneal cavity cells and analysis of B1- and B2-derived serum IgM indicated stable B1/B2 chimerism and the establishment of a functional B1 population. Monoassociation with either Morganella morganii, Bacteroides distasonis, or segmented filamentous bacteria induced germinal

center reactions in Peyer’s patches and led to the production of intestinal IgA, partially reactive with bacterial Ag. A considerable Downloaded from amount of serum IgM was B1 cell derived in both monoassociated and conventionally reared mice. However, most of the total as well as bacteria-speciﬁc intestinal IgA was produced by B2 cells. These data suggest that intestinal IgA production induced by commensal bacteria is mainly performed by B2, not B1, cells. The Journal of Immunology, 2003, 170: 4564–4571.

1 cells comprise a distinct B cell population within the hypothesis that B1 cells might be a major source of natural, poly-

immune system of mice. They differ from conventional reactive, low affinity IgA in the gut. Indeed, several studies in http://www.jimmunol.org/ B B2 cells in localization, expression of phenotypic mark- either SCID mice or irradiated animals reconstituted with bone ers, Ab repertoire, and reaction to proliferative stimuli (1–3). They marrow and peritoneal cavity cells (PeC) suggested that a signif- contribute a significant amount of natural serum IgM, thereby serv- icant number of IgA-producing plasma cells in the gut originate ing as a first line of defense against systemic viral and bacterial from B1 cells (14–19). However, it is not known whether B1 cells infections (4–7). IgA is thought to play a similar role in protecting contribute a similarly large proportion of intestinal IgA in normal, mucosal surfaces from infection with potentially pathogenic mi- immunocompetent mice. croorganisms (8, 9). Upon specific interaction with APC and T To address this issue, we generated Ig allotype chimeric mice cells in Peyer’s patches (PP),3 B2 cells can relocate to the intestinal under germfree conditions, using a protocol first described by La- lamina propria, differentiate into plasma cells, and produce Ag- lor et al. (20, 21) and applied in several studies by Baumgarth et al. by guest on September 24, 2021 specific secretory IgA (10, 11). However, a large amount of intes- (4, 5) under conventional conditions. In this model B1 and B2 tinal IgA is either nonspecific or of low affinity for a broad range cell-derived Abs can be distinguished based on different allotypes. of Ag, and the source of this natural IgA as well as its functional Germfree mice were used to establish a balanced B1/B2 cell chi- significance remain to be clarified. Production of secretory IgA is merism before microbial stimulation. Systemic and local intestinal virtually absent in germfree animals, but can be induced upon bac- immune responses were studied after colonization with Mor- terial colonization (12, 13). B1 cells have the ability to switch Ig ganella morganii, Bacteroides distasonis, or segmented filamen- class and can migrate to the intestinal lamina propria and differ- tous bacterium (SFB) and compared with conventionally reared entiate into IgA-secreting plasma cells (14–16). Furthermore, it chimeric mice. We show that B1 cells contribute a large amount of has been shown that B1 cells can be activated by either LPS or the natural serum IgM under germfree conditions, which can be fur- microbial flora, and that some B1-derived IgA is reactive with ther induced by bacterial colonization. While B2 cells respond to surface Ag of intestinal bacteria (17, 18). These features led to the intestinal colonization with the production of total and specific IgA, B1 cells contribute only a minimal amount of intestinal IgA

*Department of Biology, University of Pennsylvania, Philadelphia, PA 19104; and in this model. †Department of Cell Biology, Section Histology and Immunology, Faculty of Medical Sciences, University of Groningen, Groningen, The Netherlands Materials and Methods Received for publication July 10, 2002. Accepted for publication March 5, 2003. Ig allotype chimeric mice 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 Germfree BALB/c mice were bred and housed in the germfree facility at with 18 U.S.C. Section 1734 solely to indicate this fact. University of Pennsylvania (Philadelphia, PA) under sterile conditions in 1 This work was supported by National Institutes of Health Grant AI37108 (to J.J.C.). Trexler isolators (Standard Safety, McHenry, IL). Conventionally reared M.C.T. was supported by a fellowship from the Swiss National Science Foundation. BALB/c mice were purchased from The Jackson Laboratory (Bar Harbor, b A.W.Z. is the recipient of a fellowship from the Swiss Foundation for Medical- ME), C.B-17 (C.B-Igh1 /IcrTac) mice were obtained from Taconic Farms Biological Grants. The Flow Cytometry Facility of the Cancer Center at University of (Germantown, NY) and housed in the animal facility of University of Pennsylvania is supported by The Lucille P. Markey Trust. Pennsylvania. 2 Address correspondence and reprint requests to Dr. Nicolaas A. Bos, Department of To generate germfree Ig allotype chimeric mice, newborn germfree a Cell Biology, Section of Histology and Immunology, Faculty of Medical Sciences, BALB/c mice (a allotype) were treated with 200 ␮g of anti-IgM (clone University of Groningen, Antonius Deusinglaan 1, 9713 AV Groningen, The Neth- DS-1) twice a week starting from day 1 after birth for a total of 10 injec- erlands. E-mail address: [email protected] tions in 32 days. On day 3 after birth, 2 ϫ 106 PeC isolated by peritoneal 3 Abbreviations used in this paper: PP, Peyer’s patch(es); GC, germinal center; PeC, lavage from conventionally reared, 8- to 12-wk-old C.B-17 donors (b al- peritoneal cavity cells; PNA, peanut agglutinin; SFB, segmented ﬁlamentous lotype) were transferred by i.p. injection. Conventionally reared, Ig allo- bacterium. type chimeric animals were generated without anti-IgMa treatment by

transferring 2 ϫ 106 CB.17-derived PeC on days 1, 3, 5, and 28 after birth Total IgA was measured by RIA as described previously (23), and a into BALB/c recipients. standard curve of purified, monoclonal IgA was used to convert counts per minute to nanograms per milliliter. To assess total IgAa and IgAb levels, Bacteria and monoassociation plates were coated with 50 ␮l/well of 10 ␮g/ml anti-IgAa (HY15, a gift of M. morganii, a Gram-negative aerobic rod, related to the Proteus species, Dr. M. Pawlita, originally generated and described by Potter and Lieber- ␮ was originally isolated by Potter from mouse feces and was provided by A. man (25), produced and purified in our laboratory) in PBS or 50 l/well of ␮ b Feeny (The Scripps Research Institute, La Jolla, CA). B. distasonis (TAC: 10 g/ml anti-IgA (HISM2, originally generated and produced in our ASF 519), a Gram-negative, anaerobic rod, was purchased from Taconic laboratory, currently available at BD PharMingen) in PBS, blocking and Farms. SFB, an obligate anaerobic, spore-forming, Gram-positive, seg- incubation were performed as described, and bound Ab was detected with 125I-labeled anti-IgA (Southern Biotechnology Associates). Standard mented bacterium, related to Clostridia sp., was provided by H. Snel (Uni- a b versity of Nijmegen, Nijmegen, The Netherlands). M. morganii was cul- curves of purified IgA or IgA were used to convert counts per minute to tured in vitro in Brain Heart Infusion medium (Difco, Fisher Scientific, nanograms per milliliter. Pittsburgh, PA) under aerobic conditions. B. distasonis was grown in al- Bacteria-specific IgA was measured by RIA. Plates were coated with ␮ ␮ tered Schaedler broth (Difco), supplemented with 5% sterile FCS (Life either 50 lof10 g/ml bacterial sonicate (B. distasonis, SFB) overnight ␮ ϭ Technologies, Grand Island, NY) under anaerobic conditions using the at 4°Cor100 l/well freshly cultured bacteria in PBS (OD600 2) for 24 h Aerogen system (OXOID, Basingstoke, U.K.). No in vitro culture method at 37°C(M. morganii) and were blocked with 1% BSA in PBS. Incubation for SFB has been established to date. Therefore, spore-containing intestinal with organ fragment culture supernatant fluid was performed overnight at 4°C. Thereafter, plates were incubated for4hatroom temperature with contents were isolated from SCID mice previously monoassociated with 125 a b SFB, and fecal suspensions in PBS were used for further monoassociation. I-labeled anti-IgA. To assess levels of bacteria-specific IgA or IgA , The purity of cultures or spore-containing fecal suspensions was verified plates were coated, blocked, and incubated with organ fragment culture supernatant fluid as described above. HY15 (anti-IgAa) and HISM2 (anti- microscopically in Gram-stained smears. b 125 Eight-week-old germfree Ig allotype chimeric mice were transferred IgA ) were used for allotype-specific IgA detection, and I-labeled anti- into sterile experimental Trexler isolators and monoassociated with M. IgG1 and anti-IgG2a (Southern Biotechnology Associates) were used to Downloaded from morganii, B. distasonis, or SFB by oral inoculation with 200 ␮l(5ϫ 107 detect bound HY15 and HISM2, respectively. CFU) of in vitro cultured bacteria (M. morganii, B. distasonis) or spore- containing fecal suspensions (SFB). The expansion of bacteria in the gut Results and the persistence of intestinal colonization were monitored by in vitro Generation of neonatal allotype chimeric mice under germfree culture of fecal contents (M. morganii, B. distasonis) or analysis of Gram- and conventional conditions stained swabs from the luminal side of the intestinal tract (SFB).

To study the contributions of B1 and B2 cells to the immune re- http://www.jimmunol.org/ Flow cytometry sponse against commensal bacteria, Ig allotype chimeric mice Single-cell suspensions (2 ϫ 105–106/sample) of PeC were stained for 20 were generated under germfree and conventional conditions. min at 4°C with FITC-conjugated anti-IgDb (217-170; BD PharMingen, b Treatment of newborn germfree BALB/c mice (a allotype; recip- San Diego, CA), biotinylated anti-IgM (AF6-78; BD PharMingen), FITC- a conjugated anti-IgDa (AM89.1; BD PharMingen), and biotinylated anti- ient) with anti-IgM for 32 days and transfer of PeC cells from IgMa (DS-1; BD PharMingen). Cells were washed, and biotinylated Abs C.B-17 mice (b allotype; donor) on day 3 after birth resulted in Ig were revealed by PE-conjugated streptavidin (BD PharMingen). Single- allotype chimeric animals. As shown in Fig. 1, BALB/c- and C.B- cell suspensions of PP were stained for 20 min at 4°C with FITC-conju- 17-derived PeC cells were speciﬁcally stained with Abs against gated peanut agglutinin (PNA; Pierce, Rockford, IL; coupled to FITC in a a b b ␬ surface IgM and IgD or IgM and IgD , respectively. In Ig al-

our laboratories as described previously (22)) and PE-conjugated anti- by guest on September 24, 2021 L chain (Southern Biotechnology Associates, Birmingham, AL). Cells lotype chimeric mice, most B1 cells in the PeC were of the donor were washed and fixed in 1% paraformaldehyde in PBS and analyzed on a allotype (IgMb high IgDb low), and importantly, no donor-derived FACScan flow cytometer (BD Biosciences, Sunnyvale, CA). Germinal B2 cells (IgMb low IgDb high) were detectable (Fig. 1). Most recip- low ϩ center (GC) B cells are defined as ␬ L chain /PNA cells, while memory a low a high ␬ high Ϫ ient-derived B cells were of the B2 phenotype (IgM IgD ) B cells are excluded as L chain /PNA , as described previously (12). a high a low WinMDI2.8 software was used for evaluation (The Scripps Research even though some endogenous B1 cells (IgM IgD ) were Institute). observed. Taken together these results show that most B1 cells in Ig allotype chimeric mice were donor derived, while all B2 cells Analysis of Ab production in organ fragment cultures and serum were recipient derived. The chimerism (Table I) and cell numbers Conventionally reared, Ig allotype chimeric mice were sacrificed at 10 and in the peritoneal cavity (data not shown) remained stable over the ϭ 14 wk of age (n 4/time point). Monoassociated mice were sacrificed on entire experimental period in both germfree and conventionally days 0, 7, 14, 28, 42, 56, and 70 after colonization (n ϭ 2–3/time point), and blood was collected by heart puncture for serum analysis. The entire intestinal tract was surgically removed. For organ fragment culture (23), tissues were sterilized by sequential washes as described in detail previously (24). Pieces (3 ϫ 3 mm) of duodenum, jejunum, ileum, cecum, and colon were incubated in 1 ml of Kennett’s HY medium, supplemented with 10% FBS, L-glutamine, penicillin, streptomycin, and gentamicin (all re-

agents from Life Technologies) for 7 days under a 90% O2/10% CO2 atmosphere at 37°C. Total serum IgM was measured by RIA. Flexible polyvinyl plates (Se- rocluster; Costar, Cambridge, MA) were coated with 50 ␮l/well of 20 ␮ Ј g/ml goat anti-mouse F(ab )2 (Jackson ImmunoResearch Laboratories, West Grove, PA) and blocked with PBS containing 1% BSA (Sigma- Aldrich, St. Louis, MO), and serum was incubated overnight at 4°C. Bound IgM was detected using 125I-labeled anti-IgM (Southern Biotechnology Associates). Radioactivity of individual wells was measured using a 1272 Clini␥ gamma counter (LKB-Wallac, Gaithersburg, MD). A standard curve of monoclonal IgM was used to convert counts per minute to micrograms FIGURE 1. Chimerism of B1 and B2 cells in Ig allotype chimeric mice per milliliter. compared with BALB/c (recipients) and C.B-17 (donor) mice. PeC were a b ␮ ␮ a To determine levels of IgM or IgM ,50 l/well of 10 g/ml anti-IgM isolated from BALB/c, C.B-17, or Ig allotype chimeric animals by perito- (DS-1; BD PharMingen) in PBS or 50 ␮l/well of 10 ␮g/ml anti-IgMb (AF6-78; BD PharMingen) in PBS was used for coating. Blocking and neal lavage. Single-cell suspensions were stained with fluorescently la- incubation were performed as described above, and bound Ab was detected beled, allotype-specific anti-IgM and anti-IgD Abs. Numbers are the per- high low low high with 125I-labeled anti-IgM. To convert counts per minute to nanograms per centages of B1 cells (IgM , IgD ) and B2 cells (IgM , IgD ) milliliter, standard curves were established for each allotype using purified among gated lymphocytes. One example representative of 17 experiments monoclonal IgMa or IgMb, respectively. is shown. 4566 CONTRIBUTION OF B1 CELLS TO INTESTINAL IgA PRODUCTION

Table I. B1/B2 chimerism in PeC of neonatal chimeric mice develop from bone marrow precursors, while the antigenic stimulation by gut organisms is already present during that time for Ig Allotype b Ig Allotype a donor B1 cells. To avoid the same drawback, we tried a variation (donor) (recipient) of the original Lalor method in conventional mice by continuously Intestinal Floraa n Dayb %B1c %B2d %B1c %B2d giving B2 cells the opportunity to react to the gut ﬂora by making IgA. Using multiple injections and higher doses of B1 cells, we e M. morganii 3040 01123tried to overcome the described feedback inhibition (20). In fact, 2 14 34 0 15 13 2 28 42 0 25 19 the transferred cells did establish donor B1 cells in the PeC in 3 42 38 0 17 26 conventional mice at an ϳ2:1 ratio with respect to recipient B1 3 56 42 0 10 25 cells compared with the 6:1 ratio in germfree neonatal chimeras 2 70 51 0 18 30 (Table I). B. distasonis 3 0 38 0 10 25 3 14 28 0 16 24 B1 cells contribute to the production of natural serum IgM 3 28 61 0 11 34 Germfree Ig allotype chimeric animals showed levels of serum 3 42 31 0 15 23 3 56 43 0 21 38 IgM comparable to those in conventional naive mice, and mono- 3 70 50 0 19 32 association stimulated additional IgM production (Fig. 2A). M. SFB 5 0 35 0 12 28 morganii induced a protracted IgM response with a peak on day 28 3 7 29 0 8 18 and continuously elevated levels until day 70. After monoassocia-

2 14 41 0 15 13 tion with B. distasonis serum IgM increased over the initial 14 Downloaded from 3 28 37 0 18 26 days and thereafter waned. Colonization with SFB induced max- 2 42 42 0 18 29 imal production of serum IgM between days 14–28, followed by 3 56 55 0 19 29 a return to baseline levels during the later phase of the experiment. Conventional 4 70 30 0 18 21 To assess the contributions of B1 (b allotype) and B2 (a allo- 4 102 38 1 24 12 type) cells, IgM levels were measured for each allotype. As shown a Germ-free neonatal chimeric mice were monoassociated with M. morganii, B. in Fig. 2B, monoassociation with M. morganii induced an early http://www.jimmunol.org/ distasonis, or SFB. b b Days after monoassociation or days after birth (for conventional). increase in B1 cell-derived IgM that peaked around day 28 and c Defined as IgMhigh IgDlow cells among all lymphocytes in the PeC. thereafter waned to baseline levels. In contrast, B2 cell-derived d low high Defined as IgM IgD cells among all lymphocytes in the PeC. a e Mean percentage per time point. IgM reached maximal levels around day 56 after only a slight initial increase. Colonization with B. distasonis mainly stimulated the production of IgMa by B2 cells, while B1 cell derived IgMb reared Ig allotype chimeric mice. The germfree chimeras were remained at levels similar to those found in germfree animals. SFB constructed identically to the method used by Lalor et al. (20) and induced an increase in both IgMb and IgMa with similar kinetics can be compared directly to the original Lalor data as published. for both allotypes during the first 4 wk of colonization. While The reason why we went to the technically very difficult task of production of IgMb waned after day 28, levels of IgMa remained by guest on September 24, 2021 producing mice under germfree conditions is because of the com- elevated for a longer period and reached baseline levels around day ments on the original papers by Lalor et al. that in that instance B1 70. These distinct kinetics of IgM production resulted in different cells have an advantage over B2 cells because the latter have to relative contributions of B1- and B2-derived IgM over time (Fig.

FIGURE 2. Production of serum IgM in Ig allotype chimeric mice after monoassociation with M. morganii, B. distasonis, or SFB. Ig allotype chimeric mice were sacriﬁced at the indicated time points after monoassociation and conventionally reared, Ig allotype chimeric mouse (CNV) chimeras were sacriﬁced at the age of 70 days. Blood was collected by heart puncture. A, Total serum IgM (micrograms per milliter) was measured by RIA using anti-Fab-coated plates and radiolabeled anti- IgM for detection. B, Total serum IgMa (recipient- derived) and IgMb (donor-derived; micrograms per milliter) were measured by RIA. Plates were coated with anti-IgMa or anti-IgMb, respectively, and bound IgMa or IgMb was detected by radiolabeled anti-IgM Ab. C, Relative contributions of B1 and B2 cells to production of serum IgM. CNV, Con- ventionally reared Ig allotype chimeric mice. The Journal of Immunology 4567

2C). In conventionally reared Ig allotype chimeric mice the pro- ment culture supernatant fluid were measured by RIA (Fig. 4). The portions of B1- and B2-derived IgM were similar to those found at mucosal immune system in germfree animals was quiescent, with late time points after monoassociation in gnotobiotic animals (Fig. barely detectable levels of intestinal IgA, but secretion of IgA was 2, B and C). These data show that B1 cells contribute an appre- readily induced after monoassociation with any of the three com- ciable amount to the total natural serum IgM in germfree animals mensals. Interestingly, the kinetics as well as the localization of and mount an early IgM Ab response after colonization with com- maximal IgA production varied depending on the microorganism mensal microorganisms. used for monoassociation. M. morganii stimulated IgA production mainly in the lower gastrointestinal tract (cecum, colon). Never- Monoassociation with M. morganii, B. distasonis,orSFB theless, some stimulation of IgA secretion in the small intestine induces GC reactions was observed. Colonization with B. distasonis induced an early Next we tested the stimulatory potentials of the different commen- IgA response, most pronounced in the small intestine and cecum, sal bacteria on B1 and B2 cells in the gut. All three commensals peaking on day 14 postinoculation. SFB stimulated an early IgA led to activation of the intestinal mucosal immune system with response in the upper gastrointestinal tract, with maximal levels marked GC reaction in PP, characterized by the appearance of around days 14–28. However, a slightly delayed response in ce- PNA-binding B cells (␬ L chain low PNA binding, as defined by cum and colon followed around days 28–42. These data demon- Lebman et al. (26); Fig. 3). While the GC reaction induced by B. strate that colonization with M. morganii, B. distasonis,orSFB distasonis peaked around day 28, monoassociation with M. mor- stimulated intestinal IgA production at distinct sites and different ganii led to a prolonged activation of PNA-binding B cells in PP times for each bacterium. over 70 days. Colonization with SFB induced maximal GC reac- Downloaded from tion on day 14 after monoassociation. These results show that in- Intestinal IgA secreted in response to colonization with M. oculation of germfree Ig allotype chimeric mice with each of the morganii, B. distasonis, or SFB is reactive with bacterial Ag three commensals induced activation of a previously quiescent gut mucosal immune system. Supernatant fluid of organ fragment cultures was analyzed for the presence of bacteria-specific IgA by RIA (Fig. 5). M. morganii-

Colonization with M. morganii, B. distasonis, or SFB induces specific IgA was produced in all intestinal tissues, but in accor- http://www.jimmunol.org/ the production of intestinal IgA dance with the findings for total IgA production, maximal levels of To assess mucosal Ab responses induced by intestinal colonization specific IgA were found in the colon (note the different scale). The with M. morganii, B. distasonis, or SFB, IgA levels in organ frag- production of specific IgA in the small intestine after colonization with B. distasonis reflected the data obtained for total IgA. Sur- prisingly, the highest level of B. distasonis-specific IgA was found on day 28 in the cecum. Monoassociation with SFB led to the rapid secretion of bacteria-specific IgA in the small intestine, followed by a specific response in the lower gastrointestinal tract. These data show that monoassociation with M. morganii, B. distasonis,or by guest on September 24, 2021 SFB not only stimulated the secretion of total IgA, but led to the production of Ag-specific IgA with distinct kinetics and localization of production for each bacterium.

Contributions of B1 and B2 cells to the production of total and bacteria-specific intestinal IgA Next we analyzed the contributions of B1 and B2 cells to intestinal IgA production by determining levels of total and bacteria-specific IgAa and IgAb in organ fragment culture supernatant fluids. As shown in Fig. 6A, most of the total as well as bacteria-specific IgA induced by colonization was produced by B2 cells, with only a minor contribution of B1 cells, probably due to the paucity or even the absence of donor-derived B cells in the gut lamina propria. Similarly, analyses by immunofluorescence revealed only negligi- ble numbers of IgAb-expressing cells in the gut lamina propria of chimeric mice (data not shown). Fig. 6B depicts the relative contribution of IgAa or IgAb, respectively, to total and bacteria-specific intestinal IgA. Germfree animals produced a low amount of IgA reactive with bacterial Ag compared with maximal levels of specific IgA after colonization, and B1 cells contributed signifi- cantly (45–73%) to this minute amount of naturally occurring bacteria-specific IgA (Fig. 6B). Monoassociation with M. morganii, B. distasonis, or SFB induced intestinal IgA production by B2 cells FIGURE 3. Induction of GC reactions in PP lymphocytes after mono- with very low contribution of B1 cells. Similar ratios of B1 and B2 association with M. morganii, B. distasonis, or SFB. Monoassociated Ig allotype chimeric mice were sacrificed at the indicated time points after cell-derived total IgA were found in conventionally reared ani- colonization, PP were isolated, and single-cell suspensions were stained mals. Thus, monoassociation with M. morganii, B. distasonis,or with fluorescently labeled Ab specific for ␬ L chain and fluorescently la- SFB as well as the presence of conventional microflora locally beled PNA. Numbers indicate the percentage of GC B cells (␬ L chainlow, stimulated B2 cells to produce total and bacteria-specific intestinal PNA-binding) among total B cells (␬ L chainϩ). IgA. The contribution of B1 cells to total intestinal IgA appeared 4568 CONTRIBUTION OF B1 CELLS TO INTESTINAL IgA PRODUCTION

FIGURE 4. Induction of total intestinal IgA production after monoassociation with M. morganii, B. distasonis, or SFB. Levels of total IgA (micrograms per milliliter Ϯ

SD) in supernatant ﬂuid of organ fragment Downloaded from cultures were determined by RIA using plates coated with anti-Fab Ab and radiolabeled anti IgA Ab for detection. http://www.jimmunol.org/ by guest on September 24, 2021

to be limited (1–15%); however, some bacteria-specific B1 cell- mal immunocompetent animals. Ig allotype chimeric animals, ei- derived IgA was detectable at late time points after monoassocia- ther conventionally reared or germfree and having a fully function (15–20%). tional immune system, might more accurately reflect normal physiological conditions. Discussion Cotransfer of self-renewing B1 cells and a source of B2 cells, To try to establish a mouse model that would be informative about such as bone marrow, into conventionally reared recipients might the physiologically normal contributions of cells of the B1 lineage favor the outgrowth and differentiation of B1 cells by exposing to gut IgA plasmablasts and the role(s) of commensal enteric mi- them to a stimulatory environment before the establishment of a crobes in stimulating the development of these cells, we generated mature B2 population. We therefore generated Ig allotype chimeric Ig allotype chimeras using germfree neonatal mice. Our analysis of mice neonatally under germfree conditions, allowing a potentially B1 and B2 cell-derived IgA in these monoassociated and conven- more balanced establishment of functional populations of both B1 tionally reared Ig allotype chimeric mice revealed that most of the and B2 cells. The newborns were treated repeatedly with an anti- intestinal IgA was produced by B2 cells. Other models, using ir- IgM allotype against the host allotype while being transferred with radiated, formerly immunocompetent mice and conventionally PeC cells at 3 days of age. After 8 wk we found a predominantly reared SCID mice showed that cells of the B1 lineage could con- donor B1 cell population in the PeC, presumably normal popula- tribute Ͼ50% of the IgA plasmablasts to the gut lamina propria tions of B2 cells derived from host bone marrow and a balanced (14–18, 24). Using conventionally reared TCR-␤/␦Ϫ/Ϫ mice, it has donor/host contribution to circulating IgM. been suggested that the production of most, if not all, of the natural Monoassociation of germfree Ig allotype chimeric animals with IgA in the gut was specifically driven by particular microbial Ags either M. morganii or SFB induced a rapid increase in B1-derived and that it probably was the product of B1 cells developing in a T IgM with either simultaneous or subsequent stimulation of IgM cell-independent fashion (19). While data obtained from adoptive production by B2 cells. Likewise, in our conventionally reared, transfer models clearly demonstrated that transferred B1 cells can neonatally developed, Ig allotype chimeric mice, 27% of total se- establish a self-renewing population from which IgA-secreting rum IgM was produced by B1 cells, a ratio similar to the propor- plasma cells arise (14–18), these models might only partially re- tions found at late time points after monoassociation of gnotobiotic flect the normal physiological behavior of B1 cells and might not animals. These data support the idea that B1 cells are a major be applicable for quantitative analysis of B1 cell function in nor- source of natural IgM and might function as a link between innate The Journal of Immunology 4569

FIGURE 5. Induction of bacteria- specific intestinal IgA production after monoassociation with M. morganii, B. distasonis, or SFB. Levels of bacteria- specific IgA (counts per minute Ϯ SD) in supernatant fluid of organ fragment cultures were assessed by RIA using plates coated with either bacterial soni- Downloaded from cates (B. distasonis, SFB) or whole bacteria(M.morganii)andradiolabeledanti- IgA Ab for detection. Note the different scales. http://www.jimmunol.org/ by guest on September 24, 2021

and adaptive immune systems providing early humoral defense However, monoassociation of formerly germfree, adult mice against microbial invasion (4, 5). This first line of defense can with commensal bacteria is different from the naturally occurring provide early systemic protection from potentially dangerous mi- codevelopment of the immune system and flora in conventionally croorganisms, a vital precaution considering the importance of reared animals. The induction of GC reactions upon monoassocia- central organ systems for immediate survival. Interestingly, B. dis- tion of adult germfree mice might favor stimulation and differen- tasonis stimulated mainly B2 cells, while B1 cell-derived IgM re- tiation of B2 cells and therefore lead to the predominance of B2- mained at levels comparable to those in germfree animals. These derived IgA in our system. Conventionally reared Ig allotype results suggest that B1 cells need specific stimulation to mount chimeric mice were generated by repeated injection of donor-de- additional IgM and that this stimulation depends on the colonizing rived PeC into newborn mice without additional anti-IgMa treat- microorganism. ment. This modification allowed undisturbed development of both We used unfractionated PeC cells from conventionally reared donor-derived and endogenous B cells during natural colonization. donors as a source of B1 cells. Such unfractionated populations The resulting chimerism was comparable to that of anti- also contain donor B2 cells as well as some activated T cells. Also, IgMa-treated germfree Ig allotype chimeric mice with a majority of since these cells derived from conventionally reared donors, they donor-derived B1 cells in the PeC. However, in conventionally could have been previously selected for reactivity with a vast array reared Ig allotype chimeric mice the contribution of B1 cells to the of environmental Ags different from the microbe used to monoas- intestinal IgA production was only marginal, and most gut IgA was sociate, thereby precluding an effective response against the mi- crobes selected for our study. Our arguments to minimize these produced by B2 cells. potentially confusing aspects of our model are that 1) the donor- Monoassociation with each of the three commensals also in- derived cells were of the B1 phenotype, as shown by FACS anal- duced GC reactions in PP of the small intestine and led to the ysis of PeC isolated from Ig allotype chimeric mice, and did es- production of intestinal IgA with distinct kinetics and particular tablish in PeC, but contributed little to gut IgA production; and 2) main sites of IgA secretion for each bacterium. These data are in donor-derived B1 cells were functional, i.e., they contributed to the accordance with a recent report by Jiang et al. (27) demonstrating natural serum IgM in germfree animals as well as to the IgM re- a clear correlation between the localization of SFB in the intestine sponse after monoassociation. Thus, a rapid decay of transferred and the IgA response by the host. Taken together, our results show B1 cells with the specificities to respond to any particular microbe that all three microorganisms interact with their host in a way that seems improbable, particularly since their product is often rather leads to activation of the formerly quiescent gut mucosal immune close to a germline specificity. system. 4570 CONTRIBUTION OF B1 CELLS TO INTESTINAL IgA PRODUCTION

FIGURE 6. Intestinal IgA in gnotobiotic or conventionally reared (CNV) Ig allotype chimeric mice is predominantly produced by B2, not B1, cells. Levels of B1 and B2 cell-derived intestinal IgA were measured by RIA. A,Toas- sess total IgAa (donor-derived) and IgAb (recipient-derived), plates were coated with anti-IgAa (HY15) or anti-IgAb (HISM2); pooled organ fragment culture supernatant fluid from duodenum, jejunum, ileum, cecum, and colon was incubated in serial dilutions; and bound IgAa or IgAb (micrograms per milliliter) was detected with radiolabeled anti-IgA Ab. To determine the amount of bacteria-specific IgAa or IgAb, plates Downloaded from were coated with either bacterial sonicates (B. distasonis, SFB) or whole microorganisms (M. morganii); pooled organ fragment culture supernatant fluid from duodenum, jejunum, ileum, cecum, and colon was incubated; and bound IgAa or IgAb (counts per minute) was detected with

HY15 or HISM2, respectively. B, Relative con- http://www.jimmunol.org/ tributions of B1 and B2 cells to total and speciﬁc intestinal IgA production. by guest on September 24, 2021

Until recently the mechanisms involved in recruiting IgA-com- How can we reconcile our observations with respect to the find- mitted B cells to the intestinal lamina propria have been only ings in other systems? The potential of B1 cells to produce intes- ␣ ␤ poorly understood. It has been established that 4 7 integrin me- tinal IgA has been shown in transfer models (14–18). However, diates binding of lymphocytes to mucosal addressin cell adhesion the immunologic environment in these settings is greatly altered. molecule-1 and is therefore crucial for lymphocyte homing to the The function of B2 cells is impaired either by delayed develop- gut (28). Work by Bowman et al. (29) has shown that TECK/ ment of a functional B2 population (transfer models) or by atten- CCL25, expressed by intestinal epithelial cells, is a potent and uated GC reaction in PP and disturbed cognate B-T cell interaction selective chemoattractant for IgA-secreting B cells. Their work (TCR-␤/␦Ϫ/Ϫ mice). The SCID mouse lacks PP GC reactions, and focused on B cell populations obtained from mesenteric lymph the TCR-␤/␦Ϫ/Ϫ mouse has only minimal gut PP GC reactions, nodes, PP, and spleen. As B2 cells represent the major B cell probably due to B2 cells, and they do not lead to appreciable af- population in these organs, it might well be that the mechanisms finity maturation (point mutations) (34). Such B2 responses may described by these authors, while true for the recruitment of B2 account for the observed microbial Ag dependence of natural IgA cells, only marginally apply for B1 cells. Several mechanisms in- Ϫ Ϫ production in TCR-␤/␦ / mice (19). Possibly the cytokines volved in the accumulation of B1 cells in the peritoneal cavity, Ig present in these vestigial GC reactions are sufficient to allow min- class switching, and migration to the lamina propria have been imal clonal expansion and switching to IgA expression. Our model described (30–32), and Fagarasan et al. (33) have demonstrated ϩ ϩ system includes normal GC reactions in PP, favoring specificB2 the potential of B220 IgM lamina propria cells, presumably B1 cells, for in situ class switching and differentiation to IgA-producing cell development. Thus, we believe our proffered model is likely to cells. However, they also showed that the mechanism described be more physiologically normal and relevant to the immunocom- applies for a very small fraction of lamina propria B cells compared petent mouse than other models. with PP B cells. Taken together, these studies suggest that different In summation, our data demonstrate that B1 cells are a major mechanisms are involved in the attraction of B1 and B2 cells to the source of natural IgM and can mount an early serum IgM response lamina propria, and that the quantitative importance of these mecha- given the right antigenic stimuli, but that Ig class switching and nisms might vary. We cannot completely exclude that some of the migration to the gut after colonization with commensal bacteria are intestinal IgAa is B1 derived in our model, as the Ig allotype chimeric dominated by B2 cells and not B1 cells. Our model suggests that mice showed the presence of some endogenous B1 cells. However, if the physiologic contribution of B1 cells to IgA production in re- B1 cells contribute a major proportion of the intestinal IgA production sponse to colonization with commensal organisms is limited, and in our model, we would expect the donor-derived B1 cells to most of the total as well as specific intestinal IgA is produced by participate in a quantitatively significant way. B2 cells. The Journal of Immunology 4571

Acknowledgments 17. Bos, N. A., J. C. A. M. Bun, S. H. Popma, E. R. Cebra, G. J. Deenen, M. J. F. van der Cammen, F. G. M. Kroese, and J. J. Cebra. 1996. Monoclonal We thank Alec McKay for running the FACS and preparing radiolabeled Immunoglobulin A derived from peritoneal B cells is encoded by both germ line Abs, and Michelle Albright for maintaining the germfree facility. We also and somatically hypermutated VH genes and is reactive with commensal bacteria. thank Dr. Han-Qing Jiang for technical and intellectual support. Infect. Immun. 64:616. 18. Kroese, F. G. M., R. de Waard, and N. A. Bos. 1996. B-1 cells and their reactivity with the murine intestinal microflora. Semin. Immunol. 8:11. References 19. Macpherson, A. J., D. Gatto, E. Sainsburry, G. R. Harriman, H. Hengartner, and 1. Su, I., and A. Tarakhowsky. 2000. B-1 cells orthodox or conformist. Curr. Opin. R. M. Zinkernagel. 2000. A primitive T-cell independent mechanism of intestinal Immunol. 12:191. mucosal IgA responses to commensal bacteria. Science 288:2222. 2. Hayakawa, K., and R. R. Hardy. 2000. Development and function of B-1 cells. 20. Lalor, P. A., A. M. Stall, S. Adams, and L. A. Herzenberg. 1989. Permanent Curr. Opin. Immunol. 12:346. alteration of the murine Ly-1 B repertoire due to selective depletion of Ly-1 B 3. Martin, F., and J. F. Kearny. 2001. B1 cells: similarities and differences with cells in neonatal animals. Eur. J. Immunol. 19:501. other B cell subsets. Curr. Opin. Immunol. 13:195. 21. Lalor, P. A., L. A. Herzenberg, S. Adams, and A. M. Stall. 1989. Feedback 4. Baumgarth, N., O. C. Herman, G. C. Jaeger, L. Brown, L. A. Herzenberg, and regulation of murine Ly-1 B cell development. Eur. J. Immunol. 19:507. L. A. Herzenberg. 1999. Innate and acquired humoral immunities to influenza 22. Butcher, E. C., R. V. Rouse, R. L. Coffman, C. N. Nottenburg, R. R. Hardy, and virus are mediated by distinct arms of the immune system. Proc. Natl. Acad. Sci. I. L. Weisman. 1982. Surface phenotype of Peyer’s patch germinal center cells: USA 96:2250. implication for the role of germinal center B cell differentiation. J. Immunol. 5. Baumgarth, N., O. C. Herman, G. C. Jaeger, L. E. Brown, L. A. Herzenberg, and 129:2698. J. Chen. 2000. B-1 and B-2 cell-derived immunoglobulin M antibodies are non- 23. Logan, A. C., K. P. Chow, A. George, P. D. Weinstein, and J. J. Cebra. 1991. Use redundant components of the protective response to influenza virus. J. Exp. Med. of Peyer’s patch and lymph node fragment cultures to compare local immune 192:271. responses to Morganella morganii. Infect. Immun. 59:1024. 6. Boes, M., A. P. Prodeus, T. Schmidt, M. C. Carroll, and J. Chen. 1998. A critical 24. Kushnir, N., N. A. Bos, A. W. Zuercher, S. E. Coffin, C. A. Moser, P. A. Offit, role of natural immunoglobulin M in immediate defense against systemic infec- and J. J. Cebra. 2001. B2 but not B1 cells contribute to CD4ϩ T cell-mediated tion. J. Exp. Med. 188:2381. clearance of rotavirus in SCID mice. J. Virol. 75:5482. Downloaded from 7. Ochsenbein, A. F., T. Fehr, C. Lutz, M. Suter, F. Brombacher, H. Hengartner, and 25. Potter, M., and R. Lieberman. 1967. Genetics of immunoglobulins in the mouse. R. M. Zinkernagel. 1999. Control of early viral and bacterial distribution and Adv. Immunol. 7:91. disease by natural antibodies. Science 286:2156. 26. Lebman, D. A., P. A. Griffin, and J. J. Cebra. 1987. Relationship between ex- 8. Stokes, C. R., J. F. Soothill, and M. W. Turner. 1975. Immune exclusion is a pression of IgA by Peyer’s patch cells and functional IgA memory cells. J. Exp. function of IgA. Nature 255:745. Med. 166:1405. 9. Winner, L., III, J. Mack, R. Weltzin, J. J. Mekalanos, J. P. Kraehenbuhl, and 27. Jiang, H.-Q., N. A. Bos, and J. J. Cebra. 2001. Timing, localization and persis- M. R. Neutra. 1991. New model for analysis of mucosal immunity: intestinal tence of colonization by segmented filamentous bacteria in the neonatal mouse secretion of specific monoclonal immunoglobulin A from hybridoma tumors pro- gut depend on immune status of mothers and pups. Infect. Immun. 69:3611. tects against vibrio cholerae infection. Infect. Immun. 59:977. http://www.jimmunol.org/ 28. Berlin, C., E. L. Berg, M. J. Briskin, D. P. Andrews, P. J. Kilshaw, B. Holzman, 10. McIntyre, T. M., and W. Strober. 1999. Regulation of IgA B-cell development. I. L. Weissmann, A. Haman, and E. C. Butcher. 1993. ␣ ␤ integrin mediates In Mucosal Immunology, 2nd Ed. P. L. Ogra, J. Mestecky, M. E. Lamm, 4 7 lymphocyte binding to mucosal vascular addressin MAdCAM-1. Cell 74:185. W. Strober, J. Bienenstock, and J. R. McGee, eds. Academic Press, San Diego, p. 319. 29. Bowman, E. P., N. A. Kukin, K. R. Youngman, N. H. Lazarus, E. J. Kunkel, 11. Craig, S. W., and J. J. Cebra. 1971. Peyer’s patches: an enriched source for J. Pan, H. B. Greenberg, and E. C. Butcher. 2002. The intestinal chemokine IgA-producing immunocytes in the rabbit. J. Exp. Med. 134:188. thymus expressed chemokine (CCL25) attracts antibody secreting cells. J. Exp. 12. Schroff, K. E., K. Meslin, and J. J. Cebra. 1995. Commensal enteric bacteria Med. 195:269. engender a self limiting humoral mucosal immune response while permanently 30. Ansel, K. M., R. B. S. Harris, and J. G. Cyster. 2002. CXCL13 is required for B1 colonizing the gut. Infect. Immun. 63:3904. cell homing, natural antibody production, and body cavity immunity. Immunity 13. Talham, G. L., H.-Q. Jiang, N. A. Bos, and J. J. Cebra. 1999. Segmented fila- 16:67. mentous bacteria are potent stimuli of a physiologically normal state of the mu- 31. Hiroi, T., M. Yanagita, N. Ohta, G. Sakaue, and H. Kiyono. 2000. IL-15 and rine gut mucosal immune system. Infect. Immun. 67:1992. IL-15 receptor selectively regulate differentiation of common mucosal immune by guest on September 24, 2021 14. Kroese, F. G. M., E. C. Butcher, A. M. Stall, P. A. Lalor, S. Adams, and system independent B-1 cells for IgA responses. J. Immunol. 165:4329. L. A. Herzenberg. 1989. Many of the IgA producing plasma cells in the gut are 32. Hiroi, T., M. Yanagita, H. Iijima, K. Iwatani, T. Yoshida, K. Takatsu, and derived from self-replenishing precursors in the peritoneal cavity. Int. Immunol. H. Kiyono. 1999. Deficiency of IL-5 receptor ␣-chain selectively influences the 1:75. development of the common mucosal immune system independent IgA-produc- 15. Kroese, F. G. M., W. A. M. Ammerlaan, and A. B. Kantor. 1993. Evidence that ing B-1 cells in mucosa-associated tissues. J. Immunol. 162:821. intestinal IgA plasma cells in ␮, ␬ transgenic mice are derived from B-1 (Ly-1B) 33. Fagarasan, S., K. Kinoshita, M. Muramatsu, K. Ikuta, and T. Honjo. 2001. In situ cells. Int. Immunol. 5:1317. class switching and differentiation to IgA-producing cells in the gut lamina pro- 16. Kroese, F. G. M., J. J. Cebra, M. J. F. van der Cammen, A. B. Kantor, and pria. Nature 413:639. N. A. Bos. 1995. Contribution of B-1 cells to intestinal IgA production in the 34. Lentz, V. M., and T. Manser. 2001. Cutting edge: germinal centers can be in- mouse. Methods 8:37. duced in the absence of T cells. J. Immunol. 167:15.