In primary and telomerase-immortalized esophageal squamous cells from patients with EoE, we have shown that omeprazole, in concentrations as low as 1 µM, significantly inhibits IL-4-stimulated eotaxin-3 protein expression. We also have demonstrated that this inhibition appears to be a class effect of PPIs, because lansoprazole in low concentration (10 µM) also decreases cytokine-stimulated eotaxin-3 protein expression in EoE squamous cell lines. We have shown that omeprazole reduces IL-4-stimulated eotaxin-3 mRNA expression, and that this effect is not due to enhanced mRNA degradation, reduced STAT6 phosphorylation, or decreased phospho-STAT6 nuclear translocation. Rather, we have found that omeprazole blocks IL-4-stimulated mRNA transcription by reducing the binding of STAT6 and RNA polymerase II to the eotaxin-3 promoter, effects which are not due to alterations in the functional activity of STAT6 protein. We also have found that omeprazole reduces the levels of IL-4-stimulated H3K4me3 bound to the eotaxin-3 promoter. These findings suggest that, in EoE cells, omeprazole causes chromatin remodeling in the eotaxin-3 promoter, resulting in decreased RNA Pol II recruitment and reduced eotaxin-3 transcriptional activity. Thus, our findings elucidate molecular mechanisms whereby patients with Th2 cytokine-induced esophageal eosinophilia can respond to PPIs.
Although PPIs are prescribed primarily with the intent of controlling gastric acid secretion, these agents have been found to have a number of potentially beneficial biological actions that are independent of their antisecretory effects. For example, PPIs have demonstrable anti-oxidant properties, and have been found to inhibit certain neutrophil functions, to decrease adhesion molecule production and to block the production of the pro-inflammatory cytokine IL-8 by endothelial and epithelial cells [reviewed in 
]. In xenograft models, furthermore, PPIs decrease the growth of human tumors, presumably as a result of inhibiting the tumor cells’ vacuolar-ATPase (V-ATPase), a proton pump that regulates intra- and extracellular pH 
. In the present study, we have elucidated the molecular mechanisms underlying another gastric acid-independent, potentially beneficial PPI effect, namely the inhibition of Th2 cytokine-stimulated eotaxin-3 production by esophageal squamous cells.
In an earlier, “proof of principle” study, we found that omeprazole in high concentration (50 µM) decreased Th2 cytokine-stimulated eotaxin-3 expression by esophageal squamous cells 
. With conventional oral dosing of omeprazole, peak mean plasma concentrations of 3.2 µM have been documented, whereas levels as high as 10 µM have been reported with intravenous administration 
. In the present study, we found significant inhibition of IL-4-stimulated eotaxin-3 protein expression in esophageal squamous cells treated with omeprazole in concentrations as low as 1 µM. Thus, we have documented that the inhibitory effect of omeprazole on Th2 cytokine-stimulated eotaxin-3 expression occurs in vitro
using concentrations of the drug that are readily achieved in blood with conventional oral dosing.
For our experiments, we used acid-activated omeprazole and lansoprazole, and it is not clear whether acid activation of PPIs can occur in the esophagus. PPIs are known to accumulate and become activated in an acidic tissue microenvironment, which is found frequently around gastric parietal cells. Immune cells, including eosinophils and neutrophils, also can release protons from their exocytic granules and lysosomes into the microenvironment 
and, in the setting of inflammation associated with infection, asthma, and rheumatoid arthritis, microenvironmental acidification has been documented 
. Gastresophageal acid reflux also might acidify the esophageal microenvironment, and the Na+
exchanger on esophageal epithelial cell membranes is known to extrude intracellular protons that accumulate in the setting of injury in order to maintain intracellular pH 
. Thus, there are a number of plausible mechanisms whereby PPIs might be activated in an acidic microenvironment of the esophagus that is inflamed by EoE or GERD.
In an earlier study, we showed that IL-4-stimulated eotaxin-3 expression in esophageal squamous cells is mediated by STAT6 signaling 
, and there are a number of points in that signaling pathway where omeprazole might exert its inhibitory effects. In the stomach, omeprazole is known to alkylate cysteine residues in gastric H+
ATPase. In epithelial and non-epithelial cells, Cortes et al.
found that omeprazole decreased Th2 cytokine-induced STAT6 phosphorylation, an effect that they attributed to the PPI’s alkylating ability 
. Omeprazole did not affect total STAT6 protein levels in that study, and the authors suggested that the decreased STAT6 phosphorylation might be due to omeprazole-induced modifications either of the STAT6 protein itself or of the ability of upstream kinases to phosphorylate the protein 
. In murine myeloid cells, Perez et al.
described a different mechanism by which the drug n-alpha-tosyl-L-phenylalanine-chloromethyl ketone (TPCK) interfered with IL-4-induced STAT6 activation 
. By virtue of its alkylating properties, TPCK induced modifications of the STAT6 protein’s cysteine residues, which facilitated the degradation of total STAT6, thus resulting in decreased STAT6 phosphorylation levels 
. In contrast to these studies, we found no apparent effect of omeprazole on the levels of IL-4-induced STAT6 phosphorylation or total STAT6 protein in our EoE cells. Moreover, we found no apparent effect of omeprazole on the nuclear translocation of phospho-STAT6.
Our observation that omeprazole significantly decreases the binding of STAT6 to the eotaxin-3 promoter suggests either that omeprazole induces modifications to the eotaxin-3 chromatin structure, or that omeprazole induces modifications of the STAT6 protein that affect its function. To distinguish between these possibilities, we transiently transfected with an exogenous eotaxin-3 promoter construct that is activated by STAT6 binding 
. Unlike the endogenous promoter. this exogenous eotaxin-3 promoter has an open and accessible chromatin structure, and is not subject to complex chromatin regulatory mechanisms 
. We found no significant differences between omeprazole-treated and untreated EoE cells in the degree of activation of the exogenous eotaxin-3 promoter, showing that omeprazole does not induce modifications to the STAT6 protein that interfere with its function.
To address whether the omeprazole-induced decrease in binding of STAT6 was accompanied by decreased transcriptional activity, we assessed RNA polymerase II (Pol II) binding to the eotaxin-3 promoter by ChIP assay. Our observation that omeprazole significantly decreased binding of IL-4-stimulated Pol II to the endogenous eotaxin-3 promoter suggests that the PPI reduced eotaxin-3 transcriptional activity. Using ChIP assay, we also found that omeprazole reduced the levels of IL-4-stimulated H3K4me3 bound to the endogenous eotaxin-3 promoter. This trimethylation is a post-translational modification of histone H3 that occurs as genes are induced, and approximately 91% of all RNA Pol II binding sites correlate with H3K4me3 binding sites 
. Taken together, our observations suggest that omeprazole causes chromatin remodeling in the eotaxin-3 promoter, resulting in decreased RNA Pol II recruitment and reduced eotaxin-3 transcriptional activity in EoE cells.
In conclusion, we have shown that omeprazole, in concentrations that can be achieved in plasma with conventional dosing, significantly decreases IL-4-stimulated eotaxin-3 expression in esophageal squamous cells from patients with EoE. In those cells, omeprazole does not inhibit STAT6 phosphorylation, does not reduce the level of total STAT6 protein, and does not reduce the translocation of phopho-STAT6 to the nucleus. Rather, omeprazole reduces the transcription of eotaxin-3 mRNA by reducing the binding of STAT6 and RNA polymerase II to the exotaxin-3 promoter in association with a reduction in the levels of promoter-bound H3K4me3. These findings elucidate molecular mechanisms whereby patients with Th2 cytokine-driven esophageal eosinophilia can respond to PPIs, independent of effects on gastric acid secretion. These findings might explain the phenomenon of PPI-responsive esophageal eosinophilia, and suggest that even patients who have EoE without GERD might benefit from PPI therapy. Nevertheless, further studies are needed to establish that these PPI effects observed in cells in vitro are applicable to patients.