Who Tidies Up the Lung? Listen to Cox-2 and Nrf2

Cyclooxygenase-2 (Cox-2) recently made headline news when specific inhibitors of this enzyme were shown to increase the risk of myocardial infarction (1). Cox-2 is one of two known isoenzymes of cyclooxygenase (a Cox-3 may exist as well), and metabolizes arachidonic acid to prostaglandin H^sub 2^, the precursor of thromboxane and prostaglandins. The discovery of Cox-2 in the late 1980s quickly spawned the concept of a housekeeping Cox-1 and a proinflammatory Cox-2 isoenzyme. Activation of Cox-2 in many inflammatory processes was soon demonstrated, the lung being no exception (2, 3). Early in inflammation, newly transcribed Cox-2 appears to be responsible for the production of prostaglandins and thromboxane, whose overall function is to promote inflammation, fever, and pain (4). However, the longer Cox-2 was studied, the clearer it became that the original dichotomous concept of a constitutive Cox-1 and an inducible proinflammatory Cox-2 was not tenable. Not only do some (lung) cells express Cox-2 constitutively, and Cox-2-deficient mice develop surprisingly many maladies (5), but Cox-2 also promotes the healing of mucosal lesions (6) and the resolution of inflammation (3).

Using a model of carrageenin-induced pleurisy, Gilroy and colleagues (3) confirmed that inhibition of Cox-2 is antiinflammatory in the early phase of inflammation, but surprisingly, such inhibition aggravated inflammation at later time points, such as 24 hours after injection of carrageenin. The Cox-2 metabolite responsible for the resolution of inflammation was identified as 15-deoxy-?^sup 12,14^-prostaglandin J^sub 2^ (15-d-PGJ^sub 2^), which is derived from PGD^sub 2^ (7). Recently, it was suggested that 15-d-PGJ^sub 2^ acts by stimulating the demise of leukocytes and monocytes through apoptosis (8). Of note, acute lung injury is characterized by delayed apoptosis of neutrophils in the alveolar space (9, 10). These are important findings since resolution and containment are essential elements of inflammation, and an uncontrolled immune response may be no less detrimental than no response at all.

An important question concerned the molecular mechanisms by which 15-d-PGJ^sub 2^ stops inflammation and tidies up the lung from the remainders of inflammation. At a molecular level it was originally believed that 15-d-PGJ^sub 2^ acts primarily through activation of peroxisome proliferator-activated receptor-?, or PPAR-?, but it soon became clear that other mechanisms exist as well (7). One such mechanism is the activation of Nrf2 through removal of its cytosolic inhibitor, Keap1, by 15-d-PGJ^sub 2^ (11).

Building on these observations, in this issue of AJRCCM (pp. 1260-1266), Mochizuki and colleagues (12) now provide a molecular mechanism to explain how 15-d-PGJ^sub 2^ helps to resolve pulmonary inflammation (12). They created a novel model of acute lung injury by instilling carrageenin into the airways of mice. Carrageenin is a long-chain polysaccharide extracted from red seaweed, which is widely used in dairy products for its gelling and stabilizing properties. Carrageenin has also been used in models of inflammation, such as the rat-paw edema model (13). The deposition of carrageenin inside the body, with the exception of the gastrointestinal tract, leads to local inflammation at the site of injection, including paw edema (13), pleurisy (3), and acute lung injury (12). Mochizuki and coworkers show for the first time that intratracheal instillation of carrageenin produces alveolitis and edema formation within a day, and that the inflammation resolves within the following 2 days. However, as expected from the work in the carrageenin-induced pleurisy model, inhibition of Cox-2 reduced the production of 15-d-PGJ^sub 2^ and delayed the resolution of inflammation by several days. Mochizuki and coworkers provide the important observation that the molecular target of 15-d-PGJ^sub 2^ in this in vivo model is nuclear factor E2 p45-related factor 2 (Nrf2). The central role of Nrf2 is shown by the greatly prolonged inflammation in Nrf2-deficient mice and the failure of 15-d-PGJ^sub 2^ to resolve inflammation in the Nrf2 knockout mice.

Nrf-2 is a transcription factor that is best known for providing cellular protection against oxidative insults and chemical carcinogens via transcriptional activation of antioxidant/detoxifying enzymes (14). Genetic ablation of Nrf2 enhances susceptibility to lung injury induced by hyperoxia (15) or butylated hydroxytoluene (16) to bleomycin-induced fibrosis (17) and to cigarette smoke-induced emphysema (18). These are essentially all models of oxidative stress. Mochizuki and coworkers show that Nrf2 regulates the expression of peroxiredoxin 1 and heme oxygenase-1 (HO-1). However, in the pleurisy model, and likely in the present alveolitis model as well, resolution of inflammation is associated with increased rates of leukocyte apoptosis (8), an event not readily explained by lowered oxidative stress. Clearly, the relationship between Nrf2 and apoptosis needs further investigation. The fact that apoptosis itself is antiinflammatory and prompts release of antiinflammatory mediators also needs to be considered in this context (10).

Because oxidative stress plays a less prominent role in carrageenin-dependent inflammation, there might be another explanation for the resolution of inflammation conferred by Nrf2. Rangasamy and colleagues (18) have provided a list of nearly 50 Nrf2-dependent protective genes, and it appears reasonable to assume that Nrf2 controls genes beyond those involved in oxidative stress. In addition, some of the genes originally considered to be of importance only in oxidative stress have now been shown to have a much broader role. For example, HO-1 has both antiinflammatory and antiapoptotic effects (19). It therefore seems reasonable to speculate that Nrf2 might provide a useful target able to modify the course of pulmonary inflammation.

Many models of acute lung injury focus only on events during the first 24 hours. From a clinical perspective, there is need for models and mechanisms that can explain the pathophysiology of pulmonary inflammation beyond that time period. Both the new carrageenin model of lung injury and the findings concerning Nrf2 offer stimulating new insights into the later stages of pulmonary inflammation.

Conflict of Interest Statement: S.U. does not have a financial relationship with a commercial entity that has an interest in the subject of the manuscript.