Oncostatin M Enhances CCL21 Expression by Microvascular Endothelial Cells and Increases the Efficiency of Dendritic Cell Trafficking to Lymph Nodes

Animals, cell lines, and reagents

Female C57BL/6 mice and BALB/C mice (6–8 wk old) were purchased from The Jackson Laboratory. The National Cancer Institute Animal Care and Use Committee approved all experiments involving animals. Human DMEC and human dermal fibroblasts (HDF) were purchased from the Endothelial Cell Core Facility (Department of Dermatology, Emory University School of Medicine, Atlanta, GA) and cultured as described previously (25). Human lung microvascular EC (LMEC) were purchased from Cambrex and cultured in EGM-2 MV (Cambrex) in tissue-culture-treated plastic flasks (Corning). Recombinant mouse and human OSM, TNF-α, IL-1β, IFN-γ, and IL-6 were obtained from R&D Systems. U0126 and PD98059 were purchased from Cell Signaling Technology and were dissolved in DMSO before use. CCL19-human IgG fusion protein (CCL19-Ig) was a gift from Dr. U. von Andrian (Harvard Medical School, Boston, MA). TC2031-2, a human melanoma cell line, was a gift from Dr. J. Wunderlich (Surgery Branch, National Cancer Institute, Bethesda, MD). The Jurkat T cell leukemia line was described previously (26). The H9 T cell lymphoma line is a variant of the HUT 78 cell line and was obtained from the American Type Culture Collection.

Generation of bone marrow-derived DC (BMDC)

BMDC were prepared as described previously (27). Briefly, femurs and tibia were obtained from 8- to 12-wk-old mice, soaked in alcohol for 4 min, and rinsed with PBS three times. Bone marrow was flushed from the bones, filtered through a 45-μm filter (Falcon), and washed with PBS. Cells were treated with Ack lysis buffer for 2 min, washed two times, and maintained in culture for 7–8 days in complete RPMI 1640 containing 5% FBS that was supplemented with murine (m)GM-CSF (10 ng/ml) and mIL-4 (10 ng/ml) (PeproTech). On the last day, LPS (Sigma-Aldrich) 200 ng/ml was added to the culture to further activate DC. Cells were washed five times in PBS, and gradient centrifugation was performed using 14% Nycodenz (Sigma-Aldrich) at 1700 rpm for 10 min at 4°C to isolate DC.

Determination of OSM expression by BMDC using quantitative real-time RT-PCR and by ELISA

BMDC were prepared as described above except that no LPS was added into the culture before harvest. BMDC were isolated by gradient centrifugation and washed twice in PBS. BMDC were then seeded into a round-bottom 96-well plate at 5 × 10⁵ cells/well with medium only, 20 ng/ml mTNF-α (PeproTech), or 200 ng/ml LPS for 24 h. Cell culture supernatants were collected for ELISA, and cells were harvested for quantitative real-time RT-PCR. Primers for mOSM and GAPDH were as follows: OSM for (forward), 5′-GCA GCT GTG GCT TTC TCT GG-3′ and rev (reverse), 5′-TCG TCC CAT TCC CTG AAG AC-3′; GAPDH for, 5′-CGT GTT CCT ACC CCC AAT GT-3′ and rev, 5′-TGT CAT CAT ACT TGG CAG GTT TCT-3′.

For mOSM ELISA, 1 μg/ml anti-mouse OSM (R&D Systems) was used to coat a 96-well ELISA plate at 4°C overnight. The plate was blocked with PBS containing 10% FBS (blocking buffer) at room temperature (RT) for 1 h. The plate was washed three times with PBS containing 0.05% Tween 20 (washing buffer) between each step. Recombinant murine (rm)OSM (R&D Systems) as standards and undiluted culture supernatants from BMDC were added to the plate and incubated at RT for 2 h. After washing three times, 100 μl of biotinylated anti-mouse OSM (1 μg/ml; R&D Systems) were added to wells at RT for 1 h. Lastly, streptavidin-HRP (1/1000 dilution) was added to wells at RT for 30 min before developing with ABTS substrate (Roche). For human OSM ELISA, a commercial ELISA kit (R&D Systems) was used according to the manufacturer’s instructions.

In vitro OSM stimulation

DMEC or LMEC at passages 3–6 were cultured in EGM-2 MV with 5% serum (except as noted in the text) in the absence of gelatin or fibronectin substrate and stimulated for 24, 48, or 72 h with recombinant human (rh) OSM (100 ng/ml) or other cytokines as indicated. Total RNA was isolated using RNeasy (Qiagen). cDNA was synthesized using Superscript II (Invitrogen Life Technologies). Quantitative RT-PCR was performed as described previously (25). Primers for human CCL21 were as follows: for, 5′-GGT TCT GGC CTT TGG CAT C-3′; rev, 5′-AGG CAA CAG TCC TGA GCC C-3′. Primers for human GAPDH were as follows: for, 3′-ACC CAC TCC TCC ACC TTT GA-3′ and rev, 5′-CAT ACC AGG AAA TGA GCT TGA CAA-3′.

Western blotting and flow cytometry

EC were washed with PBS, and whole cell lysates were harvested on ice with a cell scraper in lysis buffer (1% Nonidet P-40, 0.5% sodium deoxycholate, and 0.1% SDS) containing proteinase mixture (Roche) and phosphatase inhibitors (1 mM PMSF, 10 mM NaF, and 1 mM Na3VO4). After centrifugation, whole cell lysates were separated by 14% SDS-PAGE and transferred to polyvinylidene difluoride membranes (Millipore). Membranes were blocked and soaked with a 1/1000 dilution of primary Ab at 4°C overnight. Membranes were then washed and incubated with 1:1000 of HRP-labeled secondary Abs for 1 h. After washing, membranes were incubated with working solution (SuperSignal West Pico Substrate; Pierce Scientific) for 5 min at RT and exposed to film. The Abs used in this study were goat anti-human CCL21 polyclonal Ab (R&D Systems), mouse anti-rabbit GAPDH mAb (Research Diagnostics), rabbit anti-ERK1/2 polyclonal Ab, and rabbit anti-phospho-ERK1/2 polyclonal Ab.

For flow cytometry analysis, microvascular EC were trypsinized, washed, resuspended in staining buffer (0.1% BSA/PBS), and then incubated with Abs against LYVE-1 for 45 min at 4°C (12).

Endothelial monolayer permeability assay

Permeability assays were performed using a Chemicon Vascular Permeability Kit (Chemicon International). DMEC and LMEC were grown on collagen-coated membrane inserts until confluence. For short-term permeability assays, upper chamber medium was replaced with serum-free medium containing 1 μM VEGF, 100 ng/ml human OSM (hOSM), or no treatment, together with 200 μl of FITC-dextran (50 μg/ml diluted from 2 mg/ml stock; Chemicon International) at 37°C. Medium from the lower chamber, which included any flow-through FITC-dextran, were removed at 15 min, 30 min, 1 h, and 2 h, and assayed using a CytoFluor fluorometer (Applied Biosystems). For long-term permeability assays, DMEC (grown on 1% collagen-coated, filter-based chambers) were treated overnight with OSM, other cytokines as indicated, or medium alone. FITC-dextran (per kit instructions) was placed in the upper chamber and allowed to diffuse for 1 h at 37°C before collection of buffer from the lower chamber for analysis.

In vivo permeability assay

mOSM (1 μg) or PBS was injected into the left hind footpads of C57BL/6J mice (n = 7/group). Twenty-four hours later, Evans blue dye (50 μl, 5 mg/ml; Sigma-Aldrich) was injected into the same footpads. Mice were sacrificed by cervical dislocation 1 min after dye injection, and the left popliteal LN were quickly harvested. Evans blue dye was then extracted from the draining LN by formamide (50 μl/LN) at 55°C for 2 h (28). Following measurement of OD at 630 nm in a spectrophotometer, the concentration of Evans blue dye (μg/ml) was calculated based on a standard curve of known amounts of Evans blue dye.

Chemotaxis assay

DMEC were grown in 24-well cell culture plates (Costar Transwell; Corning) and treated with rhOSM (100 ng/ml) for 48 h. Twenty minutes before the chemotaxis assay, wells were treated with 20 μg/ml neutralizing anti-CCL21 Ab (goat IgG; R&D Systems), control IgG, or medium alone. Calcein (Molecular Probes)-labeled BMDC (3 × 10⁵) were placed in Transwell inserts (Corning) and allowed to migrate toward the DMEC for 4 h at 37°C. Cells were collected and counted by flow cytometry (FACSCalibur; BD Biosciences). rhCCL21 (500 ng/ml in fresh complete RPMI 1640) was used as a positive control.

Cytokine injections and application of trinitrochlorobenzene (TNCB) hapten

Anesthetized mice were injected in ear skin with PBS, 100 ng of rmOSM, or 50 ng of mTNF-α (R&D Systems) in 20 μl of sterile PBS. TNCB (1% final concentration; Sigma-Aldrich) was dissolved in olive oil:acetone (1:4 volume:volume ratio) and then applied on the dorsal and ventral sides of the mouse ear (10 μl/side) (29). Mice were euthanized 24 h later, and total RNA was isolated from the ear. Briefly, ears were soaked in ethanol for 5 min, rinsed, and split, and then incubated (dermal-side down) in serum-free DMEM containing Liberase (150 μg/ml; Roche) for 45 min at 37°C. Ears were then washed and minced, and total RNA was extracted using RNeasy (Qiagen). Quantitative RT-PCR was performed using the following primer pairs: mCCL21 for, 5′-ATCCCGGCAATCCTGTTCTT-3′ and rev, 5′-GCCTTCCTCAGGGTTTGCA-3′; mCCR7 for, 5′-GGACACGCTGAGATGCTCACT-3′ and rev, 5′-CCATCTGGGCCACTTGGAT-3′; and mIL-8 for, 5′-ACCACACTGCGCCAACACAGAAAT-3′ and rev, 5′-TCC AGACAGAGCTCTCTTCCATCA-3′.

In vivo migration assay

BMDC (day 7 after culture) were treated with LPS for 24 h and then extensively washed before being resuspended in PBS. Mouse footpads were injected with PBS, 1 μg of rmOSM, or 500 ng of rmTNF-α 24 h before injection of 10⁶ calcein-labeled BMDC into the footpads. After another 24 h, mice were euthanized, and the popliteal LN were collected, minced, and treated with 1 mg/ml collagenase D (Roche) and 0.05% DNase (Sigma-Aldrich) for 45 min at 37°C. Cells were then filtered, washed, and resuspended in 0.1% BSA/PBS. The highly fluorescent calcein-labeled cells were identified in the FITC channel and counted by flow cytometry.

Statistical calculations

Statistical calculations were performed using Student’s t test using the GraphPad Prism analysis software (GraphPad). Values of p < 0.05 were considered statistically significant.

OSM increases CCL21 mRNA and protein expression in vitro

To determine the effects of OSM treatment on EC in vitro, we treated DMEC rhOSM over several days and measured mRNA levels of CCL21 by real-time quantitative RT-PCR (Fig. 1⇓A) and ELISA (Fig. 1⇓B). The DMEC in our studies were devoid of contaminating cell populations as suggested by uniform expression of CD31 (12). Furthermore, DMEC exhibited 40–70% expression (12) of LYVE-1, a LEC-specific marker (30). Human LMEC were previously shown to be devoid of LYVE-1 expression (12) and were deemed representative of BVEC. In the presence of 5% serum, up-regulation of CCL21 mRNA was detected in DMEC at 24 as well as 48 h after addition of OSM, but not other cytokines (IL-1β, IL-6, IFN-γ, and TNF-α) (Fig. 1⇓A). Under serum-free conditions, up-regulation of CCL21 mRNA was observed at 48 h (but not 24 h) in both LMEC (9.8-fold vs baseline) and DMEC (5.3-fold vs baseline) after treatment with OSM, but not TNF-α (data not shown). We also confirmed that OSM stimulated CCL21 protein synthesis and secretion into cell culture supernatants by DMEC within 48–72 h after stimulation (Fig. 1⇓B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

DMEC, but not other selected cell types, up-regulate protein and mRNA for CCL21 upon OSM stimulation. A, Following stimulation of DMEC in 5% serum for 24 or 48 h with IL-1β (10 ng/ml), IFN-γ (250 U/ml), TNF-α (10 ng/ml), IL-6 (10 ng/ml), or rhOSM (100 ng/ml), CCL21 mRNA expression was quantified by real-time RT-PCR. B, DMEC were stimulated with rhOSM at the indicated concentrations for 24, 48, and 72 h, after which supernatants were collected for ELISA-based measurement of CCL21 concentration. Values shown represent means and SDs of four separate experiments. C, HDF, DMEC, TC-2031-2, Jurkat, and H9 cells were treated with 100 ng/ml rhOSM (or left untreated) for 48 h. Total RNA was extracted, and quantitative RT-PCR was performed to measure CCL21 expression relative to GAPDH.

TNF-α, which has been reported to stimulate CCL21 expression in vivo by LEC following intradermal injection (10), did not stimulate CCL21 mRNA expression in either DMEC or LMEC after 24 h or 48 h of exposure at 10 ng/ml in the presence of serum (Fig. 1⇑A). Because of greatly increased death when EC were exposed to TNF-α at higher concentrations (100 ng/ml) for >24 h, assessment of CCL21 mRNA expression under these specific conditions was not possible (data not shown). In contrast to its effect on CCL21 expression, OSM did not up-regulate expression of IL-8 after either 24 or 48 h of treatment, whereas TNF-α increased expression of IL-8 in DMEC and LMEC by ∼150-fold after 24 h, suggesting that TNF-α was indeed active (data not shown). Lastly, whereas OSM stimulated CCL21 expression by DMEC and LMEC, it did not stimulate CCL21 expression in human melanoma cells (TC2031-2), T cell leukemia cells (Jurkat), lymphoblastic T cell lymphoma (H9) cells, or primary HDF, suggesting that OSM stimulated CCL21 mRNA in limited cell types (Fig. 1⇑C). Thus, rhOSM treatment specifically increased transcriptional and translational expression of CCL21 in human microvascular EC.

OSM induces phosphorylation of ERK1/2 in DMEC, which is required for expression of CCL21

OSM has previously been shown to stimulate MAPK signaling in a variety of cells (31, 32). To determine whether this pathway was involved in expression of CCL21 in OSM-stimulated EC, we used phosphorylation of ERK1/2 as a sensitive indicator of MAPK activation. Using DMEC, we found no detectable ERK1/2 phosphorylation in unstimulated cells (Fig. 2⇓A). By contrast, incubation with 100 ng/ml OSM for 30 min markedly induced ERK1/2 phosphorylation in these cells (Fig. 2⇓A). U0126, a MAPK inhibitor, blocked phosphorylation of ERK1/2 in DMEC in a dose-dependent manner (0.05–5 mM). Because CCL21 production (Fig. 2⇓B) appeared to be more sensitive to U0126 than did ERK1/2 phosphorylation (Fig. 2⇓A), CCL21 production may be dependent on other signaling pathways (besides ERK1/2) that are inhibited by U0126. Similar results were obtained using PD98059, another MAPK inhibitor (data not shown). Interestingly, both U0126 and PD98059 also blocked CCL21 production in cell-free culture supernatants of DMEC stimulated with OSM (Fig. 2⇓B and data not shown), suggesting that CCL21 protein expression in EC is, in part, dependent upon MAPK signaling.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

OSM induces phosphorylation of ERK1/2 in DMEC, which is required for expression of CCL21. A, Induction of ERK1/2 phosphorylation in DMEC stimulated for 30 min with 100 ng/ml OSM. Phosphorylation of ERK1/2 is blocked in a dose-dependent manner by U0126, a specific inhibitor of MAPK signaling. B, CCL21 protein in cell-free culture supernatants of DMEC stimulated for 24 h with OSM is also blocked by U0126 in a dose-dependent manner, implicating MAPK signaling as critical for CCL21 expression.

Conditioned medium from OSM-stimulated DMEC mediates CCL21-dependent chemotaxis of BMDC

To determine whether the increased CCL21 protein produced by DMEC could stimulate migration of CCR7-expressing BMDC, we performed a chemotaxis assay with mouse BMDC and conditioned medium from OSM-treated DMEC. Preliminary experiments indicated that rh CCL21 stimulated murine BMDC in chemotaxis assays with similar potency and efficacy as did rm CCL21 (data not shown). CCR7 surface expression was confirmed by flow cytometry in the BMDC, by binding the CCL19-Ig chimeric protein to BMDC (data not shown), and by >10-fold increased migration to recombinant CCL21 in chemotaxis assays (data not shown). Baseline migration of BMDC toward untreated DMEC could be inhibited by ∼50% with neutralizing anti-CCL21 Abs (Fig. 3⇓), suggesting that chemotactically active amounts of CCL21 were being produced by DMEC in the absence of exogenous OSM. Treatment of DMEC with OSM for 48 h resulted in enhanced migration of BMDC that was similar to that achieved by the addition of exogenous CCL21 to culture medium from nonstimulated DMEC (Fig. 3⇓). The addition of anti-CCL21, but not control IgG, Abs to cell supernatants from OSM-treated DMEC resulted in a ∼60% reduction in BMDC migration (Fig. 3⇓), suggesting that OSM treatment functionally increases the capacity of DMEC to stimulate CCL21-dependent chemotaxis.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Migration of BMDC toward OSM-treated DMEC. DMEC were grown to confluence and were either untreated or treated with rhOSM as noted in Materials and Methods. Certain wells were treated with either control IgG or neutralizing Ab against CCL21. BMDC were allowed to migrate toward the EC monolayers for 4 h at 37°C. As positive control, BMDC were allowed to migrate toward 500 ng/ml rhCCL21. Results are representative of three experiments with similar findings.

OSM treatment increases CCL21 mRNA expression in skin and draining LN in vivo

To assess the effects of OSM on CCL21 production in vivo, we injected the ears of C57BL/6 mice with PBS, 100 ng of rmOSM, or 50 ng of rmTNF-α, and extracted total skin RNA 24 h later (Fig. 4⇓A). Both rmOSM and rmTNF-α treatments increased mRNA expression of CCL21 to a similar degree, but whereas TNF-α up-regulated CCR7 mRNA expression, rmOSM treatment did not. Peripheral LN are also a rich source of expression of CCL21 (6). To determine whether OSM could enhance CCL21 expression in the LN, we injected mouse footpads with OSM and then measured CCL21 and CCR7 mRNA expression in the draining popliteal LN. OSM did not change the size of the draining LN compared with PBS treatment (data not shown). Notably, CCL21 mRNA was up-regulated by ∼10-fold in the draining LN, whereas CCR7 expression remained unchanged (Fig. 4⇓B). Therefore, both skin and skin draining LN show strong up-regulation of CCL21 upon OSM treatment, possibly contributing to enhanced DC migration into lymphatic vessels and retention, respectively, of DC in the LN.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Up-regulation of CCL21 in vivo in response to OSM treatment. Ears of C57BL/6 mice were given intradermal injections of PBS, OSM (100 ng), or TNF-α (50 ng) in a total volume of 20 μl of PBS. A, Total RNA was extracted from ears after 24 h, and quantitative RT-PCR was performed to measure CCL21 and CCR7 mRNA expression relative to GAPDH. Results are average of 2–3 mice per group and are representative of three experiments with similar findings. B, Mouse footpads were injected with PBS or 1 μg of rmOSM. Draining popliteal LN (n = 5 per treatment group) were harvested 24 h later and pooled for RNA isolation and real-time quantitative RT-PCR assessment of the indicated genes. Results are expressed relative to GAPDH and are representative of three experiments with similar findings.

OSM enhances migration of BMDC to draining LN in vivo

Based on the data above, we predicted that intradermal injection of skin with OSM would result in enhanced DC migration from peripheral tissues (i.e., skin) to draining LN in vivo, presumably by increasing expression of CCL21 by LEC. To test this hypothesis, mouse footpads were pretreated with PBS, OSM, or TNF-α before injection of BMDC. After 24 h, popliteal LN were collected, and migratory DC within LN were quantified by flow cytometry. Pretreatment with TNF-α or OSM increased BMDC migration to popliteal LN by ∼3- or 2-fold, respectively, compared with PBS (Fig. 5⇓), indicating that both cytokines were capable of enhancing BMDC migration to regional LN.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

In vivo migration of BMDC in response to OSM treatment. Calcein-labeled BMDC (1 × 10⁶) were injected into the footpads of C57BL/6 mice that had been pretreated with PBS, 1 μg of OSM, or 500 ng of TNF-α 24 h before BMDC injection. Migrated BMDC were quantified 24 h later by flow cytometry. n = 8 per group. Results are representative of two experiments with similar findings.

Others have observed that BMDC produce factors that facilitate their own migration (10). To determine whether OSM could be one of these factors, we analyzed BMDC for the expression of OSM by both RT-PCR (for mRNA) and ELISA (for protein). OSM mRNA was readily detected in BMDC in contrast to B16 melanoma cells (Fig. 6⇓A). Moreover, OSM protein was significantly elevated in the culture medium of BMDC after 24 h in vitro, which was further enhanced by treating BMDC with LPS, but not TNF-α (Fig. 6⇓B). These results were consistent with the increased transcription of OSM mRNA by BMDC following LPS treatment (Fig. 6⇓A). Thus, activated DC may be a source for OSM under inflammatory conditions.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Expression of OSM by BMDC (in vitro) and in skin (in vivo). BMDC were cultured in round-bottom 96-well plates with medium only, TNF-α (20 ng/ml), or LPS (200 ng/ml) for 24 h. A, Total RNA was extracted from BMDC, and quantitative RT-PCR was performed to measure mOSM expression relative to GAPDH. B16 melanoma cell RNA was collected as a negative control for mRNA expression of OSM. B, Cell culture supernatants were collected for ELISA to determine OSM concentration. C, Ears were injected with TNF-α as described above (see Fig. 4⇑ legend) or treated with 1% TNCB in olive oil:acetone (D) (see Materials and Methods) and assessed by RT-PCR for expression of OSM mRNA. Results are representative of three experiments with similar findings.

Topical application of a contact-sensitizing hapten increases OSM expression in skin

To determine whether TNF-α could directly stimulate OSM expression in vivo, we injected TNF-α into mouse ear skin and assessed OSM expression 24 h later. In contrast to its ability to induce CCL21 and CCR7 mRNA (Fig. 4⇑A), TNF-α alone did not result in enhanced expression of OSM mRNA (Fig. 6⇑C). TNF-α had no effect on OSM expression under either of these conditions or with lower doses of TNF-α at early time points (data not shown). This prompted us to ask whether topical application of a contact-sensitizing hapten may affect the expression of OSM. Under these conditions, skin DC would be expected to carry hapten-bearing Ag from the skin to LN via afferent lymphatic vessels, thus sensitizing mice to future antigenic challenge (33). One application of the commonly used contact dermatitis-sensitizing hapten, trinitrochlorobenzene (34), in olive oil:acetone vehicle resulted in a marked increase in OSM expression (Fig. 6⇑D). These results suggest that TNF-α does not directly stimulate OSM production. The up-regulation of OSM after application of the hapten, however, suggests that OSM is up-regulated in skin under conditions in which DC are known to become activated and to later traffic to draining LN.

OSM increases EC monolayer permeability

OSM may also increase DC migration by increasing lymphatic flow and, thus, facilitating the passive transport of DC via the lymphatic vessels to regional LN. In a manner analogous to VEGF (previously known as vascular permeability factor; Ref. 35), OSM may increase the permeability of BVEC, resulting in the leakage of fluid from the intravascular space into the tissue space. This extracellular fluid must eventually be reabsorbed by the lymphatics, leading transiently to increased lymphatic flow that may carry along DC and other leukocytes to the draining LN.

To determine whether OSM altered vascular permeability, DMEC and LMEC were treated with 100 ng/ml rhOSM, 1 μM rhVEGF, or medium alone at 37°C. Permeability of EC monolayers to FITC-dextran was measured at 30 min, 1 h, and 2 h. As expected based on other studies (36), exposure of LMEC to VEGF resulted in significantly increased permeability to FITC-dextran (37, 38) as early as 30 min after treatment (Fig. 7⇓A). Of note, OSM clearly increased permeability of LMEC monolayers beginning as early as 30 min following treatment, nearly to the same extent as VEGF, one of the most potent vascular permeability-enhancing cytokines known (35). In contrast, DMEC showed far less responsiveness to VEGF and OSM compared with LMEC (Fig. 7⇓A). DMEC that were treated with rhOSM for 18 h, however, showed a >10-fold increase in permeability compared with untreated EC (Fig. 7⇓B) without compromising viability (data not shown). TNF-α and IL-6 had no apparent effect on permeability. Of note, addition of hOSM to the lower chamber did not result in enhanced permeability of FITC-dextran (Fig. 7⇓B), suggesting that receptors for OSM were located on the apical membrane of the DMEC. Thus, exposure of DMEC and LMEC to OSM results in increased permeability of EC monolayers to FITC-dextran.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Permeability changes in DMEC and LMEC in response to OSM. A and B, DMEC and LMEC were cultured to confluence on cell membrane inserts and treated with 1 μM VEGF, 100 ng/ml OSM, or medium alone at 37°C. A, Short-term permeability to FITC-dextran was measured by quantifying fluorescence 30 min, 1 h, and 2 h after treatment as described in Materials and Methods. Results are an average from two wells and representative of two experiments with similar results. B, To measure permeability after longer exposure to OSM, IL-1β (100 ng/ml), hOSM (100 ng/ml), IL-6 (10 ng/ml), or TNF-α (10 ng/ml) was added to the upper chamber (containing DMEC growing on top of a filter membrane) of the two-part permeability assay kit for 18 h at 37°C. FITC-dextran was then added to the upper chambers of all test kits and allowed to diffuse for 1 h before collection of medium from lower chambers for analysis. In one case (OSM lower), hOSM (100 ng/ml) was added to the lower chamber for 18 h at 37°C before adding FITC-dextran. Results are an average from three wells and representative of two experiments with similar results. Fluorescence is measured in arbitrary units. C, Evans blue dye was injected into mouse footpads that had been treated with either rmOSM (1 μg) or PBS 24 h earlier. One minute after injection of dye, mice were euthanized, and the Evans blue dye was extracted from draining LN for quantification of OD measurement at 630 nm (28 ). The y-axis indicates the concentration of dye in the LN as determined by comparison with a standard curve of known amounts of dye.

To determine whether enhanced vascular permeability observed in vitro resulted in enhanced lymphatic flow in vivo, we treated mouse footpads with rmOSM, waited 24 h, and then injected Evans blue dye into the footpad. Lymphatic transport of Evans blue dye to draining LN normally occurs within minutes, and this technique has been used to document defects in lymphatic drainage in skin (28). Compared with PBS mock treatment, injection with OSM significantly increased the accumulation of dye in the LN by ∼3-fold (p = 0.0291) (Fig. 7⇑C). Therefore, OSM increases the permeability of DMEC and LMEC, which results in increased lymphatic flow and faster accumulation of Evan’s dye in the draining LN.