IL-7 production in murine lymphatic endothelial cells and induction in the setting of peripheral lymphopenia

Abstract

IL-7 is a required factor for T-cell homeostasis. Because of low expression levels and poor reagent availability, the cellular sources of IL-7 have proven challenging to characterize. In this study, we describe a reporter mouse in which enhanced GFP is expressed from the endogenous Il7 locus. We show that IL-7 is produced by lymphatic endothelial cells (LECs) distributed throughout the systemic lymphatic vasculature as well as by fibroblastic reticular cells, and that phosphorylation of STAT5 in lymphocytes is higher in lymphatics than in blood. Furthermore, in nodes depleted of lymphocytes, Il7 transcription is increased in stromal but not in myeloid subsets. These data support recent findings that lymphocyte homeostasis is influenced by access to secondary lymphoid organs and point to LECs as an important in vivo source of IL-7, bathing trafficking immune cells under both resting and lymphopenic conditions.

Keywords: HIV; endothelium; interleukin-7; lymph node; lymphatic; lymphopenia; myeloid; stromal.

Fig. 1.

Il7-eGFP reporter mice are correctly targeted and report Il7 transcription with fidelity. (A) Knock-in targeting strategy for generating Il7-eGFP reporter mice. S, SacI; K, KpnI; B, BamHI. eGFP, eGFP coding sequence; NeoR, neomycin-resistance gene cassette. (B) Southern blot analysis of tail genomic DNA digested with the restriction endonuclease SacI. The WT allele is expected to yield a 5.6-kb fragment, whereas the final targeted, NeoR-deleted allele is expected to yield a 6.6-kb fragment. (C) Flow cytometry analysis of single-cell suspensions prepared from WT or Il7 ^eGFP/eGFP BM and stained with antibodies against B220, CD43, HSA and BP1. Fractions A–C, Hardy B-cell developmental fractions: pre-pro B cells (Hardy A) and pro B cells (Hardy B). (D) Absolute lymphocyte counts in whole blood collected from WT, Il7 ^+/eGFP or Il7 ^eGFP/eGFP mice stained with antibodies against CD3, CD4 and CD8. Percentage is compared with WT. (E) Quantitative RT–PCR analysis of whole-tissue RNA collected from lymphoid (LN, inguinal LNs Spl, spleen; BM, bone marrow; Thy, thymus) or non-lymphoid tissues (gut; lung; liver; SM, skeletal muscle). Ratio of eGFP signal to Il7 signal (eGFP :Il7) is indicated. Data are representative of at least two independent experiments with at least five mice per group.

Fig. 2.

LECs are an important source of IL-7 within the resting LN. (A and B) Confocal immunofluorescence microscopy of perfusion-fixed inguinal LNs from Il7 ^+/eGFP or WT control mice. Data are representative of two separate experiments with three mice per group. Low-magnification scale bars represent 200 μm and high-magnification scale bars represent 20 μm. (A) eGFP (green) is localized to the subcapsular sinus and medullary region and co-localizes with LYVE1 (red). Open arrow in high-magnification overlay points to eGFP⁺LYVE1⁺ double-positive LEC and closed arrow points to eGFP⁻ LEC. eGFP staining in WT node is shown in the bottom panel. (B) Subcapsular and medullary eGFP (green) does not co-localize with the myeloid marker CD11b (red) in optical slices from the confocal stack. Open arrows point to eGFP⁺ cells that are non-overlapping with CD11b⁺ cells (closed arrows). (C and D) Analysis of pooled single-cell suspensions prepared by enzymatic digestion of LNs from Il7-eGFP heterozygote or WT littermate control mice. Stromal subsets enriched by CD45⁺ cell depletion were stained for CD45, gp38 and CD31. (C) eGFP signal is localized to the FRC and LEC subsets. (D) Quantitative RT–PCR analysis of Il7, gp38 and LYVE1 transcripts on flow-sorted FRCs, LECs and BECs from WT mice. Fold difference in Il7 levels between FRCs (CD31⁻gp38⁺) and LECs (CD31⁺gp38⁺) is shown. NS, not significant; P > 0.05. Data are representative of at least three independent experiments with at least three mice per group.

Fig. 3.

Systemic lymphatics draining non-lymphoid tissues also express eGFP reporter protein. Immunofluorescence analysis to assess the expression of eGFP reporter protein in perfusion-fixed diaphragm and lung from Il7 ^eGFP/eGFP mice. (A) Immunofluorescence microscopy for eGFP (green) and LYVE1 (red) in whole-mount central tendon of diaphragm. Filled arrow points to initial lymphatic and open arrow points to medium-sized collecting lymphatic. Left three panels are from Il7 ^eGFP/eGFP mice and right panel is from WT control. Scale bar represents 50 μm. (B and C) Immunofluorescence microscopy for eGFP in 200-μm lung cryosections. Magnification ×100. Scale bars represent 20 μm. (B) Lymphatic vasculature of Il7 ^eGFP/eGFP. (C) Lung parenchyma of Il7 ^eGFP/eGFP (left panel) and WT (right panel) control. Filled arrow points to a parenchymal cell. Data are representative of three experiments with three mice per group.

Fig. 4.

LECs distributed throughout peripheral tissues express important amounts of IL-7. (A) Flow cytometry analysis of pooled single-cell suspensions prepared by enzymatic digestion of lungs from Il7 ^+/eGFP heterozygote mice. Stromal cells enriched by CD45⁺ cell depletion were stained for gp38, EpCAM and CD31 and analyzed. (B) Quantitative RT–PCR analysis of flow-sorted cells derived from lungs of Il7-eGFP heterozygotes. Stromal cells enriched by CD45⁺ cell depletion were stained for CD45, gp38 and CD31 (LN) or gp38, EpCAM and CD31 (lung) and sorted by FACS into four populations: LN FRCs (gp38⁺CD31⁻), LN LECs (gp38⁺CD31⁺), lung Type II AECs (SSC^hiCD31⁻gp38^-EpCAM⁺; lung AECII) or lung LECs (SSC^loCD31⁺gp38⁺ EpCAM⁻). Purified RNA from each population was analyzed by quantitative RT–PCR. Data are from two groups of five mice and are representative of two separate experiments. *P < 0.05. (C and D) Flow analysis of pSTAT5 in T cells collected from whole blood, LNs or draining lymph of WT mice. Single-cell suspensions were stained for CD3 and intracellular pSTAT5. (C) Concatenated pSTAT5 histograms from five animals illustrating differences between compartments. (D) Median intracellular pSTAT5 fluorescence level (with isotype control subtracted) for each of the five animals in panel (C) is shown. ***P < 0.001. (E) Flow analysis of surface CD127 levels on T cells collected from whole blood or draining lymph. Cells were stained for CD3, CD4, CD8 and CD127 and median fluorescence intensity of CD127 was compared between compartments with a paired t-test. *P < 0.05.

Fig. 5.

Blockade of lymphocyte ingress into LNs induces stromal up-regulation of Il7 mRNA. (A) Whole-tissue RT–PCR analysis of Il7 transcript in tissues harvested from Il7 ^+/eGFP mice. Signal was normalized to Hprt. Data are representative of two separate experiments with five mice in each group. The ratio of Il7 transcript from WT to that from Il7 ^+/eGFP heterozygotes is shown for inguinal lymph nodes (LN), spleen (Spl), BM, thymus (Thy), gut, lung, liver and skeletal muscle (SM). ND, not detected. (B) Absolute lymphocyte count in whole blood or enzymatically dispersed LNs of WT mice after treatment with either anti-α4/αL antibody (α4/αL) or isotype control antibody (Ctrl). Percent depletion is shown for LN subsets. (C–E) Analysis of single-cell suspensions from enzymatically dispersed LNs of WT mice stained for CD45, gp38, CD11b and CD11c. (C) Flow analysis and sorting gates. (D) Quantitative RT–PCR analysis of Il7 transcript in sorted populations from LNs harvested from isotype control-treated (Ctrl) or α4/αL antibody (α4/αL) -treated (22 days) mice. CD45⁻ stromal (CD45⁻gp38⁺) and myeloid (CD11c⁺ DCs or CD11c⁻CD11b⁺ macrophages) cells were analyzed. Data are representative of at least two experiments with at least three mice per group. (E) Quantitative RT–PCR analysis of Il7 transcript in flow-sorted CD45⁻ stromal (CD45⁻gp38⁺) cells over time. Animals were treated for 10, 16 or 22 days with anti-α4/αL blocking antibodies (α4/αL) or isotype control antibody (Ctrl). *P < 0.05, ***P < 0.001.

Fig. 6.

Hematopoietic cells make a minor contribution to LN IL-7. (A–D) Bone marrow chimeras of WT BM > Il7 ^eGFP/eGFP recipients (n = 4, black) or Il7 ^eGFP/eGFP BM > WT recipients (n = 10, green). Cells derived from WT mice were CD45.1⁺, whereas those from Il7 ^eGFP/eGFP mice were CD45.2⁺. (A and B) Absolute lymphocyte counts in whole blood (A) and LNs (B) measured by staining with antibodies for CD3, CD4 and CD8. Percent depletion is shown in (B) over antibody-blocked (α4/αL) animals. Ctrl, non-antibody depleted; α4/αL, antibody-blocked. (C) Efficient seeding of SLOs with myeloid cells originating from transferred BM. Enzymatically dispersed mLNs or splenic cell suspensions were stained with CD11c, CD11b and CD45.2 and plotted. The black histogram shows reconstitution in a single representative WT BM > Il7 ^eGFP/eGFP chimera and the green histogram is representative of reconstitution in a single Il7 ^eGFP/eGFP BM > WT chimera. (D) Quantitative RT–PCR analysis of whole-tissue RNA collected from pLNs, mLNs or spleen of chimeric animals without (Ctrl) and with antibody blockade (blocked) for 22 days. Fold difference in Il7 transcript is shown in some cases. ***P < 0.001. (E) Immunofluorescence analysis of perfusion-fixed 20-μm cryosections from mLNs harvested from chimeric animals following antibody blockade (22 days): eGFP (green) and CD11b (red). Data are representative of at least four animals. M, medullary region; T, T cell area. Scale bar represents 100 μm.