Regulation of peripheral lymph node genesis by the tumor necrosis factor family member TRANCE

Abstract

Proper lymph node (LN) development requires tumor necrosis factor-related activation-induced cytokine (TRANCE) expression. Here we demonstrate that the defective LN development in TRANCE(-/)- mice correlates with a significant reduction in lymphotoxin (LT)alphabeta(+)alpha(4)beta(7)(+)CD45(+)CD4(+)CD3(-) cells and their failure to form clusters in rudimentary mesenteric LNs. Transgenic TRANCE overexpression in TRANCE(-/)- mice results in selective restoration of this cell population into clusters, and results in full LN development. Transgenic TRANCE-mediated restoration of LN development requires LTalphabeta expression on CD45(+) CD4(+)CD3(-) cells, as LNs could not be induced in LTalpha(-/)- mice. LTalpha(-/)- mice also showed defects in the fate of CD45(+)CD4(+)CD3(-) cells similar to TRANCE(-/)- mice. Thus, we propose that both TRANCE and LTalphabeta regulate the colonization and cluster formation by CD45(+) CD4(+)CD3(-) cells in developing LNs, the degree of which appears to correlate with the state of LN organogenesis.

Figure 1

Generation of TRANCE^−/− mice: a gene dosage effect on TRANCE expression. (A) Reverse transcription PCR analysis of thymocytes or activated CD8⁺ T cells from WT, TRANCE^+/−, and TRANCE^−/− mice. TRANCE-R mRNA expression was also examined and HPRT served as an internal control. (B) Flow cytometric analysis of activated CD8⁺ T cells from WT, TRANCE^+/−, and TRANCE^−/− mice. Cells were stained with hIgG₁ (gray lines) or TRANCE-R-IgG₁ (thick lines) followed by PE-coupled goat anti–human IgG₁.

Figure 2

Altered primary B cell follicles but apparently normal formation of GCs in spleen of TRANCE^−/− mice. (A) Immunofluorescence photomicrographs of representative spleen sections from WT, TRANCE^+/−, and TRANCE^−/− mice. Spleens were dissected from 3–4-wk-old mice and stained with anti-B220 (FITC) and anti-CD4 (PE). (B) Representative immunofluorescence photomicrographs of spleen sections prepared 10 d after immunization of mice with 50 μg NP₁₂-KLH adsorbed to alum. Immunofluorescence microscopy was performed to detect formation of GCs (PNA-biotin/streptavidin–PE and anti-IgM–FITC).

Figure 3

TRANCE is essential for proper development of PLNs. (A) Evan's blue dye was injected into all footpads and images of the PLNs were obtained. Shown are MLNs, iliac, and sacral (Sac.) LNs. Apart from the CLNs (see panel B), 100% of TRANCE^−/− mice lacked all PLNs and MLNs. (B) The presence of CLNs in TRANCE^−/− mice. A representative image of CLNs found in TRANCE^−/− mice (left). A representative immunofluorescence analysis of CLNs obtained from TRANCE^−/− with anti-CD4–PE and anti-B220–FITC (right). (C) Haploinsufficiency effect of TRANCE on the development of PLNs. A representative image of TRANCE^+/− mice lacking one of the inguinal (Ing.) LNs.

Figure 4

CD45⁺CD4⁺CD3⁻ cells in day 0 MLNs express TRANCE and TRANCE-R. Cells obtained from WT day 0 MLNs were stained with anti-CD4, anti-CD45, and TRANCE-R-IgG₁ (for TRANCE), 1E6.66 (for anti–TRANCE-R), LT-βR-IgG₁ (for surface LTαβ), or AF.H6 (for anti–LT-βR) (thick lines). Cells were isolated by mechanical disruption of day 0 MLNs. Similar results were also obtained when cells were isolated by collagenase digestion of day 0 MLNs (data not shown). For detection of surface TRANCE, cells obtained from day 0 MLNs were cultured in medium for 1–2 h before staining with TRANCE-R-IgG1. Gray histograms are negative controls. For TRANCE-R-IgG₁ and LT-βR-IgG₁, incubation with human IgG₁ was performed as a negative control. For anti–TRANCE-R mAb (1E6.6) and anti–LT-βR Ab (AF.H6), rat IgG and hamster anti-KLH (Ha/48) were used as negative controls, respectively.

Figure 5

Reduction of colonization by CD45⁺CD4⁺CD3⁻ cells and their failure to form clusters in TRANCE^−/− or LTα^−/− mice. (A) A representative image of MLNs from newborn WT mice and of rMLNs found in newborn TRANCE^−/− mice. Dotted circles indicate MLNs or rMLNs isolated. Similar rMLNs were observed in newborn LTα^−/− mice (data not shown). Lymphatic vessels are white, presumably due to the chyle being drained from intestines. (B) A representative flow cytometric analysis of cells taken from day 0 MLNs from WT mice or rMLNs from TRANCE^−/− and LTα^−/− mice. (C) The percentage of CD45⁺ cells showing the CD4⁺CD3⁻ phenotype in day 0 MLNs from WT mice or rMLNs from TRANCE^−/− and LTα^−/− mice. (D) Representative immunofluorescence analysis of frozen sections obtained from day 0 intestines from WT or TRANCE^−/− mice. Sections were stained with anti-CD4–Alexa-488 and anti–MAdCAM-1–PE.

Figure 6

Both TRANCE and LT-α are required for cluster formation by CD45⁺CD4⁺CD3⁻ cells and for the formation of MAdCAM-1⁺ endothelial cell network within developing LNs. Representative images taken from whole mount staining of freshly isolated MLNs from newborn WT or TRANCE⁽−^/−^,Tg) mice, and rMLNs from TRANCE^−/− or LTα^−/− mice. The whole MLNs or rMLNs were stained with anti-CD4–FITC and anti–MAdCAM-1–PE.

Figure 7

Ectopic expression of TRANCE restores LN development in TRANCE^−/− mice. (A) Transgenic TRANCE is overexpressed in T and B cells, but not in CD4⁺CD3⁻ cells found in day 0 MLNs. Freshly isolated cells were incubated with the indicated Abs and examined by FACS^® analysis. (B) Representative images of PLNs (Iliac) and MLNs observed in TRANCE⁽−^/−^,Tg) mice. 4 h after injection of Evan's blue dye, LNs were imaged. Note that one of the two iliac LNs and sacral LNs were still missing in the TRANCE⁽−^/−^,Tg) mouse shown here. In addition, MLNs in these mice were smaller than those in WT mice. (C) Representative immunofluorescence analysis of LNs from WT, TRANCE^(+/+,Tg), and TRANCE⁽−^/−^,Tg) mice. Sections were stained with anti-CD4–PE and anti-B220–FITC. Note that the righthand image in TRANCE⁽−^/−^,Tg) shows disrupted primary B cell follicle. (D) The percentage of CD45⁺ cells showing the CD4⁺CD3⁻ phenotype in day 0 MLNs from TRANCE^(+/+,Tg) and TRANCE⁽−^/−^,Tg) mice. *For comparison, the percentages of CD45⁺ cells showing the CD4⁺CD3⁻ phenotype in day 0 MLNs from WT or rMLNs from TRANCE^−/− mice are included, which were taken from Fig. 5 C. (E) The percentage of LNs restored in TRANCE⁽−^/−^,Tg) mice. *Although all of the TRANCE⁽−^/−^,Tg) mice contained MLNs, the number of MLNs averaged two, compared with four to six MLNs found in WT mice. Mand., mandibular; Super., superficial; Cerv., cervical; Axi., axial; Brach., brachial; Mesen., mesentery; Ingui., inguinal; Lumb., lumbar; Sac, sacral; Pop., popliteal.