Hematopoietic lymphoid tissue inducer (LTi) cells initiate lymph node (LN) and Peyer’s patch (PP) development during fetal life by inducing the differentiation of mesenchymal organizer cells. The growth factor signals underlying LTi cell development and LN and PP organogenesis remain poorly understood. LTi cells express the Il7r and the receptor tyrosine kinase Kit, whereas organizer cells express their cognate ligands. To determine the relative significance of Il7 and Kit signaling in LTi cell homeostasis and PP and LN development, we have analyzed mice deficient for Kit (KitW/Wv), Il7 (Il7−/−), or both (Il7−/− KitW/Wv). Unlike KitW/Wv and Il7−/− single mutants, Il7−/− KitW/Wv mice were almost devoid of LTi cells in their mesenteric LN anlage. This LTi deficiency was associated with a block in mesenchymal LN organizer cell generation and the absence of almost all LNs. In contrast, intestinal LTi cell numbers, PP organizer cell generation, and PP development were strongly affected by impaired Kit signaling, but were independent of Il7. Hence, Kit and Il7 act synergistically in LN organogenesis, whereas Kit signaling, but not Il7, critically regulates PP organogenesis and LTi cell numbers in the intestine. Consistent with these differential growth factor requirements for PP and LN development, PP organizer cells expressed higher Kitl and lower Il7 levels than did LN organizer cells. Collectively, these results demonstrate that Kit and Il7 differentially control PP and LN organogenesis through the local growth factor-driven regulation of LTi cell numbers.
Lymph nodes (LNs) and Peyer’s patches (PPs) play a central role in adaptive immunity by providing specialized microenvironments for the activation and expansion of rare Ag-specific lymphocytes. In the mouse, LN and PP organogenesis is initiated by hematopoietic lymphoid tissue inducer (LTi) cells from fetal day 12.5 onward (1). Differentiating from Il7r-expressing fetal liver (FL) progenitors (2, 3), LTi cells are among the first hematopoietic cells to colonize the fetal intestine and LN anlagen (1, 4), where they are identified as CD45+ Thy1.2+ CD4+ Il7R+ (4, 5). LTi cells are instrumental for lymphoid organ development, because mice lacking LTi cells lack PPs and LNs (6, 7). LTi cells express high levels of lymphotoxin (LT)-α1β2 (5) and thereby provide LT signals, which are required for LN and PP development during fetal life (8).
After having colonized PP and LN anlage, LTi cells trigger the differentiation of stromal organizer cells by delivering LT signals to the surrounding mesenchyma (1, 4, 9). PP and LN organizer cells express LTβR and high levels of the adhesion molecules ICAM-1 and VCAM-1 (1, 10, 11). Several LT-mediated positive feedback loops regulate and enhance the formation of LTi and organizer cell clusters. For example, upon LTβR engagement organizer cells upregulate the expression of the cytokines Il7 and TRANCE (12), which in turn increases LT expression on the LTi cell surface (1, 11). LTβR engagement also upregulates the expression of the chemokines CXCL13, CCL19, and CCL21 (13, 14), which attract more LTi cells to the anlage (11, 13). These developing lymphoid anlagen are colonized by mature lymphocytes after birth (13, 15), later developing into compartmentalized lymphoid organs.
Several pathways underlying LN and PP organogenesis have been identified, which interestingly do not affect LN and PP development equally. The TRANCE/RANK/TRAF6 pathway is necessary for LN development, but is not essential for PP formation (1, 16–18). The Il7/Il7r/Jak3/Il2rγ pathway regulates the organogenesis of several peripheral LNs (19–22), whereas PP development is completely abolished in mice deficient for Il7r, Il2rγ, and Jak3, but not in Il7–deficient mice (23, 24).
Il7 was proposed to regulate LN development by controlling the size of the LTi cell pool within the LN anlage (25). Il7−/− newborn mice, however, have a reduced but detectable LTi cell population in the developing LN (25), implying that other cytokines contribute to LTi cell homeostasis. Furthermore, Il7−/− mice have normal PP anlagen development (24), suggesting that the intestinal LTi cell population might be regulated by Il7-independent factors.
The growth factor Kit ligand (Kitl) (26) is expressed by organizer cells (27), and its receptor, Kit, is expressed by both LTi cells and their FL precursors (3, 4). This finding raises the question of whether the Kit/Kitl axis plays a role in LTi cell homeostasis and in PP and LN development. The Kit/Kitl signaling pathway has pleiotropic roles in adult and fetal life, because it is involved in hematopoiesis, gametogenesis, melanogenesis, and intestinal tract motility (28–31), but its potential role in LTi cell generation and PP and LN development remains to be determined. In this study, we analyzed KitW/Wv mice, which carry the lethal KitW allele encoding a Kit protein lacking the transmembrane domain together with the viable KitWv allele encoding for a receptor with an impaired tyrosine kinase activity (32). Hence, we have addressed LTi cell homeostasis and LN and PP development in mice with strongly reduced cell surface expression of Kit and impaired Kit signaling. Moreover, we generated and analyzed compound mutants for Kit and Il7 (Il7−/− KitW/Wv).
We show that Kit and Il7 signaling synergize in positively regulating LTi cell numbers in LN anlage and in promoting LN organogenesis. In contrast in the intestine, Kit but not Il7 signals regulate the size of the LTi cell pool and PP organogenesis. Finally, Kitl and Il7 are expressed in inverse ratios by PP and LN organizer cells, and a locally differential production of these growth factors may contribute to the specific regulation of PP and LN development.
Materials and Methods
All mice were bred and maintained in our animal facility under specific pathogen-free conditions. The animal experiments received the approval of the Cantonal Veterinary Office of the city of Basel, Switzerland. C57BL/6 mice were purchased from RCC (Itingen, Switzerland). Il7−/− (33), KitW/+, and KitWv/+ mice were described previously (30). The origin of the KitW/+ and KitWv/+ strains used in this study has been described (34).
LN enumeration was performed a week after peritoneal injection of 100 μl of 1% Chicago sky Blue 6B (Sigma-Aldrich, St. Louis, MO) ink in PBS. Images of mesenteric LNs (mLNs) were captured with stereoscopic Nikon SMZ1500 microscope coupled with a DS Camera control Unit DS-L1 (Nikon, Melville, NY).
Flow cytometry and cell sorting
FITC, PE, PE-Cy7, allophycocyanin, or biotin-conjugated anti-CD4 (GK1.5), anti-CD8α (53-6.7), anti-CD19 (1D3), anti-NK1.1 (PK136), anti-α4β7 (DATK32), and anti–ICAM-1 (3E2) Abs were purchased from BD Biosciences (San Jose, CA). anti-CD3 (145-2C11), anti-B220 (RA3-6B2), and anti-Ter119 (Ter-119) Abs were purchased from Biolegend (San Diego, CA). Anti-CD11c (N418), anti-CD45 (30-F11), anti-Kit (2B8), anti–Il7r (A7R34), anti–VCAM-1 (429), anti-NKp46 (29A1.4), and anti–Gr-1 (RB6-8C5) Abs were purchased from e-Bioscience (San Diego, CA). RORγ staining was performed with anti-RORγ Ab (AFKJS-9) and fixation/permeabilization kit (Foxp3 staining buffer kit) from e-Bioscience. As secondary reagent, streptavidin-PE and streptavidin-allophycocyanin (Biolegend) were used. Flow cytometric acquisition was performed with a FACSCalibur (BD Biosciences), and data were analyzed using FlowJo software (Tree Star, Ashland, OR). For LTi cell number quantification, mLNs from individual 0.5-d-old mice were homogenized with glass slides, and individual guts were digested with 0.5 mg/ml dispase (Life Technologies, Rockville, MD) together with 100 μg/ml DNase I (AppliChem, Darmstadt, Germany) in PBS at 37°C for 10 min under mild agitation. Cell suspensions were then filtered, and immunostaining was performed on ice. Absolute LTi cell numbers were obtained from analyzing whole organs.
For organizer cell staining, mLNs and gut samples of 0.5-d-old mice were digested with 1 mg/ml dispase and 100 μg/ml DNase I in PBS at 37°C for 25–30 min under mild agitation. LT staining was performed as previously described (20) with fusion LTβR-Fc protein (35), which was a gift from J. Browning (Biogen, Cambridge, MA). Cells were treated with anti-Fc 2.4G2 Ab and 0.5% mouse and rat serum. LTβR-Fc (35) was added and detected using biotin-goat anti-human IgG (Jackson ImmunoResearch Laboratories, West Grove, PA) pretreated for 30 min with 4% rat and mouse serum. Finally, fluorochrome-labeled streptavidin and Ab for lineage (lin; CD3, CD11c, CD19, B220, NK1.1, Ter119, Gr-1), CD4, and Thy1.2 were added.
Cell sorting was done using a FACS Aria (BD Biosciences), and reanalysis of sorted cells showed that they were >98% pure. LTi cells were FACS-sorted from 0.5-d-old mLN suspension as lin− Thy1.2+ CD4+ cells. CD45− VCAM-1high ICAM-1high organizer cells were sorted from dispase-digested mLNs and gut samples of 17.5-d-postcoitum (dpc) embryos.
FL suspensions from 14.5-dpc embryos were stained with PE-conjugated anti-α4β7 Ab, washed once, and incubated with anti-PE MACS beads (Miltenyi Biotec, Auburn, CA). Labeled cells were magnetically purified with MACS separation columns (Miltenyi Biotec), following the manufacturer’s instructions. Positively selected cells were seeded at 3 × 105 cells per well of a 48-well plate (Nunc, Naperville, IL) in supplemented IMDM alone or in the presence of combinations of 20 ng/ml Il7 (PeproTech, Rocky Hill, NJ) and 100 ng/ml Kitl (Peprotech). Ex vivo FACS-sorted LTi cells were cultured in a round-bottom 96-well plate (Falcon BD, Franklin Lakes, NJ) in the presence of various combinations of Il7 (20 ng/ml) and Kitl (100 ng/ml).
Quantitative real-time PCR
RNA extraction was performed with the Nucleospin RNA XS kit (Macherey-Nagel, Düren, Germany). RNA quantification and quality assessment were performed on Experion (Bio-Rad, Hercules, CA); 20 ng RNA was used to perform the reverse transcription with Oligo dT (Promega) and dNTPs (Roche, Mannheim, Germany) with the Superscript III Reverse Transcriptase (Invitrogen, Carlsbad, CA). Real-time PCR was performed with Sensimix (Quantace, Watford, U.K.) on a Rotor Gene RG-3000 (Corbett Research, Sydney, Australia). The following primers were used: TATA box binding protein (Tbp) FWD, 5′-CGTGAATCTTGGCTGTAAACT-3′; Tbp RVS, 5′-GTCCGTGGCTCTCTTATTCT-3′; Kitl FWD, 5′-TCAACATTAGGTCCCGAGAA-3′; Kitl RVS, 5′-CTTTGCGGCTTTCCTATTACT-3′; Il7 FWD, 5′-GATAGTAATTGCCCGAATAATGAACCA-3′; Il7 RVS, 5′-GTTTGTGTGCCTTGTGATACTGTTAG-3′. Tbp, Kitl, and Il7 primer pairs had identical efficiency. The cycling conditions were 10 min at 95°C, followed by 40 cycles of 10 s at 95°C, 15 s at 60°C, and 20 s at 72°C. The relative expression of Kitl and Il7 on Tbp was calculated with the comparative CT (DDCT) method.
Statistical significance between experimental groups was established with the Mann-Whitney U test with Prism software (GraphPad, San Diego, CA).
Kitl synergizes with Il7 for LTi cell generation and maintenance in vitro
LTi cell FL precursors were identified as lin− α4β7+ Il7r+ progenitors (3). To test the capacity of Kitl in promoting LTi cell generation from FL precursors, α4β7+ cells were purified from 14.5-dpc FL and cultured with or without Kitl and/or Il7. Cells with an LTi CD4+ Il7r+ phenotype were readily detectable in 5-d cultures (Fig. 1A). These cells expressed RORγ and LTα1β2, but not NKp46 (Fig. 1B), demonstrating that they were bona fide LTi cells. The addition of Kitl to cultures resulted in a 2-fold increased LTi cell generation (Fig. 1C). Because we have shown previously that Il7 promotes the generation of LTi cells from FL precursors in vitro (21), we tested whether the combined action of Kitl and Il7 could further promote LTi cell generation. Compared with Il7 alone, the addition of Kitl further increased the number of LTi cells 2-fold (Fig. 1C). These results demonstrate that Kitl alone promoted LTi cell generation from α4β7+ FL precursors and that it synergized with Il7 in generating LTi cells in vitro.
To assess whether Kitl would promote LTi cell maintenance in vitro, 3500 FACS-sorted LTi cells from mLNs of 0.5-d-old wild type (WT) mice were cultured with recombinant Kitl and recombinant Il7. After 4 d of culture, FACS analysis revealed that most LTi cells maintained the lin− Thy1.2+ CD4+ phenotype (Fig. 1D). In addition, regardless of Kitl, ∼70% of the number of cells initially plated was recovered from cultures without Il7 (Fig. 1E). This finding indicates that Kitl does not sustain LTi cell maintenance in vitro. In contrast, cell recovery was 2-fold the input when Il7 was added to the cultures and up to 4-fold in the presence of both Kitl and Il7 (Fig. 1E). Similar results were obtained with CD4− LTi cells (S. Chappaz and D. Finke, unpublished data). These results show that Kitl and Il7 synergize in expanding LTi cells in vitro.
Il7 was shown to induce the upregulation of LT on the surface of LTi cells in vitro (11). To test whether Kitl might influence LT expression alone or in combination with Il7, we measured LT expression levels after 6 h cytokine stimulation of FACS-sorted LTi cells. As expected (11), Il7 clearly increased LT expression, whereas Kitl did not substantially alter LT levels alone or in combination with Il7 (Fig. 1F).
Kit and Il7 signals synergize in positively regulating mesenteric LTi cell numbers and LN organogenesis
Our in vitro results prompted us to study whether Kit has a role in the regulation of mesenteric LTi cell numbers in vivo. Therefore, LTi cell populations were quantified in the mesentery of newborn WT, KitW/Wv, Il7−/−, and Il7−/− KitW/Wv mice (Fig. 2A). LTi cell numbers in KitW/Wv mice were decreased 1.5-fold and in Il7−/− mice were decreased 3-fold when compared with WT controls. Il7−/− KitW/Wv mice, however, had a 14-fold decrease in LTi cell number compared with WT controls (Fig. 2B), indicating that the Kit and Il7 signaling pathways synergize in regulating the size of the LTi cell pool in mLN anlagen.
Because LTi cells are instrumental for organizer cell generation and LN development (9), the disruption of the Kit and Il7 signaling pathways might indirectly affect the generation of mesenchymal organizer cells. We therefore analyzed the mesentery of newborn mice of each genotype for the presence of organizer cells. Newborn WT and KitW/Wv mice had a comparable percentage of organizer cells (Fig. 2C), indicating that the modest decrease in LTi cell number in KitW/Wv mice did not prevent normal organizer cell generation in the mLN anlage. Accordingly, mLNs developed normally in KitW/Wv mice (Fig. 2D) and displayed a normal chain-like anatomy (Fig. 2E). Seven of 10 KitW/Wv mice, however, lacked sacral LNs, showing that the Kit signaling pathway had a nonredundant function in the development of sacral LNs (Fig. 2D). The mLN organizer cell population was strongly reduced in Il7−/− mice (Fig. 2C). Although mLNs were present in all adult Il7−/− mice (Fig. 2D), they had an abnormal anatomy, consisting of two separated LNs (Fig. 2E). As reported previously (25), several LNs were missing in Il7−/− mice. Organizer cells were absent in the mesentery of newborn Il7−/− KitW/Wv mice (Fig. 2C), suggesting that LTi cell numbers in these mice were too low to sustain organizer cell generation. In fact, mLNs and the vast majority of peripheral LNs were absent in adult Il7−/− KitW/Wv mice (Fig. 2D). These results show that the Kit signaling pathway regulates LTi cell numbers in mLN anlage and is required for the normal development of the sacral LN. Furthermore, these results show that the synergism between the Il7- and Kitl-driven signals crucially regulates LTi cell numbers in the mLN, the generation of organizer cells, and the development of LNs.
Kit signaling regulates intestinal LTi cell numbers and PP organogenesis
We addressed whether the Kit signaling pathway can control the size of the intestinal LTi cell pool and regulate PP organogenesis. Intestinal LTi cells were identified as CD45int Il7r+ (Fig. 3A). These cells homogeneously expressed RORγ, confirming they were bona fide LTi cells (S. Chappaz and D. Finke, unpublished data). The gut of KitW/Wv mice contained 3.5-fold fewer LTi cells than did the gut of WT littermates (Fig. 3B). In line with the fact that Il7 is dispensable for PP anlagen development (24), Il7−/− mice had similar LTi cell numbers in the gut compared with WT animals. In contrast to what was observed in the mLN, newborn Il7−/− KitW/Wv mice displayed similar LTi cell numbers as KitW/Wv controls in the gut (Fig. 3B). These results show that the Kit signaling pathway regulates the size of the intestinal LTi cell pool independently of Il7. Organizer cells were present at similar percentages in WT and Il7−/− mice, but they were severely decreased in KitW/Wv and Il7−/− KitW/Wv mice (Fig. 3C). Consistent with previously published work (24), whole mount immunostaining for VCAM-1 revealed that WT and Il7−/− gut developed approximately eight VCAM-1+ PP anlagen. In contrast, gut from KitW/Wv and Il7−/− KitW/Wv mice had approximately three anlagen (Fig. 3D). In line with this finding, adult KitW/Wv mice had a 2-fold decrease in PP number compared with WT littermates (Fig. 3E). These results show that the number of intestinal LTi cells, the generation of organizer cells, and the formation of PP anlagen are dependent to a large degree on Kitl but not on Il7.
Kitl and Il7 are differentially expressed by intestinal and mLN organizer cells
The observation that Il7 and Kit signaling differentially controlled LTi cell numbers in the intestine and in the mLN suggests that Il7 and Kitl may be differentially expressed in these organs. To test this, we FACS-sorted organizer cells from 17.5-dpc mLN and intestine. Real-time PCR analysis revealed that gut organizer cells contained ∼16-fold more Kitl and 7-fold fewer Il7 transcripts than did mLN organizer cells (Fig. 4A, 4B). These results suggest that the differential effect of Kit and Il7 signaling on LTi cell numbers in the gut and the mLN results from the differential expression of Kitl and Il7 by the corresponding organizer cells.
Whereas Kitl and Il7 are known to play both unique and synergistic roles in lymphocyte development (36), their combined effects on LNs and PPs has not been determined. The gut and the mLN of newborn KitW/Wv mice contained 3.5- and 1.5-fold, respectively, lower numbers compared with WT. Kit signaling can therefore regulate the LTi cell number in vivo. This finding was consistent with the ability of Kitl to promote LTi cell generation from FL progenitors in vitro (Fig. 1A–C). In line with the fact that PP anlagen developed in Il7−/− mice (24), we show that Il7 did not regulate the size of the intestinal LTi cell pool. In contrast, Il7 plays a central role in regulating LTi cell numbers in the LN anlage (25), but the pathways accounting for LTi cells persisting in Il7−/− mLNs remained unknown. The results presented in this study identify the Kit signaling axis as one of these pathways. Indeed, LTi cells were almost absent from the mLNs of Il7−/− KitW/Wv mice, which is in line with the fact that Kitl and Il7 synergized both in the generation and growth of LTi cells in vitro. These results show that the Kit signaling pathway controls the size of the intestinal LTi cell pool independent of Il7, whereas both Kitl and Il7 positively regulate LTi cell numbers in the mLN anlage.
Because LTi cells are required for organizer cell generation (9), the size of the LTi cell pool may directly affect the mesenchymal compartment within LN and PP anlage (25). In fact, the 3.5-fold decreased intestinal LTi cell number in KitW/Wv mice was associated with a reduced number of intestinal organizer cells, VCAM-1+ anlagen, and PPs. Moreover, the 14-fold reduction in LTi cell number in the mLN of Il7−/− KitW/Wv mice was associated with the loss of mLN organizer cells and almost all LNs, which supports the proposal that insufficient LTi cell numbers preclude organizer cell differentiation (25).
The signaling requirements for PP and LN organogenesis are distinct. For example, TRANCE is required for LN organogenesis, but not PP development (17, 24), and is expressed at higher levels in mLN organizer cells than in their intestinal counterpart (10, 27). Furthermore, gene expression profile studies indicate that PP and mLN organizer cells are distinct mesenchymal populations (27), supporting the idea that organizer cell diversity accounts for the differential signaling requirement for PP and LN development. We show in this study that the intestinal LTi cell pool was strongly dependent on Kit but independent of Il7. Conversely, the mesenteric pools primarily depended on Il7 and to a lesser extent on Kit. These results show that the LTi cell pools resident in the gut and the mLN are regulated independently; they further suggest that this is due to the different availability of Kitl and Il7 in these organs. Indeed, PP organizer cells express higher levels of Kitl but lower levels of Il7 mRNA when compared with mLN organizer cells [our results and Okuda et al. (27)]. Hence, differential availability of these growth factors may provide an explanation as to why LTi cell pools and lymphoid organogenesis in the gut and the mLN are differentially affected by the disruption of the Kit and Il7 genes.
In contrast to Il7−/− and Kitw/wv mice, PP anlagen development is completely abolished in Il7r−/−, Il2rγ−/− and Jak3−/− mice (24). This finding suggests that an alternative Il7r ligand regulates PP organogenesis. Whereas thymic stromal lymphopoietin (Tslp) is a ligand for Il7r (37–39), compelling evidences argue against a crucial role in PP development. First, the Tslp receptor does not require Jak3 and Il2rγ for signaling (37–39). Second, Crlf2 (Tslpr)−/− mice develop normal PPs (40, 41). The identification of a putative third Il7r ligand will substantially improve our knowledge on the regulation of PP development.
In summary, Kit and Il7 collaborate in LN organogenesis, whereas only Kit has an essential role in PP development. Differential expression of Kitl and Il7 by LN and PP organizer cells may contribute to the observed distinct growth-factor dependence and provide regional control on LTi cell numbers. Local regulation of the LTi cell pool size may be required to ensure a stringent control on secondary lymphoid organ development, as suggested by the fact that systemic expression of an Il7 transgene leads to the enlargement of the LTi cell pool and to the subsequent formation of ectopic LN and supernumerary PP (21).
We thank Antonius Rolink and Jason Gill for helpful discussions and comments on the manuscript and Sabine Eckervogt for animal care.
Disclosures The authors have no financial conflicts of interest.
This work was supported by the Swiss National Science Foundation Grant PP00A-116894/1, the Pfizer Forschungspreis Foundation, the Jubiläumsstiftung der Schweizerischen Mobiliar, and the Julia Bangerter Rhyner Foundation (to D.F.).