ERK1/2 and p38 regulate inter-individual variability in ozone-mediated IL-8 gene expression in primary human bronchial epithelial cells

Abstract

Inter-individual variability is observed in all biological responses; however this variability is difficult to model and its underlying mechanisms are often poorly understood. This issue currently impedes understanding the health effects of the air pollutant ozone. Ozone produces pulmonary inflammation that is highly variable between individuals; but reproducible within a single individual, indicating undefined susceptibility factors. Studying inter-individual variability is difficult with common experimental models, thus we used primary human bronchial epithelial cells (phBECs) collected from many different donors. These cells were cultured, exposed to ozone, and the gene expression of the pro-inflammatory cytokine IL-8 was measured. Similar to in vivo observations, we found that ozone-mediated IL-8 expression was variable between donors, but reproducible within a given donor. Recent evidence suggests that the MAP kinases ERK1/2 and p38 mediate ozone-induced IL-8 transcription, thus we hypothesized that differences in their activation may control IL-8 inter-individual variability. We observed a significant correlation between ERK1/2 phosphorylation and IL-8 expression, suggesting that ERK1/2 modulates the ozone-mediated IL-8 response; however, we found that simultaneous inhibition of both kinases was required to achieve the greatest IL-8 inhibition. We proposed a "dimmer switch" model to explain how the coordinate activity of these kinases regulate differential IL-8 induction.

Figure 1

Variability in ozone-mediated IL-8 induction among phBECs collected from different donors. IL-8 expression was assessed following in vitro exposure of phBECs to ozone (0.5 ppm/2 h). Each data point represents a phBEC culture collected from a different human donor. IL-8 expression was normalized to β-actin expression and expressed as fold change from filtered air exposure. Mean ± SD shown, n = 16 donors. To further investigate the mechanisms underlying ozone responsiveness using phBECs, we classified cultures as being “high” or “low” responders based on whether IL-8 inductions fell above or below the group mean (4.2 fold change).

Figure 2

Donor-specificity of phBEC IL-8 induction following in vitro ozone exposure. We investigated whether phBEC cultures collected from the same donors at different times had reproducible IL-8 expression following ozone exposure. Seven phBEC donors were used with time between collections ranging from 49–453 days. The consistency of ozone response was assessed by comparing the first exposure response (X-axis) with the second exposure response (Y-axis) via linear regression. Dotted lines are drawn at X = 4.2, Y = 4.2, which is the metric by which we differentiated high and low responders. The shaded areas indicate the quadrants the data points would cluster within if cultures were consistent in their response status (high or low) in both collections one and two.

Figure 3

The influence of ERK1/2 and p38 inhibition on ozone-associated IL-8 induction. To determine whether the activation of the MAP kinases ERK1/2 and p38 were required for ozone-mediated IL-8 induction, inhibitors of these kinases (250 nM SCH772984 and LY2228820, respectively) were added to cell media two hours prior to ozone exposure. (A) The phosphorylation of downstream targets of ERK1/2 and p38, RSK and MK2, respectively, were examined to verify kinase inhibition. Extract from phBECs stimulated with 5 μg/mL LPS for four hours was used as a positive control. The LPS positive control for each mark was run on the same gel, immunoblotted, and imaged at the same time as the blots in the adjacent row. The images were cropped to allow the LPS treatments to be shown in the same lane. (B) Seven donors were used for drug treatments, four of which were high responders (shown above) and three low responders (Figure S1). IL-8 inductions were normalized to a 0.2% DMSO filtered air vehicle control. IL-8 expression was compared between the ozone-vehicle control (O₃−V) and all other treatments via 2-way ANOVA. *p < 0.05.

Figure 4

The activation of ERK1/2, p38, and associated kinases in high and low-responding phBEC cultures. Cells were considered “high responders” if their IL-8 induction was above the group mean (4.2 fold change) and low responders were below this group mean. Protein was collected from twelve donors (n = 12), six high and six low responders. (A) Blots from two representative donors (one high, one low) show the phosphorylation of ERK1/2 and an upstream (MEK1/2) and a downstream (RSK) kinase, as well as p38 and its upstream kinase (MKK4). Both sets of blots were run on the same gel, immunoblotted, and imaged at the same time. (B) Densitometry analysis was used to calculate the fold change in activation (normalized to filtered air control) for each donor. Median activation and interquartile range are shown for each group. Differences between high and low responders were determined via Mann-Whitney test. *p < 0.05.

Figure 5

IL-8 expression as a function of ERK1/2 and p38 activation. In addition to analyzing MAPK activation based on and high and low responder status, we analyzed the data in continuous format in which the ratio of activated protein (X-axis) was plotted against ozone mediated IL-8 induction. This comparison was made for both ERK1/2 (A) as well as p38 (B). Figures were created using the same donors as depicted in Fig. 4 (n = 12; 6 high and 6 low responders).

Figure 6

Comparison of p65 activation in high and low-responding cultures. (A) Using the same panel of donors as Fig. 4, we examined the p65 phosphorylation at S536, an indicator of canonical NFκB activation. Primary cells were stimulated with 10 ng/mL TNFα for 20 minutes as a positive control. Representative donors shown are the same donors depicted in Fig. 4. Both sets of blots were run on the same gel, immunoblotted, and imaged at the same time. (B) Median and interquartile range of ozone-induced phosphorylation changes across 6 high and 6 low responders (n = 12) as calculated by densitometric comparison to filtered air controls. No significant difference was found between these two groups (p = 0.07). (C) Linear regression showing the correlation between ozone-associated changes in p65 phosphorylation and IL-8 mRNA expression.

Figure 7

Paradigm describing how differences in MAPK signaling lead to inter-individual variability in phBEC ozone-mediated IL-8 induction. Previous studies have suggested that ozone exposure leads to the activation of the MAP Kinases ERK1/2 and p38 by various mechanisms including the activation of the EGFR receptor. The activation of both kinases is important for full ozone-mediated IL-8 induction, yet the magnitude of ozone-induced IL-8 expression is only correlated with the level of ERK1/2 activation. Based on these observations, we propose a ‘dimmer switch’ model describing how p38 and ERK1/2 act in concert to regulate IL-8 expression. A basal level of p38 activation is required to achieve full IL-8 expression (acting as an ‘on/off’ switch), while the level of ERK1/2 pathway activation modulates the magnitude of expression (the ‘dimmer’). These kinases both activate effector proteins and kinases such as RSK and MK2, in addition to other transcription factors and stress-associated proteins. These ultimately converge on the IL-8 promoter and regulate gene transcription.