Isolated Lymphoid Follicles Lymphocytes Drive the Formation Of

Home , Lymphotoxin

Adaptive Immune Responses Are Dispensable for Isolated Lymphoid Follicle Formation: Antigen-Naive, Lymphotoxin-Sufficient B Lymphocytes Drive the Formation of Mature This information is current as Isolated Lymphoid Follicles of September 24, 2021. Keely G. McDonald, Jacquelyn S. McDonough and Rodney D. Newberry J Immunol 2005; 174:5720-5728; ; doi: 10.4049/jimmunol.174.9.5720 Downloaded from http://www.jimmunol.org/content/174/9/5720

References This article cites 46 articles, 30 of which you can access for free at: http://www.jimmunol.org/ http://www.jimmunol.org/content/174/9/5720.full#ref-list-1

Why The JI? Submit online.

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

• No Triage! Every submission reviewed by practicing scientists by guest on September 24, 2021

• Fast Publication! 4 weeks from acceptance to publication

*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 © 2005 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Adaptive Immune Responses Are Dispensable for Isolated Lymphoid Follicle Formation: Antigen-Naive, Lymphotoxin-Sufﬁcient B Lymphocytes Drive the Formation of Mature Isolated Lymphoid Follicles1

Keely G. McDonald, Jacquelyn S. McDonough, and Rodney D. Newberry2

Isolated lymphoid follicles (ILFs) are recently appreciated members of the mucosal immune system. The architecture, composition, and inducible nature of these structures indicates that these structures are tertiary lymphoid structures. The process leading to the formation of tertiary lymphoid structures, lymphoid neogenesis, has been observed in a number of inflammatory and autoimmune conditions. Given this association, there is considerable interest in identifying the factors promoting lymphoid Downloaded from neogenesis, and understanding the steps in this process. Using murine ILF formation as a model, we have examined the roles of different cellular sources of lymphotoxin (LT) and the adaptive immune response in lymphoid neogenesis. In this study, we report that, although other cellular sources of LT may supplant B lymphocytes in the formation of immature ILFs (loosely organized clusters of B lymphocytes), LT-sufficient B lymphocytes are required for the progression of immature ILFs to mature ILFs (organized lymphoid aggregates with a follicle-associated epithelium). ILF formation occurs in the absence of T lymphocytes and Ag-specific B lymphocyte responses, and ILF B lymphocytes express elevated levels of LT in the absence of antigenic stimulation. http://www.jimmunol.org/ Consistent with a role for chemokines inducing LT expression in Ag-naive B lymphocytes, and a chemokine-driven positive- feedback loop driving mature ILF formation, mature ILFs express elevated levels of B lymphocyte chemoattractant in the absence of Ag-specific B lymphocyte stimulation. These observations indicate that ILFs contain Ag-naive lymphocytes, and suggest that events occurring within ILFs shape subsequent immune responses mediated by these lymphocytes. The Journal of Immunology, 2005, 174: 5720–5728.

he mucosal immune system is a complex network of lym- The initial identiﬁcation of ILFs used B220ϩ staining of frozen phoid compartments working together to protect higher sections of murine intestine cut at an axis perpendicular to the villi,

organisms from invading pathogens. The presence of ag- and did not allow assessment of the macroscopic architecture of by guest on September 24, 2021 T 3 ϩ gregates of mononuclear cells resembling Peyer’s patches (PP) or the B220 structures (4). Using whole-mount techniques and dis- lymph nodes in the human intestine is well documented (1–3); secting microscopy, we have identified a variety of structures con- however, the relationship of these structures to PP, the nature of taining B220ϩ cells, including loosely organized clusters of their formation, and their function has remained unclear. Analo- B220ϩ cells preferentially positioned at the base of villi, or im- gous structures, isolated lymphoid follicles (ILFs), have been iden- mature ILFs (iILFs), and well-organized lymphoid structures con- tified in the murine small intestine (4). ILFs were found to be taining a germinal center and an overlying FAE, or mature ILFs ϩ composed of predominantly B-2 B lymphocytes and CD4 T lym- (mILFs). The relationship of iILFs progressing to mILFs is sup- phocytes. And, like PP, ILF were found to possess a follicle-as- ported by several findings, which include the following: the iden- sociated epithelium (FAE) containing M cells (4). Recently, ILFs tical distribution and location of iILFs and mILFs; the require- were found to be sites for IgA class switching (5). ments for formation of iILFs are shared by mILFs, whereas not all of the requirements for formation of mILFs are shared by iILFs (indicating that iILFs precede mILFs); and the ability to promote Department of Internal Medicine, Washington University School of Medicine, St. Louis, MO 63110 the progression of iILFs to mILFs (6). Received for publication August 16, 2004. Accepted for publication February Several factors required for ILF formation have been identified. 24, 2005. These requirements parallel those required for PP and other sec- The costs of publication of this article were defrayed in part by the payment of page ondary lymphoid structure formation, but have key distinctions. charges. This article must therefore be hereby marked advertisement in accordance Like PP formation, ILF formation was found to be dependent on with 18 U.S.C. Section 1734 solely to indicate this fact. lymphotoxin (LT), LT␤R, and NF-␬␤-inducing kinase function 1 This work was supported in part by National Institutes of Health Grants DK064798 and ␤ DK060648, the Broad Medical Research Program IBD-0042, and Washington University because ILFs were absent in LT-deficient mice, LT R-deficient School of Medicine Digestive Diseases Research Core Center Grant P30-DK52574. mice, and aly/aly mice (4, 6). Although both ILF and PP formation 2 Address correspondence and reprint requests to Dr. Rodney D. Newberry, 660 South were found to be dependent on LT␤R-sufficient stromal cells, ILF Euclid Avenue, Box 8124, St. Louis, MO 63110. E-mail address: rnewberry@ and PP formation were found to differ in the cellular source of LT im.wustl.edu required for their formation. PP formation requires a LT-sufficient 3 Abbreviations used in this paper: PP, Peyer’s patch; FAE, follicle-associated epi- CD4ϩ, CD3Ϫ, bone marrow-derived cell, which may be a precur- thelium; ILF, isolated lymphoid follicle; iILF, immature ILF; mILF, mature ILF; LT, lymphotoxin; PNA, peanut lectin (agglutinin); HEL, hen egg white lysozyme; MH- sor to NK cells (7), whereas mILF formation was found to be CII, MHC class II; BLC, B lymphocyte chemoattractant; SLC, secondary lymphoid dependent on LT-sufficient B lymphocytes (6). tissue chemokine; ELC, EBV-induced molecule 1 ligand chemokine; SDF-1, stromal cell-derived factor-1; LTi, lymphoid tissue inducer; AID, activation-induced cytidine The requirement for LT-sufficient B lymphocytes is not only a deaminase. distinction between secondary lymphoid structure formation and

ILF formation, but it also provides an opportunity for critical in- Whole mounts of small intestine sight into the process of ILF formation. LT expression is induced Small intestines were removed intact, flushed with cold PBS, and opened in B lymphocytes following activation, which occurs in the context along the mesenteric border. Intestines were mounted, lumen facing up and of BCR ligation by Ag. Alternatively, LT expression may be in- fixed with cold 10% phosphate-buffered formyl saline (Fisher Scientific) duced in Ag-naive B lymphocytes following chemokine receptor for1hat4°C. Intestines were washed three times in cold PBS, incubated ligation (8). Understanding the requirement for Ag exposure in ILF in a solution of 20 mM DTT, 150 mM Tris, and 20% ethanol at room temperature for 45 min, washed three times in cold PBS, and incubated in formation not only yields insight into the mechanisms leading to a solution of 1% H2O2 for 15 min at room temperature to block endogenous ILF formation, it also yields insight into the roles ILFs play in peroxidases. Intestines were washed three times in PBS, followed by in- immune responses. The requirement for previous Ag exposure cubation in PBS containing 1% BSA and 0.3% Triton X-100 for 30 min. would suggest that ILFs are less likely to be sites where immune Intestines were incubated with HRP-conjugated lectin from Ulex euro- paeus (UEA-I) (Sigma-Aldrich) in PBS, BSA, Triton X-100 solution over- responses to new Ags are shaped, and are more likely sites where night at 4°C to facilitate the identification of PP and mILF. The following Ag-experienced lymphocytes receive ongoing stimulation to per- day, intestines were washed three times in PBS, incubated in DAB metal petuate immune responses. The absence of a requirement for Ag peroxide substrate (Pierce) for 15 min, rinsed twice in distilled water, and exposure in ILF formation would suggest that ILFs contain Ag- returned to PBS for further analysis. Investigators unaware of the treatment naive lymphocytes, and immune responses to new Ags are shaped groups determined the presence of mILFs. Under low-power microscopy (ϫ25–65) the following criteria were used to determine the presence of within ILFs. mILF: 1) presence of a nodular structure with size equal to or greater than In this study, we demonstrate that, in the absence of LT-suffi- the width of one villus, 2) nodular structure possessing an overlying dome cient B lymphocytes, ILF formation is arrested at an immature resembling the FAE of PP, and 3) nodular structures occurring singly or in stage, and ILF formation occurs in the absence of T lymphocytes groups of two (three or more nodules of approximately the same size were Downloaded from considered to be PP). and in the absence of prior Ag exposure by B lymphocytes. We For B220ϩ staining of whole mounts (modified whole mounts) to de- also observed that ILF B lymphocytes have a naive phenotype and termine the numbers of iILFs, intestines were removed intact, flushed with express LT in the absence of prior Ag exposure. Therefore, PBS, opened along the mesenteric border, and mounted as above. Intestines whereas the formation of mILFs is dependent on LT-sufficient B were then incubated three times in HBSS (BioWhittaker) containing 5 mM lymphocytes, the formation of ILFs is not dependent on prior Ag- EDTA at 37°C with shaking for 10 min to remove epithelial cells. Intes- tines were then fixed in 10% phosphate-buffered formyl saline and treated specific stimulation of these B lymphocytes. These findings indi- http://www.jimmunol.org/ with 1% H2O2 for 15 min at room temperature as above. Intestines were cate that ILF formation is driven by innate immune responses, and incubated in a solution of 50 mM Tris (pH 7.2), 150 mM NaCl, 0.6% suggest that a chemokine-driven positive-feedback loop drives the Triton X-100, and 0.1% BSA for1hat4°Ctoblock nonspecific Ab progression of iILFs to mILFs. These observations suggest that binding and then incubated with rat anti-mouse B220 Ab (BD Pharmingen) diluted in the above solution overnight at 4°C. Intestines were washed three ILFs are sites where Ag-naive lymphocytes have initial encounters times in the above solution and incubated with a HRP-conjugated goat with Ags, and events occurring within ILFs may significantly anti-rat IgG Ab (Jackson ImmunoResearch Laboratories) diluted in the shape the mucosal immune response. above solution at room temperature for 1 h. Intestines were washed three times and incubated in DAB metal peroxide substrate as above. Investi- gators unaware of the treatment groups determined the presence of iILFs. Materials and Methods

The criterion used to determine the presence of iILFs were clusters of by guest on September 24, 2021 Mice B220ϩ cells not having a nodular appearance, and occurring preferentially at the base of villi (6). Mice used for this study were housed in a specific pathogen-free facility and fed chow diet not containing hen egg white lysozyme (HEL). Animal Measurement of fecal IgA procedures and protocols were conducted in accordance with the institu- tional review board at Washington University School of Medicine. Mice used for the measurement of fecal IgA were 6–7 wk of age. Feces C57BL/6, RAG-1-deficient, B cell-deficient JHϪ/Ϫ mice (9), and MD4 were collected from individual mice, diluted 1/10 wet weight to volume mice (10) on the C57BL/6 background were purchased from The Jackson with PBS, vortexed into a uniform suspension, and centrifuged at 12,000 Laboratory. MD4 mice possess a transgenic rearranged Ig L chain and Ig rpm for 10 min in a table top microcentrifuge, and supernatants were re- H chain linked to ␮ and ␦ C region genes of the IgHa allotype, which is moved. Fecal supernatants or IgA standards (Southern Biotechnology As- specific for HEL (10). MD4 mice were bred with JHϪ/Ϫ mice to obtain the sociates) diluted in PBS containing 0.05% Tween 20 (Sigma-Aldrich) were MD4JHϪ/Ϫ mice, possessing exclusively the transgenic BCR on all B lym- incubated in 96-well Immulon 4 plates (Fisher Scientific) previously coated phocytes. LT␣-deficient mice (11) were bred onto the C57BL/6 back- with goat anti-mouse Ab (Southern Biotechnology Associates) and blocked ground for Ͼ10 generations before use in experiments. Timed pregnant with PBS containing 5% BSA and 0.05% Tween 20 at room temperature C57BL/6 female mice for use in experiments involving the injection of for 2 h. Plates were washed three times with PBS containing 0.05% Tween LT␤R-Ig fusion protein, were generated by matings with C57BL/6 male 20, and then goat anti-mouse IgA alkaline phosphatase-conjugated Ab mice. Six- to 10-wk-old LT␣Ϫ/Ϫ mice were used as recipients for bone (Southern Biotechnology Associates) diluted in PBS containing 5% BSA marrow transfers. and 0.05% Tween 20 was added to the plate and incubated for2hatroom temperature. Plates were washed three times with PBS containing 0.05% Bone marrow transfers Tween 20, and p-nitrophenyl phosphate alkaline phosphatase substrate (Sigma-Aldrich) was added. Plates were read at 405 nm using BioTek Bone marrow transfers were performed after lethal irradiation as previ- Instruments Microplate Reader. Each sample was measured in duplicate in ously described (12). A total of 1 ϫ 107 T lymphocyte-depleted bone at least three dilutions. Data are reported as the mean concentration of IgA marrow cells from gender-matched donors were injected i.v. into recipients in the fecal supernatant prepared as above. on the second day of irradiation. Mice receiving bone marrow from multiple donors (LT␣Ϫ/Ϫ and RagϪ/Ϫ,LT␣Ϫ/Ϫ, and JHϪ/Ϫ, or C57BL/6 and ELISPOT assay JHϪ/Ϫ), received 5 ϫ 106 cells from each donor. Mice were allowed 12 wk for reconstitution with donor bone marrow before use in experiments. Lamina propria cellular populations were isolated as previously described Flow-cytometric analysis was performed on splenocytes from recipients at (12). Ninety-six-well multiscreen-HA plates (Millipore) were coated with the time of sacrifice to document appropriate lymphocyte reconstitution. goat anti-mouse Ig (Southern Biotechnology Associates) overnight at room temperature. Plates were washed three times in PBS, blocked with PBS LT␤R-Ig treatment containing 5% newborn calf serum (HyClone) for1hat37°C, washed, and lamina propria cellular suspensions in IMDM (BioWhittaker), 5% FCS LT␤R-Ig production and treatment were performed as previously described (HyClone), 2 mM Glutamax I (Invitrogen Life Technologies), 50 U/ml (12). Timed pregnant female C57BL/6 mice were injected with 100 ␮gof penicillin-50 mg/ml streptomycin (Invitrogen Life Technologies), and 50 mM LT␤R-Ig via tail vein on day 16 postconception. Mice receiving LT␤R-Ig 2-ME (Fisher Scientific) were added to the plates. Plates were incubated at in utero were analyzed for the presence of mILF at 6–7 wk of age (see 37°C, 10% CO2 overnight, washed with PBS containing 0.05% Tween 20, and Fig. 2). incubated with alkaline phosphatase-conjugated goat anti-mouse IgA Ab 5722 INNATE IMMUNE RESPONSES DRIVE ILF FORMATION

2(Southern Biotechnology Associates) overnight at 4°C. Plates were CGCGGCT-3Ј; BLC, 5Ј-CAGAATGAGGCTCAGCACAGC-3Ј and 5Ј- washed with PBS and exposed to 5-bromo-4-chloro-3-indolyl phosphatase/ CAGAATACCGTGGCCTGGAG-3Ј; SLC, 5Ј-TCCCGGCAATCCTGT NBT substrate (Sigma-Aldrich), and spot-forming cells were counted un- TCTC-3Ј and 5Ј-CCTTCCTCAGGGTTTGCACA-3Ј; ELC, 5Ј-ATG der a dissecting microscope. CGGAAGACTGCTGCCT-3Ј and 5Ј-GGCTTTCACGATGTTCCCAG-3Ј; SDF-1, 5Ј-TGCTCTCTGCTTGCCTCCA-3Ј and 5Ј-GGTCCGTCAGGCTA Immunohistochemisty CAGAGGT-3Ј;LT␣,5Ј-GAGAGGGTCTACGTTAACATCAGTCA-3Ј and 5Ј-TTGCTCAAAGAGAAGCCATGTC-3Ј; and LT␤,5Ј-CCCCAGCAAG Paraffin-embedded sections containing mILF from whole-mount intestines CAGAACTCA-3Ј and 5Ј-CGCCCCGAAGAAGGTCTT-3Ј. Expression of each (performed as described above) were deparaffinized by sequential treat- target was measured in triplicate. Relative quantitation of target expression was ments with Citrosolv (Fisher Scientific) and isopropyl alcohol, rinsed with determined using the comparative CT method with 18s expression as a control, as tap water, rehydrated in PBS, treated with avidin/biotin blocking kit (Vec- described in the ABI Prism 7700 sequence detection system user bulletin. tor Laboratories), washed three times in PBS, and blocked for 15 min at room temperature in PBS containing 1% BSA, and 0.1% Triton X-100. Statistical analysis Sections were then incubated with biotin-conjugated lectin from Arachis hypogaea (peanut lectin (agglutinin) (PNA)) (Sigma-Aldrich), and diluted Data analysis using one-way ANOVA followed by Tukey’s multiple com- in PBS containing 1% BSA, and 0.1% Triton X-100 overnight at 4°C. For parison posttest was performed using GraphPad Prism (GraphPad Soft- detection of PNA staining, we used biotinyl-tyramide signal amplification ware). A value of p Ͻ 0.05 was used as a cutoff for statistical significance. (DuPont/NEN) followed by incubation with streptavidin-conjugated cya- nine 2 dye (Jackson ImmunoResearch); sections were counterstained with Results Hoechst dye (Sigma-Aldrich) to visualize nuclei. LT-sufficient B lymphocytes are not required for the formation Cell isolation from mILFs, PP, and spleen of iILFs, but are required for the progression of iILFs to mILFs

Spleens were removed and disrupted by mechanical disassociation. Intes- Previous observations have demonstrated that LT and the LT␤R Downloaded from tines were flushed with cold PBS, opened along the mesenteric border, and play a crucial role in the formation of ILFs, and the requirement for mounted with the lumen facing up in cold PBS, as described above. PP LT and the LT␤R occurs early in this process, because both mILFs were identified, cut out, and disrupted using mechanical disassociation. ␤ Using the dissecting microscope, and a blunt-end 26-gauge needle and and iILFs are absent in LT-deficient and LT R-deficient mice (4, syringe, the contents of multiple mILFs were aspirated and placed in cold 6). We have previously demonstrated that, in the absence of LT- PBS. RBC were lysed from each cellular suspension and then used for sufficient B lymphocytes, mILFs do not form (6). However, it is flow-cytometric analysis as described below. Average yield of viable unclear whether LT-sufficient B lymphocytes are required through- http://www.jimmunol.org/ mononuclear ILFs cells ranged from 3 to 7 ϫ 105 cells/intestine. ϩ out the process of ILF formation, or whether other LT-sufficient ILF cell populations were enriched for CD19 cells using anti-CD19 microbeads and magnetic separation columns (Miltenyi Biotec) per the cells may suffice to drive ILF formation at earlier stages. To ad- Ϫ Ϫ manufacturer’s recommendations. ILF cellular populations enriched in this dress this question, we reconstituted lethally irradiated LT␣ / ϩ fashion contained Ͼ90% B220 cells; populations depleted in this fashion mice with bone marrow from LT␣Ϫ/Ϫ and RAGϪ/Ϫ mice, LT␣Ϫ/Ϫ Ͻ ϩ Ϫ Ϫ Ϫ Ϫ Ϫ Ϫ contained 20% B220 cells by flow-cytometric analysis. and JH / mice, C56BL/6 and JH / mice, and LT␣ / mice, Flow-cytometric analysis thus reconstituting the LT/LT␤R axis with different cellular sources of LT. The recipient’s intestines were examined for the Single-cell suspensions were obtained as above, and flow-cytometric analysis was performed as previously described (12). Reagents used for analysis were presence of iILFs and mILFs using whole-mount techniques. We Ϫ/Ϫ Ϫ/Ϫ by guest on September 24, 2021 FITC-conjugated rat anti-mouse CD19, PE-conjugated rat anti-mouse CD19, observed that reconstitution of LT␣ mice with LT␣ bone FITC-conjugated hamster anti-mouse CD69, hamster anti-mouse LT␤, bi- marrow did not induce the formation of iILFs or mILFs. Recon- otin-conjugated mouse anti-hamster IgG mixture, streptavidin-PE, strepta- stitution of LT␣Ϫ/Ϫ recipients with a combination of bone marrow vidin-FITC, appropriate isotype control Abs (all from BD Biosciences), from LT␣Ϫ/Ϫ mice and RAGϪ/Ϫ mice, such that the recipients FITC-conjugated rat anti-mouse MHC class II (MHCII) (Southern Bio- technology Associates), biotin-conjugated rat anti-mouse CD86, and FITC- would have LT-sufficient NK cells and LT-deficient T and B lym- conjugated rat anti-mouse CD23 (eBioscience). HEL (catalog no. L-6876; phocytes, induced the formation of iILFs but not mILFs. Recon- Sigma-Aldrich) was conjugated to Alexa Fluor 488 (Molecular Probes) per stitution of LT␣Ϫ/Ϫ recipients with a combination of bone marrow the manufacturer’s recommendations and incubated with isolated cell pop- from LT␣Ϫ/Ϫ mice and JHϪ/Ϫ mice, such that recipients would ulations at a concentration of 500 ng/ml for the determination of Ag binding. Dead cells were excluded based on forward- and side-light scatter. have LT-sufficient NK cells and T lymphocytes and LT-deficient B Gates for positive staining were defined such that 1% of the analyzed lymphocytes, induced the formation of more iILFs but not mILFs. Ϫ Ϫ population stained positive with the appropriate isotype control Ab. In comparison, reconstitution of LT␣ / recipients with a combi- To document modulation of the expression of cell surface markers as- nation of C57BL/6 and JHϪ/Ϫ bone marrow, such that the recip- Ϫ/Ϫ sociated with Ag activation, splenocytes from MD4JH mice were cul- ients now have LT-sufficient B lymphocytes, induced the forma- tured at a density of 1 ϫ 106 cells/ml with hen egg lysozyme (1 ␮g/ml) in medium consisting of RPMI 1640 (BioWhittaker), 2 mM L-glutamine (In- tion of iILFs and mILFs (Fig. 1). vitrogen Life Technologies), 10 mM HEPES (BioWhittaker), 1 mM so- These findings are consistent with previous observations of an dium pyruvate (BioWhittaker), 5 ϫ 10Ϫ5 M 2-ME (Sigma-Aldrich), 1% essential role for LT in ILF formation, because we observed iILF penicillin/streptomycin (Invitrogen Life Technologies), and 10% FCS (Hy- formation only when a cellular source of LT was supplied. Also Clone) at 37°C, 5% CO for 12 h. 2 consistent with this observation, we noted that, as the numbers of RNA isolation and real-time PCR cell types expressing LT increased, the numbers of iILFs formed Ϫ Ϫ increased. Although the formation of iILFs was not dependent on Spleen, PP, and mILFs were isolated from MD4JH / mice as described above. mILFs used for RNA isolation contained both bone marrow-derived LT-sufficient B lymphocytes, the progression from iILFs to mILFs cell populations and stromal cell populations. Non-PP non-mILF-bearing was dependent on LT-sufficient B lymphocytes. intestine from the distal small intestine of MD4JHϪ/Ϫ mice was identified using a dissecting microscope and removed. RNA was isolated from the ILF formation does not require Ag-specific lymphocyte appropriate cell populations and tissues using TRIzol (Invitrogen Life stimulation Technologies) and treated with DNase I (Ambion) to remove contaminat- ing DNA, and cDNA was synthesized from 2 ␮g of total RNA using The above findings demonstrate that LT-sufficient lymphocytes are Superscript II RNase HϪ reverse transcriptase (Invitrogen Life required for the formation of iILFs, and the progression of iILFs to Technologies). Expression of targets was detected by real-time PCR using mILFs is driven by LT-sufficient B lymphocytes. LT is largely ABI Prism 7700 sequence detection system and SYBR Green PCR master mix (Applied Biosystems). The following primers were used for detection expressed by activated lymphocytes. In addition to up-regulation of the targets (forward primers are listed first, followed by reverse primers): of LT expression, lymphocyte activation affects a number of prop- 18s, 5Ј-CGGCTACCACATCCAAGGAA-3Ј and 5Ј-GCTGGAATTAC erties that may be important for the formation of ILFs, including The Journal of Immunology 5723 Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021 FIGURE 1. LT-sufficient B lymphocytes are not required for the formation of iILFs, but are essential for the progression of iILFs to mILFs. Lethally irradiated LT␣Ϫ/Ϫ mice were reconstituted with LT␣Ϫ/Ϫ bone marrow, a combination of LT␣Ϫ/Ϫ and RagϪ/Ϫ bone marrow, a combination of LT␣Ϫ/Ϫ and JHϪ/Ϫ bone marrow, or a combination of C57BL/6 and JHϪ/Ϫ bone marrow as described in Materials and Methods. Following reconstitution, mice were sacrificed, and their small intestines were examined for the presence of iILFs (a) and mILFs (b) as described in Materials and Methods. Representative images of B220-stained whole mounts and UEA-I-stained whole mounts from each group are shown in c–f and g–j, respectively. k shows an enlarged view of a B220-stained whole mount demonstrating a cluster of B220ϩ cells at the base of a villus. Representative sections of H&E- and PNA-stained (green) sections of mILFs from LT␣Ϫ/Ϫ mice reconstituted with wild-type bone marrow are shown in l and m, respectively. Consistent with a requirement for LT in the formation of ILFs, no iILFs or mILFs were identified in LT␣Ϫ/Ϫ mice reconstituted with LT␣Ϫ/Ϫ bone marrow. iILFs, but not mILFs were formed in LT␣Ϫ/Ϫ recipients of a combination of LT␣Ϫ/Ϫ and RagϪ/Ϫ bone marrow and in LT␣Ϫ/Ϫ recipients of a combination of LT␣Ϫ/Ϫ and JHϪ/Ϫ bone marrow, documenting that LT-sufficient NK cells and T cells can supply the LT-dependent signals for iILF formation, but not mILF formation. LT␣Ϫ/Ϫ recipients of wild-type bone marrow developed both iILFs and mILFs, confirming that LT-sufficient B lymphocytes are essential for the progression of iILFs to mILFs. Black arrows denote structures counted as iILFs, black arrowheads denote structures counted as mILFs, and blue arrowheads (k) outline a villus p Ͻ 0.05 when compared with LT␣Ϫ/Ϫ recipients of C57BL/6 ,ء .arising from a B220ϩ cluster. Original magnification: c–f, ϫ20; g–j, ϫ40; k–m, ϫ200 and JHϪ/Ϫ bone marrow. the regulation of adhesion molecules and chemokine receptor ac- The rearranged Ig gene is specific for HEL, is of the IgM or IgD tivity (13–15). To define the requirement for Ag-specific lympho- isotype, and cannot undergo class switching to other isotypes. The cyte activation in ILF formation, we examined the intestines of MD4JHϪ/Ϫ mice, in comparison to MD4 mice on the wild-type back- TCR␤␦Ϫ/Ϫ mice and the intestines of MD4 BCR transgenic mice ground, express exclusively one BCR on all B lymphocytes. We have onaJHϪ/Ϫ background. Previous observations have demonstrated observed that MD4 mice on the wild-type background produce intes- that T lymphocyte-deficient mice contain ILFs (4); however, these tinal IgA at a level of 20% of that produced by wild-type mice due to studies could not assess whether both iILF and mILFs were formed endogenously expressed Ig genes (data not shown), whereas, in in the absence of T lymphocytes. Examination of the intestines of MD4JHϪ/Ϫ mice, there is no detectable intestinal IgA production TCR␤␦Ϫ/Ϫ mice revealed that these animals possess iILFs and (Fig. 2c). Examination of the intestines of MD4JHϪ/Ϫ mice revealed mILFs (Fig. 2), demonstrating that T lymphocytes are dispensable that these animals possess iILFs and mILFs (Fig. 2, a and b). for the formation of ILFs. Consistent with previous reports regard- The mILFs in the MD4JHϪ/Ϫ mice are present in the absence of ing the presence of ILFs in nu/nu mice (4), we noted that the ILFs manipulations to ablate PP formation, whereas in C57BL/6 mice, in the TCR␤␦Ϫ/Ϫ mice were generally smaller than those seen in significant numbers of mILFs are induced only following PP ab- their wild-type counterparts. lation. We feel this is consistent with a protective role for intestinal The MD4 mice possess a transgenic rearranged Ig L chain and Ig IgA in preventing inflammatory responses leading to the formation H chain linked to ␮ and ␦ C region genes of the IgHa allotype (10). of mILFs. MD4JHϪ/Ϫ mice are unable to produce IgA, and 5724 INNATE IMMUNE RESPONSES DRIVE ILF FORMATION

FIGURE 2. T lymphocytes and Ag-specific B lymphocyte responses are dispensable for ILF formation. Intestines from C57BL/6 mice, MD4JHϪ/Ϫ mice, TCR␤␦Ϫ/Ϫ mice, and C57BL/6 mice receiving LT␤R-Ig in utero to ablate PP formation were examined for the presence of iILFs (a), and/or mILFs (b). iILFs and mILFs were present in MD4JHϪ/Ϫ mice, and TCR␤␦Ϫ/Ϫ mice, indicating that T lymphocyte responses and Ag-specific B lymphocyte responses were dispensable for the formation of both iILFs and mILFs (a and b). Consistent with our previous observations, iILFs were present in unmanipulated C57BL/6 mice, but significant numbers of mILFs were only formed after manipulations to ablate PP formation. Evaluation of fecal IgA levels revealed that MD4JHϪ/Ϫ mice did not produce intestinal IgA, and that C57BL/6 mice receiving LT␤R-Ig in utero were relatively deficient in intestinal IgA production at 6–7 wk of age (c). TCR␤␦Ϫ/Ϫ mice had total fecal IgA levels that were significantly less than age-matched C57BL/6 mice (c). Consistent with this, we observed that TCR␤␦Ϫ/Ϫ mice had significantly fewer IgA-producing cells, 1329 Ϯ 254 per 106 small intestine lamina propria cells (n ϭ 6), when compared with age-matched C57BL/6 .p Ͻ 0.05 when compared with results from C57BL/6 mice ,ء .mice, 5863 Ϯ 886 per 106 small intestine lamina propria cells (n ϭ 3); p Ͻ 0.05 Downloaded from

C57BL/6 mice treated with LT␤R-Ig in utero, are relatively IgA induced molecule 1 ligand chemokine (ELC). These chemokines deficient until 7 wk of age, a time at which mILF formation can have several properties of interest relative to the formation of ILFs. easily be appreciated in these animals (Fig. 2c). Consistent with a The production of these chemokines in secondary lymphoid struc- relative IgA deficiency promoting mILF formation, we observed tures is diminished by abrogation of LT␤R signaling (18). The ␤␦Ϫ/Ϫ that TCR mice have reduced IgA production when com- overexpression of these chemokines in transgenic animals results http://www.jimmunol.org/ pared with age-matched wild-type mice, and have a reduced num- in lymphoid neogenesis (19, 20). And notably, BLC induces the ber of IgA-producing plasma cells in their small intestine (Fig. 2c). expression of cell surface LT in Ag-naive B lymphocytes (8). We examined the expression of these chemokines, and stromal ILF B lymphocytes have a naive phenotype and express cell-derived factor-1 (SDF-1) in spleen, mILFs, PP, and non-PP-, membrane-bound LT in the absence of Ag-specific stimulation non-ILF-bearing intestine from MD4JHϪ/Ϫ mice. SDF-1 is a che- In the absence of exogenous HEL, B lymphocytes in the mokine induced in response to inflammation, and has a capacity to MD4JHϪ/Ϫ mice do not receive stimulation through their BCR attract naive T lymphocytes and B lymphocytes (21, 22). In addi- and therefore are Ag naive. Consistent with this, we noted that ILF tion, SDF-1 also has a capacity to induce LT expression in naive Ϫ Ϫ B lymphocytes from the MD4JH / mice expressed the high-af- B lymphocytes (8). We observed that the mRNA expression of by guest on September 24, 2021 finity receptor for HEL and had a naive phenotype expressing low BLC and ELC were significantly increased in mILFs when com- levels of CD69 and CD86, and expressing high levels of CD23 pared with non-PP-, non-mILF-bearing small intestine (Fig. 4). (Fig. 3). ILF B lymphocytes from the MD4JHϪ/Ϫ mice expressed SLC expression was also elevated in mILFs; however, this did not higher levels of MHCII when compared with splenocytes from the reach statistical significance. Parallel changes in BLC, SLC, and same animals, but this expression was still lower than that seen in ELC expression were also seen in the spleen and PP. SDF-1 ex- splenocytes from MD4JHϪ/Ϫ mice stimulated with Ag (Fig. 3d). pression was increased in the spleen, but its expression was not Although events independent of Ag stimulation can modulate the increased in ILFs or PP when compared with non-PP-, non-ILF- expression of these molecules, the overall pattern of expression of bearing small intestine (Fig. 4). these molecules is supportive of an Ag-naive phenotype. Consistent with our findings of a requirement for LT-sufficient Discussion B lymphocytes in driving the progression of iILFs to mILFs, we ILFs are a recently appreciated member of the mucosal immune Ϫ Ϫ observed that mILF B lymphocytes from MD4JH / mice express system. Previous studies have demonstrated that ILFs are orga- cell surface LT␤ in the absence of HEL, and we observed that LT␣ nized, ectopic lymphoid structures that can be formed de novo in Ϫ Ϫ mRNA expression was increased in MD4JH / mILF cell popu- the intestine of adult animals (4, 6). These observations indicate lations enriched for B lymphocytes (Fig. 3). We observed a modest that ILFs are tertiary lymphoid structures. Lymphoid neogenesis, increase in LT␤ mRNA expression in the CD19-enriched popula- the process leading to tertiary lymphoid structure formation, has tion that was not significantly different from the CD19-depleted been observed in a number of chronic inflammatory and autoim- population (data not shown). These observations are consistent mune conditions including rheumatoid arthritis, Sjo¨gren’s syn- with expression of the membrane-bound LT because the cell sur- drome, primary sclerosing cholangitis, chronic hepatitis C infec- face expression of LT␤ occurs only in association with LT␣ and tion, myasthenia gravis, multiple sclerosis, and Hashimoto’s therefore requires the expression of LT␣, and LT␤ exhibits con- thyroiditis (23, 24). Tertiary lymphoid structures have also been stitutive mRNA expression in murine lymphocytes (16, 17). observed in a number of inflammatory conditions involving the gastrointestinal tract including inflammatory bowel disease, gas- mILFs express “homeostatic” chemokines in the absence of B tritis, and animal models of intestinal inflammation (2, 25–27). lymphocyte Ag stimulation Although the role ILFs and other tertiary lymphoid structures play A basal level of expression of chemokines has been described in in immune responses is unclear, their architecture and composition secondary lymphoid structures. Due to this constitutive expression, suggest that these structures allow efficient interactions between Ag, these chemokines have been referred to as homeostatic chemo- APCs, and lymphocytes. It has been suggested that tertiary lymphoid kines and include the following: B lymphocyte chemoattractant structures may contribute to inappropriate immune responses by (BLC), secondary lymphoid tissue chemokine (SLC), and EBV- facilitating interactions of Ag, APCs, and lymphocytes in an The Journal of Immunology 5725 Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021

FIGURE 3. ILF B lymphocytes have a naive phenotype and express cell surface LT in the absence of Ag stimulation. Single-cell suspensions from spleens, PP, and mILF were isolated from MD4JHϪ/Ϫ mice and analyzed by flow cytometry as described in Materials and Methods. ILF cellular populations were pooled from two mice for flow-cytometric analysis; splenic and PP populations were analyzed from individual mice. Consistent with an Ag-naive phenotype, B lymphocytes (CD19ϩ cells) from MD4JHϪ/Ϫ mILFs (filled gray histograms) expressed low levels of CD69 and CD86 and elevated levels of CD23 similar to the expression of these markers by MD4JHϪ/Ϫ splenic B lymphocytes (unfilled dark-line histograms) (a). MD4JHϪ/Ϫ mILF B lymphocytes expressed higher levels of MHCII when compared with freshly isolated MD4JHϪ/Ϫ splenic B lymphocytes; however, MHCII expression in MD4JHϪ/Ϫ splenocytes could be augmented above this level by overnight culture with HEL (unfilled light-line histograms). In a similar manner, culture of MD4JHϪ/Ϫ splenocytes with Ag up-regulated CD69 and CD86 expression and down-regulated CD23 expression on B lymphocytes, consistent with an Ag-activated phenotype (a). B lymphocytes from MD4JHϪ/Ϫ mILFs expressed high levels of a BCR-specific for HEL (b). B lymphocytes from each tissue expressed cell surface LT␤ (d), a component of the membrane-bound LT complex (LT␣1␤2). Notably, B lymphocytes represented the majority of LT-expressing cells in mILFs from MD4JHϪ/Ϫ mice (d). Consistent with this, CD19-enriched MD4JHϪ/Ϫ mILF B lymphocytes demonstrated increased p Ͻ 0.05 when ,ء .expression of LT␣ when compared with CD19-depleted populations (c). Results shown are representative of one of two experiments comparing CD19-enriched and -depleted populations.

environment that lacks the “normal” regulatory environment of it is not dependent on LT-sufficient B lymphocytes; LT-sufficient secondary lymphoid structures (24). Given this hypothesis and the T lymphocytes and NK cells can provide the necessary source of association of tertiary lymphoid structures with chronic inflamma- LT for iILF formation. However, LT-sufficient B lymphocytes are tory and autoimmune conditions, there has been considerable in- essential for the progression of iILFs to mILFs. In contrast to mILF terest in understanding the steps leading to lymphoid neogenesis. formation, LT-sufficient B lymphocytes are not required for sec- Similar to the formation of PP and other secondary lymphoid ondary lymphoid structure formation, and LT-sufficient B lympho- structures, the formation of ILFs is dependent on LT and LT␤R. cytes are dispensable for the development of M cells in PP FAE The requirement for LT and LT␤R occurs early in ILF formation (28). Recent studies have suggested an alternative cellular source as LT- and LT␤R-deficient mice lack both mILFs and iILFs. We of LT inducing the formation of iILFs, the adult counterpart of the have previously observed that LT-sufficient B lymphocytes are re- fetal lymphoid tissue inducer (LTi) cell. Fetal LTi and their adult quired for mILF formation (6); whether other LT-sufficient lym- counterparts are among the few cell types that express the nuclear phocytes can provide the source of LT for iILF formation has not hormone receptor ROR␥␶ (29, 30). In the absence of ROR␥␶, these been previously addressed. In this study, we demonstrate that, al- cells do not develop, and consequently lymph nodes, PP, crypto- though iILF formation is dependent on LT-sufficient lymphocytes, patches, and ILFs fail to form (29, 30). The fetal LTi cells express LT 5726 INNATE IMMUNE RESPONSES DRIVE ILF FORMATION

LT-sufficient B lymphocytes are essential for the progression of iILFs to mILFs, this process occurs in the absence of adaptive immune responses, and therefore is driven by the innate immune system. Due to their constitutive expression in secondary lymphoid organs, SLC, ELC, and BLC have been referred to as homeostatic chemokines. These chemokines regulate lymphocyte entry and or- ganization within secondary lymphoid structures (31). SLC and ELC are ligands for CCR7, which is highly expressed by naive T lymphocytes, maturing dendritic cells, and activated B lymphocytes (15, 32–34). BLC is the only known ligand for CXCR5, which is expressed by B lymphocytes and subsets of T lymphocytes (35, 36). Although the formation of tertiary lymphoid struc- FIGURE 4. ILFs express homeostatic chemokines and LT in the ab- tures in CXCR5-deficient and CCR7-deficient mice has not been sence of Ag-specific B lymphocyte stimulation. RNA was isolated from evaluated formally, the ectopic expression of BLC, SLC, and ELC spleen, PP, mILFs, and non-PP-, non-mILF-bearing small intestine from in transgenic mice has been shown to induce ectopic organized Ϫ Ϫ MD4JH / mice, and used to evaluate differential target gene expression aggregates of lymphocytes, suggesting that these chemokines play as described in Materials and Methods. Consistent with prior observations a significant role in lymphoid neogenesis (19, 20). Additional ob- regarding the expression of homeostatic chemokines in secondary lym- servations supporting a role for these chemokines in lymphoid phoid structures, the spleen and PP demonstrated elevated expression of neogenesis include the demonstration of BLC and/or SLC expres- Downloaded from BLC, SLC, and ELC. ILFs also demonstrated elevated expression of BLC, sion in tertiary lymphoid structures in ulcerative colitis, primary SLC, and ELC when compared with non-PP-, non-ILF-bearing intestine. The spleen, but not PP, mILFs, nor non-PP-, non-ILF-bearing small intes- sclerosing cholangitis, Sjo¨gren’s syndrome, and rheumatoid arthri- tine, demonstrated elevated expression of SDF-1, a chemokine that can be tis (23, 37–39). induced in response to inflammatory stimuli, and has been demonstrated to LT and LT␤R have established roles in lymphoid neogenesis. up-regulate LT expression in Ag-naive B lymphocytes. Tissues from three Animal models demonstrated that overexpression of LT can in-

mice were pooled for analysis. The data presented are representative of one duce lymphoid neogenesis, that LT- and LT␤R-deficient animals http://www.jimmunol.org/ -p Ͻ 0.05 when compared with results from non-PP-, do not form tertiary lymphoid structures, and that tertiary lym ,ء .of two experiments non-mILF-bearing small intestine. phoid structure formation can be inhibited by continuous LT␤R blockade (6, 20, 40). One role that LT and LT␤R may play in this process is the production of homeostatic chemokines, because oth- and are believed to be the cellular source of LT required during em- ers have demonstrated that blockade of LT␤R signaling in wild- bryogenesis for the formation of secondary lymphoid structures. Al- type mice reduces splenic expression of BLC, SLC, and ELC (18). though the expression of LT on the adult counterparts to the fetal LTi Further evidence supporting a role for the production of these che- has not been formally investigated, it is felt that these cells may de- mokines downstream of LT and the LT␤R in lymphoid neogenesis liver the LT-dependent signal required for cryptopatch and ILF for- is the observation of increased BLC and SLC expression at sites of by guest on September 24, 2021 mation. Whether this signal occurs very early, forming cryptopatches lymphoid neogenesis in transgenic mice with ectopic expression of or other iILF precursors, or whether it is the signal driving iILF for- LT␣ (40). mation is not clear. Our observations of increasing numbers of iILFs Additional studies have implicated that these chemokines also as we add increasing cellular sources of LT (NK cells vs NK cells and act upstream of the LT␤R in the formation of tertiary lymphoid T lymphocytes vs NK cells, T lymphocytes, and B lymphocytes) structures. In transgenic mice with ectopic expression of BLC, would favor this step being driven by mature LT-expressing lympho- lymphoid neogenesis was found to be LT␣ dependent, suggesting cytes as opposed to the adult counterparts to LTi cells. that LT may have a function downstream of BLC (20). Other stud- The differential requirements for LT-sufficient B lymphocytes in ies have demonstrated that SLC and ELC can induce the expres- ␣ ␤ ϩ the formation of PP and ILFs not only provides a distinction be- sion of LT 1 2 in naive CD4 T lymphocytes (19). And, most tween these structures, it also provides a focal point for under- relevant to our observations, BLC has been shown to induce ␣ ␤ standing the events leading to lymphoid neogenesis in the small LT 1 2 expression in Ag-naive lymphocytes, setting up a posi- ␣ ␤ intestine. LT expression is induced in lymphocytes following ac- tive-feedback loop in which BLC induces LT 1 2 expression in tivation. In B-2 B lymphocytes, which exclusively comprise the B Ag-naive B lymphocytes and these B lymphocytes induce further lymphocyte population of mILFs, this activation occurs in re- BLC production by ligation of the LT␤R on stromal cells (8). In sponse to CD40 ligation in the context of T cell help. BCR Ag this study, we demonstrate that ILF formation is independent of B binding, Ag internalization and processing, and cognate interac- lymphocyte Ag exposure and that ILF formation is arrested at an tions of MHCII/peptide complexes with TCR are prerequisites for immature stage in the absence of LT-sufficient B lymphocytes. We this process. We observed that ILF formation occurs in the absence also demonstrate that Ag-naive ILF B lymphocytes have elevated of T lymphocytes and therefore in the absence of T cell help. expression of cell surface LT, and that ILFs have increased ex- Further supporting this observation, we demonstrate that ILF for- pression of homeostatic chemokines in the absence of B lympho- mation is preserved in the absence of BCR ligation by Ag, and that cyte Ag exposure. These observations implicate the existence of a Ag-naive ILF B lymphocytes express LT. The form of LT, soluble BLC-driven positive-feedback loop in ILF formation and specifi- ␣ ␣ ␤ (LT 3) or membrane-bound (LT 1 2), driving this step is not cally in the progression of iILFs to mILFs. known. We have previously demonstrated that TNFRI, a receptor Although mILFs are infrequent in unmanipulated C57BL/6 ␣ Ϫ/Ϫ for LT 3, is required for the transition of iILFs to mILFs (6), mice, their formation occurred spontaneously in MD4JH mice suggesting that LT-sufficient B lymphocytes may be the source of and TCR␤␦Ϫ/Ϫ mice. Treatment of C57BL/6 mice with LT␤R-Ig TNFRI ligands driving this transition. The findings we present here in utero to ablate PP formation can induce ILF formation in ␣ could be consistent with either soluble LT 3 or membrane-bound C57BL/6 mice. Consistent with the hypothesis that mILF forma- ␣ ␤ LT 1 2 driving this step. Therefore, although the process of ILF for- tion is induced by inflammatory responses precipitated by a rela- mation is dependent on LT-sufficient lymphocytes, and specifically tive deficiency of intestinal IgA, we observed that 7-wk-old The Journal of Immunology 5727

C57BL/6 mice receiving LT␤R-Ig in utero were relatively defi- immunization in the absence of their Ag (45). In the present study, cient in intestinal IgA production when compared with unmanipu- we describe the formation of ILFs in the absence of BCR-depen- lated C57BL/6 controls. In addition, we observed “spontaneous” dent and TCR-dependent responses, and in conjunction with our formation of mILFs in MD4JHϪ/Ϫ mice, which lack the capacity previous studies demonstrating that the ILF formation can be to produce IgA, and in TCR␤␦Ϫ/Ϫ mice, which produce lower driven by lumenal flora, we suggest that innate immune responses intestinal IgA levels when compared with their wild-type counter- to lumenal bacteria are driving this event. The apparent differences parts. In the absence of T cell help, the IgA produced is primarily, between these studies can be reconciled by breaking ILF formation if not exclusively, derived from B-1 B lymphocytes. In comparison down into a series of events, all of which may be influenced by to B-2 B lymphocytes, B-1 B lymphocytes have a restricted V lumenal bacteria. The studies presented here indicate that the for- region repertoire, and undergo little-to-no somatic hypermutation, mation of ILFs, or the events leading to the accumulation of lym- resulting in the production of Ab with relatively low affinity. phocytes into a loosely organized collection with an overlying Therefore, in addition to the reduced amount of IgA, the IgA pro- dome-shaped FAE, can occur independent of adaptive immune duced by TCR␤␦Ϫ/Ϫ mice may be less effective at preventing an responses. Studies on AIDϪ/Ϫ mice suggest that ILF expansion, or inflammatory response to lumenal bacteria. In support of this hy- hyperplasia, can be driven by B lymphocytes “sensing” bacterial pothesis, studies of mice deficient in activation-induced cytidine stimuli (42, 43). This expansion is consistent with the known re- deaminase (AID), which have impaired class switch recombination sponses of B lymphocytes to both cognate Ags, recognized and somatic hypermutation resulting in impaired IgA production through the BCR, and noncognate stimuli, such as LPS, leading to (41), revealed that these mice have hyperplastic ILFs and altered polyclonal activation and expansion. Although our studies indicate bacterial flora (42). The ILF hyperplasia could be reversed with that adaptive immune responses are not required for the formation Downloaded from antibiotic treatment to decrease the bacterial flora, or normalization of ILFs, adaptive immune responses play a role in the function of of the IgA production by the generation of parabiotic AIDϪ/Ϫ and ILFs. The function of ILFs is still being investigated, but current wild-type mice (43). studies suggest that ILFs are sites of B-2 B lymphocyte stimula- Recent studies investigated the ability of lumenal stimuli to in- tion, driving their subsequent differentiation into IgA-producing fluence the gut-associated lymphoid tissues, and suggested varying plasma cells that contribute to the production of intestinal IgA (5,

pathways in which these stimuli mediate their effects. The ability 46). In this context, the development of a germinal center, which http://www.jimmunol.org/ of exogenous stimuli to induce ILFs was demonstrated by the for- facilitates the stimulation and differentiation of B-2 B lymphocytes mation of mILFs in germfree mice reconstituted with normal cecal into plasma cells, would be a later event in ILF formation, and flora (6). Studies of AID-deficient mice support the role of bacte- could be driven by cognate interactions of TCR with Ag/MHC rial flora in expanding ILFs in size and cellularity, and have sug- complexes. In line with these observations, we have not been able gested that B cell sensing of lumenal bacterial drives this ILF to demonstrate the presence of germinal centers in mILFs from hyperplasia (42). Studies of BCR-deficient mice expressing limit- TCR␤␦Ϫ/Ϫ mice. ing levels of the EBV protein LMP2A to rescue B lymphocyte The findings presented here expand the previous model of ILF survival in the absence of BCR signaling revealed that these mice formation (Fig. 5). ILF formation occurs independent of adaptive developed germinal centers in their PP and mesenteric LN, but not immune responses and is driven by the innate immune system. by guest on September 24, 2021 in their spleen (44). The formation of germinal centers in the PP Any LT-sufficient lymphocyte may deliver the necessary signals to and mesenteric LN was driven by lumenal flora, because treating allow accumulation of B lymphocytes into clusters at the base of the mice with antibiotics could abolish these germinal centers (44). villi and form iILFs; however, the progression of iILFs to mILFs In addition, the formation of these germinal centers was dependent is dependent on LT-sufficient B lymphocytes, is independent of on the presence of T lymphocytes because germinal center forma- Ag-specific stimulation of B lymphocytes, and occurs in the ab- tion was absent when these animals were bred onto a recombina- sence of T cell help. Consistent with previous observations, we tion activation gene-deficient background (44). This is most con- propose that the progression of iILFs to mILFs is driven by a sistent with TCR-mediated recognition of a bacterial product chemokine-driven positive-feedback loop. because other studies demonstrated that TCR transgenic mice have These observations have several implications for the potential an impaired ability to develop splenic germinal centers following role of ILFs in the immune response. Because ILF formation is

FIGURE 5. Putative sequence of events in ILF formation. Events elucidated by this study are indicated in bold type. The formation of iILFs and mILFs is driven by innate immune responses because it occurs in the absence of T lymphocytes and in the absence of Ag-specific B lymphocyte responses. The requirement for LT and LT␤R occurs early in this process, because LT␣Ϫ/Ϫ mice and LT␤RϪ/Ϫ mice lack both iILFs and mILFs. The formation of iILFs can be rescued in LT␣Ϫ/Ϫ mice by supplying a cellular source of LT; LT-sufficient T lymphocytes and/or LT-sufficient NK cells are sufficient sources of LT to drive the formation of iILFs. The progression of iILFs to mILFs requires LT-sufficient B lymphocytes, and occurs independent of Ag-specific stimulation of B lymphocytes. The Ag-independent expression of LT by B lymphocytes and the requirement for B lymphocyte LT expression is consistent with a chemokine-driven, positive-feedback loop driving the progression of iILFs to mILFs. Based upon prior observations, the formation of germinal centers is likely dependent on T lymphocyte responses. 5728 INNATE IMMUNE RESPONSES DRIVE ILF FORMATION driven by the innate immune system, ILF formation should be 21. Bleul, C. C., J. L. Schultze, and T. A. Springer. 1998. B lymphocyte chemotaxis reproducible in response to the appropriate inducing stimuli and regulated in association with microanatomic localization, differentiation state, and B cell receptor engagement. J. Exp. Med. 187:753. should be able to occur throughout life. ILFs preferentially contain 22. Bleul, C. C., R. C. Fuhlbrigge, J. M. Casasnovas, A. Aiuti, and T. A. Springer. Ag-naive lymphocytes, suggesting that ILFs are sites of initial in- 1996. A highly efficacious lymphocyte chemoattractant, stromal cell-derived factor 1 (SDF-1). J. Exp. Med. 184:1101. teractions of lymphocytes with Ag and APCs. Therefore, events 23. Grant, A. J., S. Goddard, J. Ahmed-Choudhury, G. Reynolds, D. G. Jackson, occurring within ILFs will shape the phenotype of subsequent im- M. Briskin, L. Wu, S. G. Hubscher, and D. H. Adams. 2002. Hepatic expression mune responses mediated by these lymphocytes. of secondary lymphoid chemokine (CCL21) promotes the development of portal- associated lymphoid tissue in chronic inflammatory liver disease. Am. J. Pathol. 160:1445. Acknowledgments 24. Hjelmstrom, P. 2001. Lymphoid neogenesis: de novo formation of lymphoid We thank W. Stenson, R. Lorenz, and E. Newberry for assistance with the tissue in chronic inflammation through expression of homing chemokines. J. Leu- kocyte Biol. 69:331. preparation of this manuscript. 25. Dohi, T., K. Fujihashi, P. D. Rennert, K. Iwatani, H. Kiyono, and J. R. McGhee. 1999. Hapten-induced colitis is associated with colonic patch hypertrophy and T Disclosures helper cell 2-type responses. J. Exp. Med. 189:1169. The authors have no financial conflict of interest. 26. Yeung, M. M., S. Melgar, V. Baranov, A. Oberg, A. Danielsson, S. Hammarstrom, and M. L. Hammarstrom. 2000. Characterisation of mucosal lymphoid aggregates in ulcerative colitis: immune cell phenotype and TcR-␥␦ References expression. Gut 47:215. 1. Fujimura, Y., R. Kamoi, and M. Iida. 1996. Pathogenesis of aphthoid ulcers in 27. Eidt, S., and M. Stolte. 1993. Prevalence of lymphoid follicles and aggregates in Crohn’s disease: correlative findings by magnifying colonoscopy, electron mi- Helicobacter pylori gastritis in antral and body mucosa. J. Clin. Pathol. 46:832. croscopy, and immunohistochemistry. Gut 38:724. 28. Tumanov, A. V., D. V. Kuprash, J. A. Mach, S. A. Nedospasov, and A. V. Chervonsky. 2004. Lymphotoxin and TNF produced by B cells are dis-

2. Kaiserling, E. 2001. Newly-formed lymph nodes in the submucosa in chronic Downloaded from inflammatory bowel disease. Lymphology 34:22. pensable for maintenance of the follicle-associated epithelium but are required for 3. Lockhart-Mummery, H. E., and B. C. Morson. 1960. Crohn’s disease (regional development of lymphoid follicles in the Peyer’s patches. J. Immunol. 173:86. ␣␤ enteritis) of the large intestine and its distinction from ulcerative colitis. Gut 1:87. 29. Eberl, G., and D. R. Littman. 2004. Thymic origin of intestinal T cells re- ␥ ϩ 4. Hamada, H., T. Hiroi, Y. Nishiyama, H. Takahashi, Y. Masunaga, S. Hachimura, vealed by fate mapping of ROR t cells. Science 305:248. S. Kaminogawa, H. Takahashi-Iwanaga, T. Iwanaga, H. Kiyono, et al. 2002. 30. Eberl, G., S. Marmon, M. J. Sunshine, P. D. Rennert, Y. Choi, and D. R. Littman. Identification of multiple isolated lymphoid follicles on the antimesenteric wall of 2004. An essential function for the nuclear receptor ROR␥(t) in the generation of the mouse small intestine. J. Immunol. 168:57. fetal lymphoid tissue inducer cells. Nat. Immunol. 5:64. 5. Shikina, T., T. Hiroi, K. Iwatani, M. H. Jang, S. Fukuyama, M. Tamura, T. Kubo, 31. Muller, G., U. E. Hopken, and M. Lipp. 2003. The impact of CCR7 and CXCR5 H. Ishikawa, and H. Kiyono. 2004. IgA class switch occurs in the organized on lymphoid organ development and systemic immunity. Immunol. Rev. 195:117. http://www.jimmunol.org/ nasopharynx- and gut-associated lymphoid tissue, but not in the diffuse lamina 32. Campbell, J. J., E. P. Bowman, K. Murphy, K. R. Youngman, M. A. Siani, propria of airways and gut. J. Immunol. 172:6259. D. A. Thompson, L. Wu, A. Zlotnik, and E. C. Butcher. 1998. 6-C-kine (SLC), 6. Lorenz, R. G., D. D. Chaplin, K. G. McDonald, J. S. McDonough, and a lymphocyte adhesion-triggering chemokine expressed by high endothelium, is R. D. Newberry. 2003. Isolated lymphoid follicle formation is inducible and an agonist for the MIP-3␤ receptor CCR7. J. Cell Biol. 141:1053. dependent upon lymphotoxin-sufficient B lymphocytes, lymphotoxin ␤ receptor, 33. Dieu, M. C., B. Vanbervliet, A. Vicari, J. M. Bridon, E. Oldham, S. Ait-Yahia, and TNF receptor I function. J. Immunol. 170:5475. F. Briere, A. Zlotnik, S. Lebecque, and C. Caux. 1998. Selective recruitment of 7. Mebius, R. E., P. Rennert, and I. L. Weissman. 1997. Developing lymph nodes immature and mature dendritic cells by distinct chemokines expressed in different collect CD4ϩCD3ϪLT␤ϩ cells that can differentiate to APC, NK cells, and fol- anatomic sites. J. Exp. Med. 188:373. licular cells but not T or B cells. Immunity 7:493. 34. Kellermann, S. A., S. Hudak, E. R. Oldham, Y. J. Liu, and L. M. McEvoy. 1999. 8. Ansel, K. M., V. N. Ngo, P. L. Hyman, S. A. Luther, R. Forster, J. D. Sedgwick, The CC chemokine receptor-7 ligands 6Ckine and macrophage inflammatory J. L. Browning, M. Lipp, and J. G. Cyster. 2000. A chemokine-driven positive protein-3␤ are potent chemoattractants for in vitro- and in vivo-derived dendritic feedback loop organizes lymphoid follicles. Nature 406:309. cells. J. Immunol. 162:3859. by guest on September 24, 2021 9. Gu, H., Y. R. Zou, and K. Rajewsky. 1993. Independent control of immunoglob- 35. Schaerli, P., K. Willimann, A. B. Lang, M. Lipp, P. Loetscher, and B. Moser. ulin switch recombination at individual switch regions evidenced through Cre- 2000. CXC chemokine receptor 5 expression defines follicular homing T cells loxP-mediated gene targeting. Cell 73:1155. with B cell helper function. J. Exp. Med. 192:1553. 10. Goodnow, C. C., J. Crosbie, S. Adelstein, T. B. Lavoie, S. J. Smith-Gill, 36. Ansel, K. M., L. J. McHeyzer-Williams, V. N. Ngo, M. G. McHeyzer-Williams, and R. A. Brink, H. Pritchard-Briscoe, J. S. Wotherspoon, R. H. Loblay, K. Raphael, J. G. Cyster. 1999. In vivo-activated CD4 T cells upregulate CXC chemokine receptor et al. 1988. Altered immunoglobulin expression and functional silencing of self- 5 and reprogram their response to lymphoid chemokines. J. Exp. Med. 190:1123. reactive B lymphocytes in transgenic mice. Nature 334:676. 37. Takemura, S., A. Braun, C. Crowson, P. J. Kurtin, R. H. Cofield, W. M. O’Fallon, 11. De Togni, P., J. Goellner, N. H. Ruddle, P. R. Streeter, A. Fick, S. Mariathasan, J. J. Goronzy, and C. M. Weyand. 2001. Lymphoid neogenesis in rheumatoid S. C. Smith, R. Carlson, L. P. Shornick, J. Strauss-Schoenberger, et al. 1994. synovitis. J. Immunol. 167:1072. Abnormal development of peripheral lymphoid organs in mice deficient in lym- 38. Carlsen, H. S., E. S. Baekkevold, F. E. Johansen, G. Haraldsen, and P. Brandtzaeg. photoxin. Science 264:703. 2002. B cell attracting chemokine 1 (CXCL13) and its receptor CXCR5 are 12. Newberry, R. D., J. S. McDonough, K. G. McDonald, and R. G. Lorenz. 2002. expressed in normal and aberrant gut associated lymphoid tissue. Gut 51:364. Postgestational lymphotoxin/lymphotoxin ␤ receptor interactions are essential for 39. Salomonsson, S., P. Larsson, P. Tengner, E. Mellquist, P. Hjelmstrom, and the presence of intestinal B lymphocytes. J. Immunol. 168:4988. M. Wahren-Herlenius. 2002. Expression of the B cell-attracting chemokine 13. Spaargaren, M., E. A. Beuling, M. L. Rurup, H. P. Meijer, M. D. Klok, CXCL13 in the target organ and autoantibody production in ectopic lymphoid S. Middendorp, R. W. Hendriks, and S. T. Pals. 2003. The B cell antigen receptor tissue in the chronic inflammatory disease Sjo¨gren’s syndrome. Scand. J. Immu- controls integrin activity through Btk and PLC␥2. J. Exp. Med. 198:1539. nol. 55:336. 14. Liao, F., A. K. Shirakawa, J. F. Foley, R. L. Rabin, and J. M. Farber. 2002. 40. Hjelmstrom, P., J. Fjell, T. Nakagawa, R. Sacca, C. A. Cuff, and N. H. Ruddle. Human B cells become highly responsive to macrophage-inflammatory protein- 2000. Lymphoid tissue homing chemokines are expressed in chronic inflamma- 3␣/CC chemokine ligand-20 after cellular activation without changes in CCR6 tion. Am. J. Pathol. 156:1133. expression or ligand binding. J. Immunol. 168:4871. 41. Muramatsu, M., K. Kinoshita, S. Fagarasan, S. Yamada, Y. Shinkai, and 15. Ngo, V. N., H. L. Tang, and J. G. Cyster. 1998. Epstein-Barr virus-induced T. Honjo. 2000. Class switch recombination and hypermutation require activa- molecule 1 ligand chemokine is expressed by dendritic cells in lymphoid tissues tion-induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell and strongly attracts naive T cells and activated B cells. J. Exp. Med. 188:181. 102:553. 16. Koni, P. A., R. Sacca, P. Lawton, J. L. Browning, N. H. Ruddle, and 42. Fagarasan, S., M. Muramatsu, K. Suzuki, H. Nagaoka, H. Hiai, and T. Honjo. R. A. Flavell. 1997. Distinct roles in lymphoid organogenesis for lymphotoxins 2002. Critical roles of activation-induced cytidine deaminase in the homeostasis ␣ and ␤ revealed in lymphotoxin ␤-deficient mice. Immunity 6:491. of gut flora. Science 298:1424. 17. Millet, I., and N. H. Ruddle. 1994. Differential regulation of lymphotoxin (LT), 43. Suzuki, K., B. Meek, Y. Doi, M. Muramatsu, T. Chiba, T. Honjo, and lymphotoxin-␤ (LT-␤), and TNF-␣ in murine T cell clones activated through the S. Fagarasan. 2004. Aberrant expansion of segmented filamentous bacteria in TCR. J. Immunol. 152:4336. IgA-deficient gut. Proc. Natl. Acad. Sci. USA 101:1981. 18. Ngo, V. N., H. Korner, M. D. Gunn, K. N. Schmidt, D. S. Riminton, 44. Casola, S., K. L. Otipoby, M. Alimzhanov, S. Humme, N. Uyttersprot, M. D. Cooper, J. L. Browning, J. D. Sedgwick, and J. G. Cyster. 1999. Lym- J. L. Kutok, M. C. Carroll, and K. Rajewsky. 2004. B cell receptor signal strength photoxin ␣/␤ and tumor necrosis factor are required for stromal cell expression determines B cell fate. Nat. Immunol. 5:317. of homing chemokines in B and T cell areas of the spleen. J. Exp. Med. 189:403. 45. Wu, Z. Q., Y. Shen, A. Q. Khan, C. L. Chu, R. Riese, H. A. Chapman, 19. Luther, S. A., A. Bidgol, D. C. Hargreaves, A. Schmidt, Y. Xu, J. Paniyadi, O. Kanagawa, and C. M. Snapper. 2002. The mechanism underlying T cell help M. Matloubian, and J. G. Cyster. 2002. Differing activities of homeostatic che- for induction of an antigen-specific in vivo humoral immune response to intact mokines CCL19, CCL21, and CXCL12 in lymphocyte and dendritic cell recruit- Streptococcus pneumoniae is dependent on the type of antigen. J. Immunol. ment and lymphoid neogenesis. J. Immunol. 169:424. 168:5551. 20. Luther, S. A., T. Lopez, W. Bai, D. Hanahan, and J. G. Cyster. 2000. BLC 46. Lorenz, R. G., and R. D. Newberry. 2004. Isolated lymphoid follicles can func- expression in pancreatic islets causes B cell recruitment and lymphotoxin-depen- tion as sites for induction of mucosal immune responses. Ann. NY Acad. Sci. dent lymphoid neogenesis. Immunity 12:471. 1029:44.