Regulation of IL-17A responses in human airway smooth muscle cells by Oncostatin M

We first examined the response of HASMC to recombinant OSM in the regulation of MCP-1/CCL-2 and IL-6 using ELISA systems for protein analysis. We observed a dose-responsive induction of MCP-1/CCL-2 (Figure 1A, left panel), detectable at 0.5 ng/ml and further elevated at higher concentrations of OSM. When stimulating HASMC cultures with IL-6, LIF or IL-31, at concentrations ranging up to 50 ng/ml, we did not observe detectable changes in MCP-1/CCL-2 supernatant levels. OSM stimulation also elevated IL-6 in supernatants of HASMC (Figure 1A, right panel), while we did not detect any changes in IL-6 upon stimulation with LIF or IL-31 in this system. To assess the effects of IL-17A in regulation of HASMC, we cultured cells with increasing doses of IL-17A (Figure 1B and C). We observed minimal responses to doses of IL-17A tested alone (0.1 to 10 ng/ml) when measuring IL-6 or MCP-1 levels in supernatants. However, we were able to detect synergistic induction of IL-6 (Figure 1B) and MCP-1 (Figure 1C) levels in supernatants upon co-stimulation with 10 ng/ml OSM and IL-17A. This was consistently evident in different HASMC cell lines examined, albeit the magnitude of responses varied somewhat between cell lines. Baseline levels of MCP-1 (approximate mean of 170 ug/ml, Figure 1C and 175 ug/ml in Figure 2A) were not statistically altered by stimulation with 10 ng/ml IL-17A (approximate mean of 225 ug/ml in Figure 1C and 235 ug/ml in Figure 2A). To determine whether lower concentrations of OSM could induce synergy with IL-17A, we cultured cells with increasing doses of IL-17A in the presence of 0, 0.1, 1 or 10 ng/ml OSM. Levels of MCP-1/CCL-2 (Figure 2A, left panel) and IL-6 (Figure 2A, right panel) showed that addition of OSM at 1 or 10 ng/ml could induce significant increases in MCP-1/CCL-2 or IL-6 responses to IL-17A.

We next assessed whether OSM could synergize with typical Th1 (IFNγ) or Th2 (IL-4, IL-13) cytokines in the regulation of MCP-1/CCL-2 by HASMC. Figure 2B and C show that OSM (0.5 ng/ml) co-stimulation with IL-4 or IL-13 resulted in synergistic increases in MCP-1/CCL-2 (left panels) and IL-6 levels were also elevated upon co-stimulation (right panels). On the other hand, OSM and IFNγ co-stimulations did not result in MCP-1/CCL-2 or IL-6 modulation differently than either cytokine alone (Figure 2D left and right panels). In the assessment of eotaxin-3 mRNA (Additional file 1: Figure S2), the function of OSM or IL-17A was negligible in comparison to that of IL-4 or IL-13. This is in contrast to IL-8 responses, which showed marked up-regulation by OSM/IL-17A combinations and not by IL-4/OSM or IL-13/OSM (Additional file 1: Figure S3).

The gp130 cytokine family shares several biological functions, mediated through gp130 signaling, in cells that express the specific ligand receptor chains. We thus assessed whether LIF (utilizes the complex of gp130 and LIFRα), IL-31 (utilizes the complex of IL-31R and OSMRβ chain), or other members of the gp130 cytokine family could regulate HASMC expression of MCP-1/CCL-2 in context of IL-4, IL-13 or IL-17A co-stimulation (Figure 3). MCP-1/CCL-2 levels (Figure 3A) were synergistically induced by the combination of OSM with IL-17A, IL-4 or IL-13. IL-6 levels (Figure 3B) were most pronounced with OSM and IL-17A combinations. Eotaxin-1 levels were induced synergistically by OSM in combination with IL-4 or IL-13, but not in combination with IL-17A (Figure 3C). Parallel testing of IL-31 or LIF (used at 10 ng/ml) shown in Figure 3A-C, showed no effects in the modulation of MCP-1/CCL-2, IL-6, or eotaxin-1 in this system alone or in combination with IL-17A, nor did IL-11 or IL-6 (shown in Additional file 1: Figure S1). In testing the effects of stimulation in serum-free conditions (2 days quiesed in RPMI/1%BSA before stimulation), the response to OSM or IL-17A, and the OSM/IL-17A combination, was identical in trend in the resulting levels of MCP-1and IL-6 found in the cell supernatants (Additional file 1: Figure S4 A, B), although we observed lower absolute quantities than the 2% FBS/RPMI conditions. We also observed identical trends with IL-4, OSM and IL-4/OSM regulation of eotaxin-1 in the same serum-free culture conditions (Additional file 1: Figure S4C) compared to 2% FBS conditions.

Since activation of the JAK-STAT signaling cascade is a prominent intracellular signaling pathway in response to gp130 cytokines, we assessed levels of tyr-phosphorylated STAT1, STAT3, STAT5 and STAT6 in HASMC lysates after cytokine stimulation (Figure 3D) as well as MAPK (p38 and JNK) and Akt pathway activation by standard immunoblots (Figure 3D). Quantitative assessment of band intensities through densitometry (Figure 3E) show that signals for P-STAT-1, 3 and 5 were upregulated by OSM but not by the other members of the gp130 cytokines tested in these cells. Signals for p-(T/Y)-JNK, p-(T/Y)-p38 and p-(S)-Akt were also upregulated by OSM, but minimal or no signals were observed upon stimulation with the other gp130 cytokines tested in these cells. This was consistent in two separate cell lines examined. We also observed that the regulation of STAT3 and p38 cell signaling (phosphorylation) in serum-free conditions were also identical in trend (Additional file 1: Figure S5) to that in 2% FBS/RPMI conditions. OSM-induced signals were evident and other gp130 cytokine signals were low/absent, although again, the levels of signals were much lower in serum free (RPMI/ 1% BSA) conditions.

To assess whether the regulation of the protein was linked to mRNA regulation, we measured mRNA levels in HASMC after stimulation with OSM, IL-4, IL-13 and IL-17A at both 6 hours (Figure 4A) and at 18 hours (Figure 4B). Changes in levels of mRNA reflected trends seen in the protein levels. OSM and IL-17A co-stimulation synergistically regulated MCP-1, IL-6 and IL-8, but not eotaxin-1 mRNA (Figure 4). OSM with either IL-4 or IL-13 synergized to induce MCP-1, IL-6 and eotaxin-1 but not IL-8 mRNA (Figure 4). Although eotaxin-3 mRNA was regulated by IL-4 or IL-13 (Additional file 1: Figure S2), we did not observe eotaxin-3 regulation by IL-17A, OSM, or the combinations. Furthermore, we were not able to detect eotaxin-3 in HASMC culture supernatants by ELISA (data not shown).

Since the regulation of receptor expression may influence responses by HASMC, we assessed the mRNA expression of receptor chains IL-17RA, IL-17RC, gp130, OSMRβ and IL-4Rα (Figure 5). OSM did not alter levels of IL-17RA or IL-17RC mRNA at 6 hours. Both IL-4 and IL-13 enhanced IL-17RA levels (over 4-fold and 2-fold, respectively) but did not regulate OSMRβ mRNA. OSM stimulation for 6 hours resulted in induction of mRNA levels for IL-4Rα 2.5-3 fold, and its own receptor chains OSMRβ and gp130 (approximately 4 fold and 2.3-fold, respectively). These altered levels showed the same trend at 18 hours (not shown). IL-17A stimulation did not regulate IL-4Rα, IL-17RA, IL-17RC, gp130 or OSMRβ, in context of its marked elevation of IL-8 mRNA levels (Figure 4).

To determine if selective signaling pathways were amplified with the combinations of cytokines, we assessed phosphorylation of STAT3, Akt at serine 473 and p38 by immunoblots (Figure 6A) and quantitative densitometry for 3 different HASMC cell lines (Figure 6B). p-(Y)-STAT3, p-(S)-Akt and p-(T/Y)-p38 levels were up-regulated by OSM alone and in combination with the other stimuli but not by IL-4, IL-13 and IL-17A alone in this system (Figure 6A). Levels of STAT3 or Akt activation were not statistically different in combination stimuli verses OSM alone in averaging the 3 cell line responses (Figure 6B, left and middle panels). However, levels of p38 phosphorylation on average were further increased (approx.1.7-fold, statistically significant) with OSM/IL-17A combination stimulations of HASMC (Figure 6B right panel). To examine whether inhibition of these pathways could alter HASMC responses to OSM and the synergistic responses induced by OSM combined stimuli, we assessed HASMC responses in the presence or absence of pharmacological inhibitors of STAT-3 (Stattic), p38 (SB203580) and Akt (Akt X). To test the effective concentrations and specificity of Stattic on STAT inhibition, 1.25, 2.5 or 5 uM Stattic was used to pre-treat HASMC cell cultures before stimulating with OSM (1 ng/ml). Levels of phosphorylated STAT1, 3, 5, p38 and Akt were assessed by immunoblots (Figure 6C). p-(Y)-STAT-1,3,5 signals at 20 minutes OSM stimulation were inhibited by 1.25 uM, more so by 2.5 uM and nearly complete reduction at 5 uM Stattic pre-incubation. Stattic is described as a STAT3 selective inhibitor [46,47] however; in our system Stattic markedly inhibited STAT-1 and STAT-5 activation as well. When we examined the effects of Stattic on other cell signaling pathways MAPK (p38) and Akt, we observed minimal reduction in p38 and Akt phosphorylation and a lack of dose-dependent effects (Figure 6C). To assess Stattic effects on MCP-1/CCL-2 or IL-6 production, 2.5 uM Stattic was used to pre-treat HASMC cultures before stimulation with the cytokines and combinations indicated (Figure 7A). Cell viability was not altered at this concentration of the inhibitor. Elevation of MCP-1/CCL-2 levels in response to OSM, or the synergistic MCP-1/CCL-2 responses to OSM/IL-4, OSM/IL-13 or OSM/IL-17A were reduced by 50% or more at 1.25 uM Stattic (not shown) and by 60% or greater at 2.5 uM Stattic (Figure 7A left panel). In contrast, Stattic did not alter IL-6 levels in response to the various stimuli, with the exception of the OSM/IL-17A combination (Figure 7A, right panel). This suggests that STAT-3 has a selective role in OSM/IL-17A synergy.

We next examined whether inhibition of p38 (Figure 7B) and Akt (Figure 7C) would alter cytokine responses in HASMC. MCP-1/CCL2 levels induced by IL-17A alone or in combination with OSM were significantly inhibited by SB203580 at the concentration used but were not significantly reduced in the other combinations (Figure 7B, left panel). IL-6 levels (Figure 7B right panel) were significantly inhibited by SB203580 regardless of the stimulation used in HASMC cultures. The Akt X inhibitor reduced MCP-1/CCL2 levels in supernatants of cultured HASMC independent of the stimuli used (Figure 7C, left panel) but did not inhibit IL-6 levels (Figure 7C, right).