During fetal life, CD4+CD3− lymphoid tissue inducer (LTi) cells are required for lymph node and Peyer’s patch development in mice. In adult animals, CD4+CD3− cells are found in low numbers in lymphoid organs. Whether adult CD4+CD3− cells are LTi cells and are generated and maintained through cytokine signals has not been directly addressed. In this study we show that adult CD4+CD3− cells adoptively transferred into neonatal CXCR5−/− mice induced the formation of intestinal lymphoid tissues, demonstrating for the first time their bona fide LTi function. Increasing IL-7 availability in wild-type mice either by IL-7 transgene expression or treatment with IL-7/anti-IL-7 complexes increased adult LTi cell numbers through de novo generation from bone marrow cells and increased the survival and proliferation of LTi cells. Our observations demonstrate that adult CD4+lineage− cells are LTi cells and that the availability of IL-7 determines the size of the adult LTi cell pool.
During fetal life, the development of secondary lymphoid organs in mice is determined by the interactions between lymphotoxin (LT)3 α1β2+CD45+CD4+CD3− lymphoid tissue inducer (LTi) cells and mesenchymal LTβR+VCAM-1+ organizer cells (1, 2). The first lymph node (LN) and Peyer’s patch (PP) anlagen in fetal mice develop between embryonic day (E)13.5 and E15.5 from clusters of LTi and organizer cells (3, 4). The generation of LTi cells is dependent on the helix-loop-helix protein Id2 and the retinoic acid-related orphan receptor (ROR) γ (5, 6). In the absence of LTi cells or when the LTβR signaling pathway is perturbed, LNs and PPs do not form (7, 8). Moreover, IL-7 has an important role in LN and PP development, and we have recently shown that this relies on the IL-7-dependent survival of fetal LTi cells and their fetal liver (FL) progenitors (9). The number and localization of secondary lymphoid organs is developmentally fixed in mice. By increasing the pool of fetal LTi cells in vivo through IL-7 transgene expression, this restriction can be overcome, leading to the formation of additional ectopic lymphoid organs (9).
Among CD3− cells, RORγt is an exclusive marker of fetal LTi cells (10). Fate-mapping experiments with ROR(γt)+/GFP mice led to the conclusion that LTi cells persist in the gut of adult mice (11). In wild-type (WT) adult mice, CD4+CD3− cells were found in the spleen where they help organize the architecture and optimize immune responses (12, 13). In addition, CD4+CD3− cells accumulate in the spleens of lymphocytic choriomeningitis virus-infected mice (14). Neonatal and adult splenic CD4+CD3− cells display similar genetic profiles that distinguish them from other cells. Both express LTα, TNF-α, c-Kit (CD117), IL-7Rα (CD127), and TNF receptors (TNFRII, TRANCER, and HVEM) (15); but in contrast to adult cells, fetal LTi cells express neither CD30 ligand nor OX40 ligand (16) unless treated in vitro with IL-7 (16) or TNFSF15 (15), respectively. The mechanism for the generation and persistence of adult CD4+CD3− cells and their possible function as inducers of newly formed lymphoid organ remain to be addressed. In the present study, we first sought to determine whether adult CD4+CD3− cells have bona fide LTi activity and, secondly, whether they could be expanded by increased availability of IL-7 either in transgenic or IL-7/anti-IL-7 complex-treated mice (17). In this study, we show that adult CD4+CD3− cells are functional LTi cells that are dependent on RORγt and IL-7.
Materials and Methods
Mice and in vivo treatment
C57BL/6 (CD45.2) and C57BL/6 (CD45.1) mice were purchased from Harlan. IL-7/CIITA double-transgenic, RORγ−/−, CXCR5−/−, IL-7Rα−/−, and IL-7−/− mice were previously described (9, 18, 19). RAG-2−/− and RAG2−/−γc−/− mice on C57BL/6 background were kindly provided by J. Kirberg (Max Planck Institute, Freiburg, Germany). H-2k bcl-2 transgenic mice were originally generated in the laboratory of I. L. Weissman (Stanford University School of Medicine, Stanford, CA) (20) and provided by A. Trumpp (Swiss Institute for Experimental Cancer Research, Epalinges, Switzerland). Immune complexes containing 10 μg of recombinant human IL-7 (R&D Systems) and 50 μg of anti-IL-7 neutralizing mAb (M25) were i.p. injected into mice on days 0, 2, and 4, and mice were analyzed on day 7. If indicated, IL-7/anti-IL-7 complex treatment was prolonged over 3 wk. FACS-sorted LTi cells were i.v. injected into neonatal CXCR5−/− mice as described (18). Splenocytes, LTi, bone marrow (BM), or FL (E13.5) cells were i.v. injected into sublethally (600 rad) irradiated mice (RAG2−/−γc−/−, IL-7Rα−/− mice). Total BM was treated with anti-CD4 (clone RL172.1) and anti-CD8 (clone M31) mAbs followed by incubation with low Tox-M rabbit complement (Cedarlane Laboratories) at 37°C before adoptive transfer. Mice were housed under standard conditions in a pathogen-free mouse facility. The study received the approval of the Cantonal Veterinary Office of the city of Basel, Switzerland.
Abs either biotinylated or conjugated to fluorochromes (FITC, PE, PE-Cy5, PE-Cy7, or allophycocyanin) and purchased from BioLegend, eBioscience or BD Pharmingen were used against the following mouse Ags: CD4 (clone: RM4-5), c-Kit/CD117 (2B8), CD3 (145-2C11), CD8 (53-6.7), TCRαβ (H57-597), TCRγδ (UC7-13D5), CD11c (N418), CD19 (6D5), B220 (RA3-B2), NK1.1 (PK136), Gr-1 (RB6-8C5), TER119 (TER119), IL-7Rα (A7R34), CD18 (M18/2), β1 integrin (HMb1-1), α4β7 integrin (DATK32), Sca-1 (D7), CD44 (IM7), CD62L (MEL-14), LTβR-human Fc (gift from J. Browning, Biogen, Cambridge MA), TRANCE (IK22/5), TRANCER (R12-31), ICAM-1 (3E2), CD184 (2B11/CXCR4), CD45.1 (A20), CD45.2 (104), VCAM-1 (429), IL-7 (lot no. AWR01; R&D Systems), M25 (mouse anti-human; Amgen), and RORγ (AFKJS9; e-Bioscience). Secondary Abs were streptavidin (purchased from BioLegend) conjugated to FITC, PE, PE-Cy5, PE-Cy7 or allophycocyanin, HRP-conjugated goat anti-rat IgG (Biosource), Alexa Fluor 488 donkey anti-goat IgG (Invitrogen), and Cy3-conjugated goat anti-rat IgG (Jackson ImmunoResearch Laboratories).
Flow cytometry and cell sorting
Cells (106) derived from various lymphoid organs were stained with mAbs using standard protocols (9). Cells were resuspended in PBS containing 3% FCS, stained with biotinylated or fluorochrome-conjugated Abs, and analyzed using a FACScalibur flow cytometer (BD Biosciences). To analyze for the presence of LTi cells, total splenocytes from 6- to 8-wk-old mice were depleted of erythrocytes, stained with biotinylated anti-CD4 Ab, and enriched using MACS separation (Miltenyi Biotec). MACS-enriched cells were labeled with anti-CD4, anti-c-Kit, and a mixture of Abs specific for CD3, CD8, CD11c, CD19, B220, TCRαβ, TCRγδ, Gr-1, NK1.1, and TER119 (lineage (lin) mixture). Before the intracellular staining with an anti-human/mouse RORγ-PE Ab, cells were fixed and permeabilized with an eBioscience kit (catalog no. 00-5523-00) following the manufacturers protocol. Data were analyzed with the FlowJo software (Tree Star). A FACSaria device (BD Biosciences) was used to sort adult LTi cells.
RORγt- and TATA binding protein-specific RT-PCRs were conducted as previously described (9).
In vitro culture and proliferation assay
FACS-sorted LTi cells from H-IL-7 spleen were cultured on primary splenic stroma (14) in IMDM/FCS supplemented with antibiotics (penicillin/streptomycin, Ciproxin, and kanamycin), 0.5 mg of insulin per 100 ml, 0.1% 2-ME, 1% nonessential amino acids without or with a blocking anti-IL-7Rα Ab (clone A7R34; 50 μg/ml). The presence and number of LTi cells was analyzed by flow cytometry after 8 days.
To determine whether LTi cells proliferate in the presence of IL-7, sorted LTi cells from RAG2−/− spleens were labeled with 1.25 μM CFSE (10 min at 37°C) and cultured on primary splenic stroma (14) with or without IL-7. After 7 days, cells were analyzed by flow cytometry. To test in vivo proliferation, 108 CD4+ MACS-enriched CFSE-labeled splenocytes were adoptively transferred into irradiated mice.
Immunofluorescence and immunohistochemistry
Tissues were snap frozen in Tissue-Tek OCT compound (Sakura). Sections (5–8 μm) were fixed in acetone and rehydrated in PBS. For intracellular staining, slides were blocked in PBS with 0.2% gelatin (porcine skin, type A; Sigma-Aldrich) and 1% donkey serum for 30 min and then incubated with primary Ab in PBS with 0.2% gelatin overnight at 4°C. For surface staining, tissue sections were blocked in PBS with 5% goat serum and 1% BSA and then incubated with primary Ab for 1 h at room temperature. Sections were washed and incubated with a secondary Ab for 45 min. For enzymatic immunohistochemistry, peroxidase activity was developed using 3,3′-diaminobenzidine tetrahydrochloride (Sigma-Aldrich). Slides were embedded in Mowiol or Medi-Mount (Medite).
Data were analyzed using Student’s t test and are shown as mean values and SDs.
Results and Discussion
Adult CD4+CD3− cells are LTi cells
We investigated the LTi activity and number of CD4+CD3− cells in normal and IL-7-overexpressing IL-7/CIITA double-transgenic (referred to as H-IL-7) adult mice. The percentage of CD4+CD3− cells was increased 5-fold (Fig. 1⇓A) and their number 30-fold in the spleens of H-IL-7 mice (Fig. 1⇓B). In addition, adult CD4+CD3− numbers were increased in LNs (Fig. 1⇓, C and D). IL-7 protein was present in VCAM-1+CD45− cells in the splenic red pulp and was substantially increased in H-IL-7 mice (supplemental Fig. S1, A and B).4 Altogether, our findings demonstrate that increased IL-7 availability was associated with increased CD4+CD3− cell numbers in adult mice.
H-IL-7 mice were used as a source for isolating and characterizing adult CD4+CD3− cells. Adult CD4+CD3− cells expressed IL-7Rα, CD18, c-Kit, β1 integrin, α4β7 integrin, and CD44 and were negative for CD62L (supplemental Fig. S2A). Furthermore, they expressed LTαβ, TRANCE (TNF-related activation-induced cytokine), and TRANCER (TRANCE receptor) (Fig. S2B). Transcripts for RORγt were found in adult CD4+CD3− cells (supplemental Fig. S2C). Hence, freshly isolated fetal and adult cells share a common phenotype in vivo (Table I⇓). Our data add new information to a previous study demonstrating that in vitro cultured fetal and adult CD4+CD3− cells display similar genetic fingerprints (15). In contrast to a previous report (16), we were unable to detect significant levels of OX40L and CD30L in adult CD4+CD3− cells (Table I⇓). This discrepancy may rely on the fact that in our study we investigated freshly isolated cells, whereas in other studies adult CD4+CD3− cells were cultured overnight before testing.
To investigate the function of adult CD4+CD3− cells, we isolated the cells by FACS sorting either from H-IL-7 or from RAG2−/− mice, thereby avoiding contamination with CD4+ T cells and adoptively transferred them into CXCR5−/− newborn mice. CXCR5−/− mice lack almost all PPs and isolated lymphoid follicles (ILFs) unless neonatally reconstituted with fetal LTi cells from WT origin (9, 18, 21). Thus, this assay was established to monitor the LTi activity of CD4+CD3− cells in the gut. Recipient mice were analyzed 3 wk after adoptive transfer by anti-VCAM-1 immunohistochemistry of the whole intestine. In this study, we show that in CXCR5−/− mice receiving adult CD4+CD3− cells from RAG2−/− mice, the number of VCAM-1 spots comprising both PPs and ILFs was increased 10-fold (Fig. 2⇓A). The reconstitution was in relative proportion to the number of transferred adult CD4+CD3− cells from H-IL-7 mice (Fig. 2⇓B). Total splenocytes (106) from WT animals were unable to restore lymphoid tissue formation in CXCR5−/− mice, suggesting that insufficient LTi cells were present among this number of cells. These data demonstrate for the first time that CD4+CD3− cells from adult mice are bona fide LTi cells. As compared with fetal LTi cells, LTi cells from adult H-IL-7 mice were ∼5-fold less efficient (18). In a previous study it was shown that adult CD4+CD3− cells help reorganize the splenic architecture after lymphocytic choriomeningitis virus infection (14). This effect appeared to be mediated by accelerated restoration of the splenic stromal cell compartment. In addition, a role for adult CD4+CD3− cells in creating organized T/B segregation in the adult spleen was proposed (13). Our study clearly shows that adult LTi cells can mediate the de novo generation of lymphoid tissue. In line with this, H-IL-7 mice developed not only ectopic LNs and additional PPs, but also formed tertiary lymphoid organs in nonlymphoid organs such as salivary glands (S. Schmutz and D. Finke, unpublished observations). Altogether, adult CD4+CD3− cells have lymphoid tissue-inducing activity that may contribute not only to lymphoid tissue organization but also to the neoformation of secondary and tertiary lymphoid organs.
Adult LTi cells are strictly dependent on IL-7 and RORγt
To test whether adult LTi cells and their precursors respond to IL-7, we treated WT and IL-7−/− mice with IL-7 or with a combination of IL-7 plus anti-IL-7 mAb. IL-7/anti-IL-7 complexes have been shown to display 50- to 100-fold higher biological activity than free IL-7 (17). In line with this, within 7 days the in vivo administration of IL-7/anti-IL-7 complexes significantly increased splenic LTi numbers in WT (Fig. 3⇓A) and IL-7−/− (Fig. 3⇓B) mice. IL-7 treatment alone had no significant effect. Under steady state conditions, splenic CD4+CD3− cell numbers in WT and IL-7−/− mice were comparable. Because IL-7−/− mice have no peripheral LNs, LTi cells mainly accumulated in the spleen. Our in vivo data prompted us to speculate that IL-7 increased the number of adult LTi cells through support of survival and/or proliferation. Indeed, survival of sorted splenic LTi cells cultured for 8 days was reduced by ∼50% in the presence of neutralizing an anti-IL-7R mAb (Fig. 3⇓C). IL-7 had a slight effect on the proliferation of LTi cells in vitro (Fig. 3⇓D) and after adoptive transfer into RAG2−/−γc−/− mice in vivo (Fig. 3⇓E). In RAG2−/−γc−/− mice, adoptively transferred LTi cells proliferated vigorously even when no IL-7/anti-IL-7 complexes were injected. It is likely that in RAG2−/−γc−/− mice, elevated levels of IL-7 and/or additional factors supported their in vivo proliferation, which was increased after treatment with IL-7/anti-IL-7 complexes. Overexpression of either IL-7 or bcl-2 did not rescue LTi numbers in the spleen of adult RORγ−/− mice, thus demonstrating their strict dependence on RORγt (Fig. 3⇓F). Additionally, we also demonstrated that the amount of IL-7 was able to regulate the pool of adult LTi cells.
Fetal LTi cells have been reported to originate from multilineage FL precursor cells (22, 23), whereas postnatally the BM is the major site of hematopoiesis. Although BM chimera experiments have given indirect evidence that the BM harbors cells or precursors that can help in organizing the spleen, clear data about the generation of peripheral CD4+CD3− LTi cells from the BM are still missing. To test whether IL-7 can promote the generation of adult CD4+CD3− LTi cells from BM cells, we transplanted BM from adult WT (Ly5.1) mice into irradiated IL-7Rα−/− (Ly5.2) recipients and treated the mice with an IL-7/anti-IL-7 complex or left them untreated. Before adoptive transfer, BM cells were depleted of CD4+ and CD8+ cells to avoid contamination with T and mature LTi cells. As controls, we injected FL cells from E13.5 (CD45.1) WT mice into irradiated IL-7Rα−/− (CD45.2) recipients. As expected, in control mice receiving FL cells a significant percentage of donor LTi cells (4.7% of lin− donor cells) was found in the spleen, whereas donor LTi cells from WT BM were almost undetectable (0.8% of lin− donor cells) (Fig. 4⇓A). IL-7/anti-IL-7 complex treatment, however, significantly enhanced the generation of LTi cells (2.7% of lin− donor cells) from the BM. Fetal LTi cells can be identified by the expression of RORγ (10). We found that after adoptive transfer of BM or FL cells (CD45.1) into RAG2−/− CD45.2 mice, donor-derived splenic CD4+CD45.1+lin− cells were RORγ+ (Fig. 4⇓B) and that the absolute cell numbers significantly increased after treatment with IL-7/anti-IL-7. We observed a striking increase in the number of ILFs in the small intestine of mice receiving adult BM or FL cells (Fig. 4⇓C). As in normal WT mice, the ILFs in the small intestine of BM-reconstituted mice were mainly composed of LTi cells and B cells of donor origin (supplemental Fig. 3 and data not shown). These data are in agreement with a previous study demonstrating that the reconstitution of LTi cell-deficient Rorc(γt)gfp/gfp mice with RORγt+ LTi cells induced the formation of ILFs (24). Our data show that functional LTi cells can be generated from adult BM and that exogenous IL-7 increases the frequency of BM-derived LTi cells. Mixed chimera experiments confirmed that FL cells were almost 10-fold more efficient at generating LTi cells as compared with BM cells (data not shown). This is likely a result of a stronger proliferation of fetal hematopoietic cells (25). As previously shown, administration of BrdU to pregnant mice labeled fetal LTi cells (9). Seven weeks after continuous treatment of pregnant mice (E12.5–E19.5) with BrdU, all LTi cells in the offspring had lost BrdU, whereas 20% of maternal LTi cells were BrdU+ (data not shown). Therefore, LTi cells in fetal and adult mice differ in their proliferative status.
Collectively our data demonstrate that bona fide LTi cells can be generated from adult BM, respond to IL-7, colonize the gut, and induce additional lymphoid follicles in the small intestine. The increased availability of IL-7 enlarges the adult LTi cell pool through survival and proliferation of pre-existing LTi cells and the de novo generation of LTi cells. In vivo, LTi cells develop from BM precursors relatively inefficiently unless exogenous IL-7 is provided. Considering the role of LTi cells in normal and ectopic lymphoid tissue development (9), adult LTi cells may play an important role in chronic inflammatory diseases such as rheumatoid arthritis, where local IL-7 availability is increased (26).
We thank J. Kirberg for providing RAG2−/− mice, D. Littman for RORγ−/− mice, A. Trumpp for bcl-2 transgenic mice, T. Schueler for an anti-IL-7 immunostaining protocol, Amgen Inc. for providing M25 mAb, and I. Seibert for technical help.
Except for O.B., who is a shareholder in Nascent Biologics Inc., the other authors have no conflicting financial interests.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵1 This work was supported by the Swiss National Science Foundation Grant PPOOA-68855, the Gottfried and Julia Bangerter-Rhyner Foundation, and a grant from Swiss Mobiliar (to D. F.). A.R. is holder of the Chair of Immunology endowed by Hoffman-La Roche Ltd., Basel, Switzerland.
↵2 Address correspondence and reprint requests to Prof. Daniela Finke, Department of Biomedicine, University of Basel, Mattenstrasse 28, CH-4058 Basel, Switzerland. E-mail address:
↵3 Abbreviations used in this paper: LT, lymphotoxin; BM, bone marrow; E, embryonic day; FL, fetal liver; lin, lineage; ILF, isolated lymphoid follicle; LN, lymph node; LTi, lymphoid tissue inducer; PP, Peyer’s patch; ROR, retinoic acid-related orphan receptor; WT, wild type; H-IL-7, high IL-7.
↵4 The online version of this article contains supplemental material.
- Received September 8, 2008.
- Accepted June 26, 2009.
- Copyright © 2009 by The American Association of Immunologists, Inc.