Characterisation of the interleukin-4 signalling cascade involved in the induction of cyclooxygenase-2 and the polarisation of macrophages into an anti-inflammatory phenotype

The inflammatory response can be broken down into the pro-inflammatory phase, followed by the resolution phase. Both phases are regulated by the expression or release of immunomodulatory mediators such as cytokines e.g. TNFα and IL-10. The two phases are also associated with different macrophage polarization states where the M1 phenotype is seen during the pro-inflammatory phase and M2 during resolution. Whereas much focus has been on the pro-inflammatory phase, more recently, emphasis has been placed on initiating the resolution phase of inflammation. Historically one of the classical signs of inflammation is the expression of the COX-2 enzyme in neutrophils and macrophages early in the pro-inflammatory phase. However, over the last 20 years, a few research groups have observed a delayed but elevated expression of COX-2 in macrophages, along with a range of cytokines such as IL-10, normally associated with the resolution of inflammation. In addition, other researchers have found that administration of COX-2 inhibitors late in the inflammatory process can exacerbate some chronic inflammatory diseases in vivo. The concept of an anti-inflammatory COX-2 was further supported by studies that showed prolonged exposure to diclofenac in vitro, induced a COX-2 enzyme and high levels of anti-inflammatory cytokines at a time when resolution would be expected. These observations led to the suggestion that the expression of COX-2 in the resolution phase may in fact be a key mediator of the resolution process. To explore these ideas further, this study focuses on the effects of IL-4, an anti-inflammatory cytokine that is widely associated with the resolution phase of inflammation. A key objective will be to investigate the role of other factors such as peroxisome proliferator-activated receptor gamma (PPARγ), cyclic adenosine monophosphate (cAMP) and p38 mitogen activated kinase (MAPK) in the signalling pathways through which IL-4 could potentially induce COX-2 protein and other mediators of resolution.

The experiments were designed to investigate whether there are similarities between the induction of COX-2 and the polarisation of macrophages down an anti-inflammatory line. As such, the macrophage J774.2 cell line was treated with 1μg/ml lipopolysaccharide (LPS), 500μM diclofenac or 10ng/ml IL-4 for 24 and 48 hours. These cells were assessed for COX-2 expression and activity, the production of the pro-inflammatory cytokine, tumour necrosis factor alpha (TNFα) and the anti-inflammatory cytokine, interleukin 10 (IL-10) respectively and the anti-inflammatory mediator, cAMP. The importance of p38 and PPAR was assessed using the specific p38 inhibitor, SB203580 and a PPAR antagonist, Bisphenol A diglycidyl ether (BADGE). Following preliminary pharmacology experiments, the working concentration of each drug was decided: 10μM SB203580 and 150μM BADGE. In a subset of experiments, the effect of inhibition of COX-2 by 0.1-100μM acetaminophen, 1-100μM indomethacin and 1-5μM dexamethasone (DEX) was also investigated; with the aim of identifying differential effects depending on the stimulant.

LPS treated macrophages produced high levels of TNFα and low levels of IL-10 within 24 hours of exposure. By contrast IL-4 produced high IL-10 and low TNFα levels at 48 hours. This confirms the opposing phenotypic state of the macrophages post-stimulation. Following IL-4 stimulation, the studies identified a possible sequence of events leading up to the induction of COX-2 and prostaglandin E2 (PGE2) synthesis. Adding antagonists of cAMP, PPAR or p38 MAPK to cells stimulated with IL-4 reduced or blocked COX-2 expression and activity leading to the conclusion that these factors become activated upstream to COX-2 induction by IL-4. In addition, cAMP was regulated at the level of both PPAR and p38 while PPAR was regulated by p38. Exposure to indomethacin reduced and blocked TNFα and IL-10 induced by LPS and IL-4 respectively. DEX specifically blocked TNFα secretion but induced IL-10 secretion.

Blocking both p38 and PPAR in IL-4 treated cells, prevents secretion of anti-inflammatory cytokine IL-10. The novel finding that the p38 kinase and PPAR receptor are required for both the induction of COX-2 and secretion of IL-10, suggests an overlap between macrophage polarisation to an anti-inflammatory phenotype and late COX-2 induction pathways. P38, PPAR and cAMP would therefore be reasonable drugs targets to induce resolution. These studies suggest that in addition to being involved in driving inflammation, COX-2 may also drive resolution; however, this is yet to be confirmed with further experiments. These studies may go some way to explaining the adverse reactions and an impaired immune response reported in some individuals administered COX-2 inhibiting NSAIDs.