Increased lymphangiogenesis in joints of mice with inflammatory arthritis

Abstract

Angiogenesis is involved in the pathogenesis of inflammatory arthritis, but little is known about the role of lymphangiogenesis in this setting. Here, we examined whether tumor necrosis factor (TNF) stimulates osteoclast precursors (OCPs) to produce the lymphatic growth factor, vascular endothelial growth factor-C (VEGF-C), and induce lymphangiogenesis. We used TNF-transgenic (Tg) mice and mice with serum-induced arthritis. OCPs were purified by fluorescence-activated cell sorting of CD11b+/Gr-1-/lo blood or bone marrow cells and subjected to microarray analysis or were generated from spleen or joint cells and treated with TNF. Expression of VEGFs was analyzed and examined by real-time reverse transcription-polymerase chain reaction and Western blotting. Immunostaining and magnetic resonance imaging were used to quantify lymphatic vessels and volumes of synovium and draining lymph nodes. TNF stimulated VEGF-C expression by OCPs and increased nuclear factor-kappa B (NF-kappaB) binding to an NF-kappaB sequence in the VEGF-C promoter. OCPs from joints of TNF-Tg mice express high levels of VEGF-C. Lymphatic vessel numbers and size were markedly increased in joint sections of TNF-Tg mice and mice with serum-induced arthritis. The severity of synovitis correlated with draining lymph node size. In summary, TNF induces OCPs to produce VEGF-C through NF-kappaB, leading to significantly increased lymphangiogenesis in joints of arthritic mice. The lymphatic system may play an important role in the pathogenesis of inflammatory arthritis.

Figure 1

Differential expression of vascular endothelial growth factor (VEGF) and platelet-derived growth factor (PDGF) family genes in CD11b⁺/Gr-1^-/lo osteoclast precursors (OCPs) from tumor necrosis factor-transgenic (TNF-Tg) and wild-type (WT) mice. CD11b⁺/Gr-1^-/lo cells from peripheral blood and bone marrow from TNF-Tg (7 mice per array) and WT (23 mice per array) mice were pooled and purified by flow sorting. The RNA samples were subjected to microarray analysis using the GeneChip mouse genome 430A 2.0 array from Affymetrix. The array data on angiogenic gene expression were analyzed using GeneTraffic software. The expression ratio was calculated by dividing the mean value of the intensity of RNA signals from two separate arrays of TNF-Tg OCPs by that from WT cells.

Figure 2

Tumor necrosis factor (TNF) increases expression of vascular endothelial growth factor-C (VEGF-C) by osteoclast precursors (OCPs). Wild-type (WT) spleen cells were cultured with macrophage colony-stimulating factor (M-CSF) for 3 days to enrich for OCPs. OCPs were cultured in 10% serum with M-CSF and treated with phosphate-buffered saline (PBS) or TNF (10 ng/mL). (a) Expression of VEGF-C and β-actin mRNA was measured by real-time reverse transcription-polymerase chain reaction (RT-PCR) at various times (left panel), and protein levels were measured by Western blot analysis (right panels). (b) Expression of VEGF-A, -B, -C, -D, and placental growth factor (PLGF) mRNA was examined in samples treated for 24 hours. The fold increases in TNF-treated over PBS-treated cells at a given time were calculated. Values are the means of triplicate loadings plus standard deviation. The effect of different doses of TNF (c) or a combination with interleukin 1 (IL-1) (TNF and IL-1 10 ng/mL) (d) on VEGF-C expression at 24 hours was examined by RT-PCR. Experiments were repeated twice with similar results.

Figure 3

Tumor necrosis factor (TNF) promotes the binding of nuclear proteins to the nuclear factor-kappa B (NF-κB) binding sequences of the vascular endothelial growth factor-C (VEGF-C) promoter. Raw264.7 osteoclast/macrophage precursors were cultured in 0.5% bovine serum albumin overnight. Cells were treated with TNF for 30 to 60 minutes. Nuclear extracts were isolated and subjected to Western blot analysis using anti-NF-κB p65 and p50 antibodies (a) or to electrophoretic mobility shift assay using a ³²P-labeled probe consisting of the putative NF-κB binding sequence of the mouse VEGF-C promoter (b). The specificity of binding was confirmed by using 50-fold more unlabeled wild-type (WT) or mutated probe in which the putative NF-κB binding sequence was mutated and could not be bound by NF-κB in a competition reaction. An SP-1 probe was used as a loading control. WT osteoclast precursors were treated with TNF ± an NF-κB inhibitor for 24 hours, and expression of VEGF-C was determined by real-time reverse transcription-polymerase chain reaction. Values are the means of three mice plus standard deviation. (c) The fold decrease in expression in the NF-κB inhibitor-treated over vehicle-treated cells was calculated. Experiments were repeated two times with similar results.

Figure 4

Cells from joints of tumor necrosis factor-transgenic (TNF-Tg) mice express high levels of vascular endothelial growth factor-C (VEGF-C). Ankle and wrist joints of TNF-Tg mice and wild-type (WT) littermates were subjected to collagenase digestion to isolate cells, which then were stained with fluorescein isothiocyanate (FITC)-anti-CD11b and phycoerythrin-anti-Gr-1 and subjected to fluorescence-activated cell sorting analysis. (a) The percentage of CD11b⁺/Gr-1^-/lo cells is shown in a representative histogram from one pair of TNF-Tg and WT mice. Cells (3 × 10⁵) were cytospun onto glass slides and double-stained with FITC-anti-CD11b and rabbit anti-VEGF-C followed by anti-rabbit Alexa 546. TO-PRO-3 iodide was used as a DNA dye for nuclear staining. (b) Co-localization of CD11b and VEGF-C proteins was observed, and pictures were taken under a confocal microscope at a power of × 20. Cells were cultured with M-CSF for 3 days, and adherent cells were harvested. (c) Expressions of VEGF-A, -C, and -D mRNA were measured by real-time reverse transcription-polymerase chain reaction (RT-PCR). The fold increases in expression in TNF-Tg over WT cells were calculated. Values are the means of triplicate loadings plus standard deviation (SD). (d) The expression levels of TNF and IL-1 mRNA in joint extracts from TNF-Tg mice and WT littermates were determined by real-time RT-PCR. Values are the means of three mice plus SD. Experiments were repeated using four additional pairs of TNF-Tg and WT mice with similar results.

Figure 5

Enlargement of lymphatic vessels in the joints of tumor necrosis factor-transgenic (TNF-Tg) mice. Joint sections from TNF-Tg mice or wild-type (WT) littermates were immunostained with anti-LYVE-1 or CD31 antibody. (a) Micrographs (× 20) show increased numbers and diameters of LYVE-1⁺ lymphatic vessels in the synovium of TNF-Tg mice. (b) The area and number of lymphatic vessels within the synovium were counted. Values are the means plus standard deviation of five TNF-Tg mice and six WT mice. *p < 0.05 versus WT samples. LYVE-1, lymphatic endothelial hyaluronan receptor 1.

Figure 6

Increased volume of popliteal lymph nodes in tumor necrosis factor-transgenic (TNF-Tg) mice. (a) A representative post-contrast magnetic resonance imaging slice from a TNF-Tg and wild-type (WT) mouse (5 months old). (b) Using a semi-automated segmentation procedure in Amira software, the synovial (yellow) and lymph node (red) volumes were quantified and visualized in TNF-Tg and WT mice. *p < 0.05 versus WT samples (n = 5). (c) Representative hematoxylin-and-eosin sections (× 4) demonstrate differences in lymph node size between TNF-Tg and WT animals.

Figure 7

Increased lymphangiogenesis in joints of mice with serum-induced arthritis. Wild-type mice received serum from K/B × N mice to induce arthritis and were sacrificed at 0, 14, and 35 days after serum injection (n = 3 or 4 mice at each time point). Ankle joint sections were immunostained with anti-LYVE-1 antibody. (a) Representative pictures show inflamed pannus and large numbers of LYVE-1⁺ lymphatic vessels in an adjacent section (pink arrows) at day 35. (b) The area and number of lymphatic vessels within the pannus were determined by histomorphometric analysis. Values are the means plus standard deviation of 3 or 4 mice at each time point. H&E, hematoxylin and eosin; LYVE-1, lymphatic endothelial hyaluronan receptor 1.