Abstract
The discovery of endogenous molecules involved in counterregulation of inflammatory responses that may lead to tissue injury provides an opportunity to explore new therapeutic approaches based on manipulation of new pathways. Natural counterregulatory pathways may reduce the possibility of unwanted toxic side-effects. Lipoxins are trihydroxytetraene-containing eicosanoids that are generated within the vascular lumen during platelet-leukocyte interactions and at mucosal surfaces via leukocyte-epithelial cell interactions. During cell-cell interactions, transcellular biosynthetic pathways are the major lipoxin biosynthetic routes, and thus, in humans, lipoxins are formed in vivo during multicellular responses, such as inflammation and asthma. This branch of the eicosanoid cascade generates specific tetraene-containing products that serve as "stop signals" for neutrophils that regulate key steps in leukocyte trafficking and prevent neutrophil-mediated tissue injury. These novel anti-inflammatory lipid mediators also appear to facilitate the resolution of the acute inflammatory response. In this review, recent findings and new concepts pertaining to the generation of lipoxins and their impact on the resolution of acute inflammation, and organ protection from leukocyte-mediated injury, are presented. The parallels and possible associations with periodontal diseases are discussed.
(I) Introduction
Inflammation is a physiologic series of responses generated by the host in response to infection or other insults. Inflammation can have rapid onset and last a short period of time (acute inflammation), or it can persist due to a continuous stimulus or injury (chronic inflammation). The initial events of inflammation are derived from vascular reactions at the site of injury. Vascular changes are important for the induction of the response and are characterized by redness, heat, and swelling, usually accompanied by pain and loss of function, and collectively represent the "cardinal signs" of inflammation (Cotran et al., 1999). These signs of inflammation are the result of vasodilatation and increased vascular permeability, leading to exudation of fluid and plasma proteins and recruitment of leukocytes to the site of injury (Larsen and Henson, 1983; Sharma and Mohsin, 1990; Trowbridge and Emling, 1997).
Inflammation can be initiated by physical, chemical, or biological stimuli. Inflammation has the net effect of confining the injury or insult to an isolated area and serves as the first step in the initiation of the immune response, through which the infection is eliminated and the injury is repaired (Cotran et al., 1999). In general, once the insult is eliminated, the inflammatory response resolves and subsequent immune reactions diminish. In the case of chronic inflammation, however, the persistence of the response leads to host tissue destruction and may result in irreversible pathological changes. Periodontal diseases are important chronic infections leading to chronic inflammation that results in a host-mediated destruction of the supporting tissues of the dentition. The pathogenesis of periodontal diseases has been the subject of intense study over the past two decades, and is reviewed elsewhere in depth (Kornman et al., 1997; Landi et al., 1997).
Inflammatory events are initiated, enhanced, or coordinated by the actions of various chemical mediators. Several cell types, such as mast cells, platelets, and leukocytes, are responsible for the release of inflammatory mediators and play an important role in the development of inflammation. Elevated levels of these chemical signals (cytokines and chemokines) promote inflammation and further function to amplify the response (Goldsby et al., 2000). Cytokines and chemokines are generally proteins, although low-molecular-weight lipids derived from arachidonic acid play an important pro-inflammatory autocrine role. In addition to these molecules, gases like nitric oxide and carbon monoxide, reactive oxygen species, and lately nucleotides are also recognized as inflammatory mediators (Serhan, 2001; Wada et al., 2001).
(II) Lipid Inflammatory Mediators
(A) Eicosanoid generation and arachidonic acid metabolism
Eicosanoids are compounds derived from eicosa- (20-C) polyenoic fatty acids and consist of the prostanoids and leukotrienes (LT). Prostanoids can be further grouped as prostaglandins (PG), prostacyclins (PGI), and thromboxanes (TX). Although the term ’prostaglandin’ is often loosely used to include all prostanoids, this group of eicosanoids refers to a specific group of fatty acids acting as local hormones with important physiologic and pharmacological activities (Murray et al., 1999). Altogether, eicosanoids elicit their biochemical effects through G-protein-linked receptors. Arachidonic acid (5,8,11,14-cis-eicosatetraenoic acid) is an ω6 series eicosanoic acid with 4 double bonds. Initially isolated from peanut (arachis) oil together with linoleic acid, arachidonic acid is recognized to be an important component of phospholipid metabolism in animals. It is generated from linoleic acid or α-linoleic acid in most mammals by the action of a desaturase enzyme system (Kelley, 2001). Biosynthesis of most eicosanoids starts at arachidonic acid. Initially, molecules are derived from the 2-position of phospholipids in the cell membrane through the action of phospholipase A2 (PLA2) (Murakami et al., 1997).
Prostanoids are structurally similar in the sense that they contain a cyclic ring and are produced via the action of the enzyme cyclooxygenase (COX). The in vivo synthesis of prostaglandins takes place by cyclization of the center of the carbon chain of the eicosanoic polyunsaturated fatty acids to form a cyclopentane ring. The cyclopentane ring of thromboxanes is interrupted with an oxygen atom and is thus referred to as an "oxane" ring. Thromboxanes were first discovered in platelets, whereas prostaglandins were named after their initial detection in the seminal fluid (Hamberg and Samuelsson, 1966; Hamberg et al., 1975). Later research has shown that virtually all mammalian tissues contain these important molecules and that they play a major role in the host defense (Vane, 1976; Samuelsson, 1987; Yoshikai, 2001). The nomenclature of prostanoids reflects two important structural features. First, the number of double bonds in the side-chains can be between 1 and 3, indicating three different eicosanoic fatty acids, and this number is designated in the subscript (e.g., PG1, PG2, and PG3). Second, the capital letters (e.g., A, B, C, etc.) depict the variations in groups attached to the rings. For example, with this terminology, PGE2 would refer to a molecule with two double bonds in the side-chains and the presence of a keto group in position 9 (Coleman et al., 1995).
Arachidonic acid is converted into PGG2 by the action of COX (also known as prostaglandin synthase) that introduces 2 oxygen molecules into its structure. This first intermediate of arachidonic metabolism is catalyzed into the highly unstable PGH2, which is rapidly transformed into other prostaglandins, prostacyclin (PGI2), and thromboxanes (Coleman et al., 1995; Murray et al., 1999). COX has been found to have two isoenzymes: COX-1 and COX-2. These isoenzymes share 64% identity and similar enzymatic activity (O’Banion et al., 1992). Moreover, there are also important common structural features. Both types of COX have a conserved hydrophobic region, a heme-containing domain, and glycosylation sites. The most important difference between COX-1 and COX-2 lies in their functional properties. While COX-1 is expressed constitutively, COX-2 is generated by ligand-coupled stimulation. Therefore, COX-2 is alternatively named the "inflammatory cyclooxygenase" (O’Banion, 1999). As an example, COX-2 is transcribed and synthesized de novo after stimulation of monocytes by IL-1β or LPS. Furthermore, it has been shown that the PGE2 production is independent of COX-1 induction (Xie et al., 1993).
The second group of eicosanoid derivatives consists of the leukotrienes. In contrast to the cyclization of the fatty acid chain in prostanoids, leukotrienes are generated through the lipoxygenase (LO) pathway (Lee and Austen, 1986). Their name reflects their initial discovery in leukocytes. The first leukotriene that is formed is LTA4, which in turn is metabolized into either LTB4 or LTC4. Through this process, a peptide glutathione is added to LTC4. LTC4 is converted into LTD4 and LTE4 by the removal of glutamate and glycine. Three different lipoxygenases insert oxygen into the 5, 12, and 15 positions of arachidonic acid, giving rise to hydroperoxides of eicosatetraenoic acid (HPETE). The lipoxygenases 5-LO, 12-LO, and 15-LO are found in neutrophils, platelets, and endothelial/epithelial cells, respectively, and their products are also named accordingly (5-HPETE, 12-HPETE, and 15-HPETE) (Samuelsson et al., 1987; Westlund et al., 1988; Serhan, 1997; Funk, 2001; Bandeira-Melo et al., 2002).
(B) Physiological significance of eicosanoids
Eicosanoid metabolism plays an important role in the regulation of many physiological events in various organ systems (Table). Arachidonic acid derivatives and lipid mediators of inflammation play critical roles in health and disease states. Animals do not have the capacity to desaturate the fatty acids fully compared with plants (Rivers and Frankel, 1981). Therefore, it is necessary for many animals to acquire certain polyunsaturated "essential" fatty acids derived from a plant source as a part of their diet. In humans, certain symptoms, such as skin lesions, occur when the diet lacks essential fatty acids (Rivers and Frankel, 1981; Simopoulos, 1999; Schaefer, 2002). Linoleic-acid-rich diets or essential fatty acid intake that consists of 1-2% of the total caloric requirement can reverse such pathological symptoms (Lanzmann-Petithory, 2001). In addition to the lack of dietary intake of essential fatty acids, abnormal metabolism of these molecules can also be associated with diseases such as cystic fibrosis, Crohn’s disease, cirrhosis, and alcoholism (Rivers and Frankel, 1981; Laffi et al., 1997; Freedman et al., 2000; Esteve-Comas and Gassull, 2001). Diet with a high polyunsaturated fatty acid content is beneficial in decreasing the serum cholesterol levels and low-density lipoproteins, and balanced consumption of essential fatty acids is essential to prevent the risk of cardiovascular diseases associated with high cholesterol levels (Schaefer, 2002). An interesting study on Greenland Eskimos showed that low incidence of heart disease and prolonged clotting times were correlated with high intake of fish oils containing 20:5 ϖ3-eicosapentaenoic acid (Goto et al., 1993). Eicosapentaenoic acid is the substrate in the generation of PG3 series and TXA3 that inhibit the release of arachidonic acid from phospholipids and the formation of PG2 series and TXA2. Through this mechanism, platelet aggregation is lowered and plasma concentrations of cholesterol, triacylglycerol, and low-density lipoproteins are decreased, while the high-density lipoproteins are raised.
A classic role of lipid mediators is seen in the example of anaphylaxis. Slow-reacting substance of anaphylaxis (SRS-A) is a mixture of LTC4, LTD4, and LTE4 (Ford-Hutchinson, 1991). These molecules are 100-1000 times more potent than histamine or prostaglandins in their constriction of the bronchial airway muscles. Together with LTB4, these leukotrienes cause vascular permeability and leukocyte accumulation and activation. Thus, the leukotrienes are important regulators in many diseases involving inflammatory or immediate hypersensitivity reactions, such as asthma.
Prostaglandins are also responsible for various important physiological events, such as the contraction of smooth muscles, inflammation, induction of labor at term, prevention of gastric ulcers, and relief of asthma (Turini and DuBois, 2002). However, pathological processes can affect these functions. Data on the levels of PGE2 in periodontal diseases and the association of the periodontal diseases with low birthweight suggest that periodontal infections alter the prostaglandin metabolism and might lead to systemic problems (Offenbacher et al., 1996). In a similar fashion, other recent studies on the role of periodontal disease in increased risk for cardiovascular diseases imply that the link between these two diseases might be through lipid mediators (Beck and Offenbacher, 2001). Thus, periodontal diseases present an important model for inflammatory diseases that may also modify the course of systemic disease.
The products and the substrates of eicosanoid metabolism are associated with multiple functions, and the production of these molecules is through the cyclooxygenase (COX) and lipoxygenase (LO) pathways. Since many molecules in this metabolism are associated with pro-inflammatory roles, blocking the actions of arachidonic acid cascade has been realized to be an effective means of blocking the inflammation. Based on this principle, several drugs have been developed and are used to arrest or modify inflammation by blocking the enzymatic pathways that lead to the generation of lipid mediators. In this context, part of the activity of corticosteroids is thought to be through blockage of the release of arachidonic acid from the membrane phospholipids (Vane and Botting, 1987). This non-specific inhibition of the eicosanoid generation, although highly effective, leads to multiple side-effects, including suppression of the acquired immune system. Likewise, non-steroidal anti-inflammatory drugs (NSAIDs) or specific COX inhibitors can be used to block prostaglandin production. NSAIDs specifically suppress COX-mediated inflammation without blocking acquired immunity (Vane and Botting, 1987). NSAIDs are effective and used widely. However, these agents also lead to significant side-effects, due to their lack of selectivity, blocking both the constitutive (COX-1) and the inducible (COX-2) isoforms. Drugs that are more specific for COX-2 are now available, as well (Matheson and Figgitt, 2001).
Aspirin has been widely used for a long time to decrease inflammation, pain, and fever. It was discovered that aspirin inhibits the synthesis of prostaglandins by inhibiting the prostaglandin synthase (COX) pathway (Vane, 1971). The specific action of aspirin has long been thought to be attributable to its irreversible blockage of COX activity by acetylation of a specific serine hydroxyl group. Since aspirin suppresses the synthesis of prostaglandins at this initial step, for many years, the activity of aspirin was attributed to this mechanism. It was also discovered that aspirin prevents excessive blood clot formation through inhibition of TXA2 formation (Vane and Botting, 1987). New information on the activity of aspirin will be presented later in this review.
(III) Lipoxins
Once inflammation is initiated, the cascade of inflammatory events is an amplified loop until the infection is contained via immune reactions, or injury is confined. These early actions of the host response are later replaced by more specific mechanisms and eventually become redundant. Thus, it is important for the inflammatory response to be limited and resolved. Many molecules play a counterregulatory role in this resolution stage of inflammatory response to control the magnitude and duration of the inflammatory response (Serhan and Prescott, 2000).
Endogenous lipid-derived mediators have been demonstrated to moderate the host response and coordinate the resolution of inflammation (Diamond et al., 1999). Recently, several novel lipid mediators have been described as potential anti-inflammatory molecules, illustrating the importance of endogenous generation of lipid mediators with anti-inflammatory properties. An important example of lipid mediators with inflammation-resolving properties is the lipoxins (LX). Lipoxins contain a trihydroxytetraene group and are members of the eicosanoid family that are produced within the vascular lumen, primarily via platelet-leukocyte transcellular biosynthesis (Samuelsson et al., 1987; Serhan and Sheppard, 1990; Edenius et al., 1991). Lipoxins can be generated by several different pathways. In general, cell-cell interactions result in the generation of lipoxins, while single cells also can produce lipoxins (Serhan, 1994). Lipoxin generation is a very rapid process that is activated by inflammation, atherosclerosis, and thrombosis (Levy et al., 2001). Cell-cell interactions, that occur during these events and lead to generation of lipoxins, can also induce transcellular biosynthetic routes that lead to amplification signals such as leukotrienes and prostaglandins or to braking signals that involve novel compounds (Brady and Serhan, 1992). Therefore, lipoxin production is an important stage of the inflammatory response.
(A) 15-lipoxygenase-initiated pathway
Biosynthesis of lipoxin was first demonstrated in 1984 (Serhan et al., 1984). It was shown that insertion of molecular oxygen into the 15-carbon (C15) position of arachidonic acid is essential for lipoxin production. In addition, it was also reported that lipoxin generation through this pathway required the action of 15-LO. Once oxygenated at the C15 position, arachidonic acid is converted into 15-hydroperoxyeicosatetraenoic acid (15-HPETE), which is a substrate for 5-LO in leukocytes (Serhan et al., 1984; Ford-Hutchinson, 1991). This molecule is rapidly converted by hydrolases to either 5S,6R,15S-trihydroxy-7,9,13-trans-11-cis-eicosatetraenoic acid (lipoxin A4; LXA4), and/or, via lipoxin B4 hydrolase, to 5S,14R,15S-trihydroxy-6,10,12-trans-8-cis-eicosatetraenoic acid (lipoxin B4; LXB4). LXA4 and LXB4 are vasoactive molecules, primarily vasodilatory in vivo, and regulate leukocyte functions (Badr et al., 1989; Lee et al., 1991; Serhan et al., 1995). While the 15-LO-initiated route synthesizes lipoxin, 5-LO activation blocks leukotriene synthesis (Lindgren and Edenius, 1990, 1993). Thus, these series of events can be viewed as an inverse relationship between leukotriene and lipoxin synthesis (Serhan, 1994; Claria and Serhan, 1995). When neutrophils produce lipoxins by carrying the alcohol group in 15-HETE to either the R or S configuration, leukotriene formation is dramatically reduced.
The role of the neutrophil in lipoxin generation is crucial. It has been shown that primed neutrophils are another source of LX biosynthesis (Brezinski and Serhan, 1990). This mechanism is mediated by the esterification of 15-HETE in inositol-containing phospholipids within the cell membrane. Cells rapidly convert 15-HETEs into their inositol-containing lipids. When stimulated by various agonists, neutrophils release 15-HETE, which is further transformed into lipoxin. This pathway suggests that precursors of lipoxin synthesis can be stored within membranes of inflammatory cells and then released by stimuli activating PLA2 (Brezinski and Serhan, 1990). Furthermore, membrane priming of neutrophils generates bioactive lipid mediators that have implications for second messengers, such as 15-hydroxy-phosphatidyl inositol diphosphate and diacylglycerol, which contain 15-HETE that may alter their intracellular signaling activities. As is the case with 15S-HETE, 15R-HETE is also rapidly esterified into membrane phospholipids and could probably serve as a reservoir for epimeric lipoxin formation (Claria and Serhan, 1995).
(B) 5-lipoxygenase-initiated pathway
The second route for LX biosynthesis occurs with the interaction of human neutrophils with platelets in the blood vessels. In this model, cell-to-cell interaction involves 5-LO in neutrophils and 12-LO in platelets for the insertion of molecular oxygen into arachidonic acid (Edenius et al., 1988; Fiore and Serhan, 1990). Under resting conditions, unstimulated neutrophils do not generate considerable amounts of lipoxin, and the majority of neutrophil-generated LTA4 by the 5-LO pathway is released into the extracellular environment (Fiore and Serhan, 1990). When platelets adhere to neutrophils, however, they convert LTA4 into a cation (carbonium) by 12-LO (Serhan and Sheppard, 1990). As a result, platelet 12-LO reduces hydrogen at the C13 position and inserts oxygen into the C15 position of LTA4, converting LTA4 either to LXB4 at the C14 position or LXA4 at the C6 position (Romano and Serhan, 1992; Sheppard et al., 1992; Romano et al., 1993). Both LXB4 and LXA4 generation is exclusively mediated by 12-LO. Thus, 12-LO functions as a "lipoxin synthase" in platelets (Serhan and Sheppard, 1990). These observations with isolated intact platelets from human peripheral blood were also confirmed with recombinant 12-LO (Romano et al., 1993; Lehr et al., 1994).
Transcellular lipoxin synthesis requires cell adhesion. Therefore, cell adhesion characteristics play an important role in lipoxin metabolism. Indeed, a study on P-selectin-deficient knockout mice has showed that the adhesion is selectin-dependent, and when adhesion between the cells is blocked by specific antibody, transcellular lipoxin biosynthesis was also blocked (Mayadas et al., 1996). This in vitro observation was also demonstrated in ConA-induced glomerular nephritis (Papayianni et al., 1995). LXA4 levels were restored and neutrophil infiltration was normalized when the P-selectin-deficient mice were infused with wild-type platelets. These observations suggest that platelet-neutrophil adherence and transcellular biosynthesis are important inflammatory events that regulate neutrophil recruitment by initiating the formation of lipid mediators that suppress pro-inflammatory responses. Thus, the role of platelets in lipoxin generation during platelet-neutrophil interaction within the blood vessels might be an important factor in regulating the extravasation of neutrophils (Lehr et al., 1994). The vascular route of platelet-leukocyte interactions elevates the formation of lipoxin by transcellular conversion of LTA4 (Fiore and Serhan, 1989; Krump et al., 1997). This appears to be a major route for lipoxin generation, particularly when the platelet COX-1 is inhibited by non-steroidal anti-inflammatory drugs (NSAIDs).
(C) Additional pathways for lipoxin generation
Neutrophil-released LTA4 that is converted by the 15-LO in epithelial cells, particularly in tracheal epithelial cells, can also generate lipoxins by an LTA4-dependent mechanism (Edenius et al., 1990; Garrick and Wong, 1991). In this pathway, the substrate for 15-LO, LTA4, also plays a critical role (Edenius et al., 1990; Serhan and Sheppard, 1990). Another biosynthetic route involves 5,6-dihydroxyeicosanoids, which are also substrates for conversion to lipoxin. These reactions can include both 15-LO or 12-LO conversion of 5,6-dihydroxyeicosanoid substrates to lipoxin, which are enhanced by cell-cell adhesion interactions (Edenius et al., 1991; Tornhamre et al., 1992). Lipoxin generation can also be mediated by single cells. In this model, primed neutrophils in inflammatory disorders such as asthma generate lipoxin entirely from endogenous sources of arachidonic acid from a single cell type (Chavis et al., 1995, 1996; Thomas et al., 1995).
(D) Aspirin-triggered lipoxins
Aspirin may also play an important role in the generation of lipoxins (Claria and Serhan, 1995). In this transcellular biosynthetic scheme, cyclooxygenase-2 (COX-2) switches its catalytic activity in the presence of aspirin, generating 15R-HETE instead of prostaglandins. Thus, aspirin inhibits prostaglandin biosynthesis by both COX-1 and COX-2 (Herschman, 1996). COX-2, when acetylated by aspirin in endothelial or epithelial cells, is enzymatically active and converts arachidonic acid to 15R-HETE, which is released and transformed through transcellular routes to form 15-epi-lipoxins by leukocytes. The activated and adherent leukocytes possess 5-LO and transform 15R-HETE to a 5(S)-epoxytetraene within leukocytes, which carries its C15 position alcohol in the R configuration. This intermediate leads to the formation of both 15-epi-LXA4 and 15-epi-LXB4, which carry the R configuration at C15. 15-epi-LXA4 is more potent than native LXA4 in inhibiting neutrophil adhesion, and 15-epi LXB4 inhibits cell proliferation (Claria and Serhan, 1995; Serhan et al., 1995). COX-2 is increased in inflammatory reactions and in disease states including colon cancer, and is normalized upon aspirin intake (Herschman, 1996; Levy, 1997). Aspirin-triggered lipoxins (ATL) can serve as potential endogenous anti-inflammatory signals or mediators of some of aspirin’s beneficial actions. These beneficial actions of aspirin include prevention of myocardial infarction and protection from colorectal adenoma, as well as other forms of cancer (Giovannucci et al., 1995; Savage et al., 1995; Levy, 1997). Thus, in addition to serving as an inhibitor of eicosanoid biosynthesis, aspirin can also trigger the biosynthesis of various compounds, such as the endogenous generation of 15-epimeric lipoxins, via acetylation of COX-2 at sites of inflammation in vivo (Vane, 1971; Claria and Serhan, 1995; Chiang et al., 1998; Fierro et al., 2002). Meanwhile, aspirin also affects COX-2-bearing vascular endothelial cells or epithelial cells and their co-activation with neutrophils. In this model, inflammatory cytokines induce COX-2 to generate 15R-HETE when aspirin is administered (Claria and Serhan, 1995). This intermediate carries a C15 alcohol in the R configuration that is rapidly converted by activated neutrophils to 15-epi-lipoxins rather than 15S native lipoxins, which result from LO interactions (Serhan, 1997). These actions of LXA4 and ATL, their endogenous epimeric counterpart (15-epi-LX), were first found in experiments with isolated cell types in vitro and have been confirmed and demonstrated in several acute murine models of inflammation and second-organ reperfusion injury (Serhan et al., 1995; Takano et al., 1997, 1998; Chiang et al., 1998, 1999; Clish et al., 1999; Hachicha et al., 1999; Bandeira-Melo et al., 2000).
Neutrophil infiltration in lung, skin, and sites of wound healing is dramatically inhibited by both intravenous and topical application of stable analogs of both LXA4 and aspirin-triggered 15-epi-LXA4 (ATL). 15-epi-LXA4 and LXA4 analogs inhibit interleukin-1β (IL-lβ), tumor necrosis factor-α (TNF-α), and IL-8 expression while stimulating IL-4 release in vivo, and interact with a common receptor on human and murine leukocytes (Gewirtz et al., 1998; Hachicha et al., 1999). Bioactive ATL and LXA4 analogs compete with [3H]-LXA4 binding to LXA4 receptors (Takano et al., 1997). Thus, these inhibitory actions of LX and ATL analogs are likely to be mediated by specific lipoxin receptors present in rodent and human cells (Gronert et al., 2001). LXB4 is a positional isomer of LXA4, carrying alcohol groups at carbon 5S, 14R, and 15S positions, instead of the C-5S, 6R, and 15S positions present in LXA4. Aspirin-triggered LXB4 carries a 15R alcohol, hence 15-epi-LXB4. Although LXA4 and LXB4 show similar activities in some biologic systems, in many others they have distinct actions (Tamaoki et al., 1995; Serhan, 1997).
(IV) Lipoxins and Disease
Lipoxins are generated in human organs and are associated with a variety of inflammatory events (Table). The first demonstration of lipoxin in clinical inflammation was in bronchial lavage (Lee et al., 1990). Furthermore, LXA4 and LXB4 are formed in nasal polyps, and LXA4 is generated in nasal lavage from aspirin-sensitive asthmatics and in experimental nephritis (Edenius et al., 1990; Brezinski et al., 1992; Papayianni et al., 1995). It has also been proposed that lipoxins are useful biomarkers of asthma and long-term clinical improvement in arthritic patients (Thomas et al., 1995; Chavis et al., 2000). Rupture of atherosclerotic plaque leads to rapid generation of LXA4 in the intracoronary artery (Brezinski et al., 1992). Lipoxins are also generated by normal human bone marrow, and, during chronic myelocytic leukemia, platelets lose 12-LO. They lose their ability to generate lipoxins, and this finding may be related to the blast crisis observed in chronic myelocytic leukemia (Stenke et al., 1994). To determine whether 15-epi-LXA4 could be detected in animal experimental models or in patient-derived materials, experiments were carried out with a mouse peritonitis model (Chiang et al., 1998). In this model, COX-2 protein levels were up-regulated by intraperitoneal injection of lipopolysaccharide (LPS) aspirin. The results revealed a low level of 15-epi-LXA4 generation in these LPS-treated animals, and the presence or absence of aspirin did not make any significant difference. This study suggested that LPS alone is not sufficient to elicit neutrophil infiltration. Furthermore, these experiments also demonstrated that aspirin administration in murine peritonitis yields inflammatory exudates that generate 15-epi-LXA4 in appreciable levels from endogenous substrates.
In humans, ATL and LXA4 formation was studied in aspirin-tolerant and aspirin-intolerant asthmatics. The aspirin-tolerant subjects generated both LXA4 and ATL, but the aspirin-intolerant patients proved to have a diminished capacity to generate ATL and lipoxins upon aspirin challenge (Sanak et al., 2000). Furthermore, a reduction and alteration in lipoxin generation were found in patients with chronic liver disease and chronic myelogenous leukemia (Stenke et al., 1991; Claria et al., 1998). In contrast to these disease states, LXA4 production is up-regulated, following atherosclerotic plaque rupture, and with nasal polyps (Edenius et al., 1990; Brezinski et al., 1992).
(A) Lipoxins in periodontal diseases
Neutrophils are within the first line of host defense, and, by their ability to phagocytose microbes, they can protect the host from infection. They can also give rise to neutrophil-dependent vascular injury and contribute to increased vascular permeability, edema, and further release of chemoattractants, with a net pro-inflammatory effect. The involvement of the inducible cyclooxygenase isoform (COX-2) and the role of novel lipid mediators in the pathogenesis of periodontal disease are under study, and data derived from these observations have showed that periodontitis represents an important inflammatory model for the investigation of lipid mediators. The working hypothesis is that COX-2 could have multiple role(s) in the development and progression of the periodontal disease (Pouliot et al., 2000). First, crevicular fluid samples from localized aggressive periodontitis (LAP) patients were examined and found to contain prostaglandin PGE2 and 5-LO-derived products, LTB4, and the biosynthesis interaction product, lipoxin LXA4. As opposed to early suggestions that monocytes and macrophages were the major source of PGE2 production in periodontal disease (Dewhirst et al., 1983), neutrophils were found to generate considerable arachidonic acid metabolites in this study (Pouliot et al., 2000). This finding suggests that neutrophils contribute to the pathogenesis of periodontal disease in ways that were not previously anticipated. Furthermore, neutrophils from peripheral blood of LAP patients, but not from healthy volunteers, also generated LXA4, suggesting that this immunomodulatory molecule may also have a role in periodontal disease (Pouliot et al., 2000).
The role of lipid mediators in the neutrophil response to Porphyromonas gingivalis was also characterized in an animal model. When P. gingivalis was introduced into murine dorsal air pouches, leukocyte infiltration was initiated. Elevated PGE2 levels in the cellular exudate and up-regulated COX-2 expression in the leukocyte infiltrate accompanied neutrophil accumulation. In addition, human neutrophils exposed to P. gingivalis also demonstrated up-regulation of COX-2 mRNA expression. Interestingly, P. gingivalis introduced into the epithelium-lined air pouch caused significant increases in the murine tissue levels of COX-2 mRNA associated with both heart and lung, supporting a potential role for this oral pathogen in the evolution of systemic events. The administration of metabolically stable analogues of lipoxin and of aspirin-triggered lipoxin potently blocked neutrophil traffic into the dorsal pouch cavity and lowered PGE2 levels within exudates without allowing infection to spread. These results show that neutrophils can provide an important source of PGE2 in periodontal tissues. Moreover, they provide strong support for the notion that lipoxin can have a protective role in periodontitis, limiting further neutrophil recruitment and neutrophil-mediated tissue injury that can lead to loss of inflammatory barriers that prevent tissue invasion by oral microbial pathogens (Pouliot et al., 2000). It has been further proposed that lipoxin generation and its relationship to PGE2 and LTB4 can be important markers for the pathogenesis of periodontal disease. Indeed, activated neutrophils from LAP patients produced lipoxin, whereas healthy neutrophils did not.
(B) Lipoxins as potential drugs
Lipoxin-stable analogs were studied for their clinical inhibition of inflammation. Several LXA4 and ATL stable analogs were synthesized by means of a recombinant dehydrogenase screen assay as a relatively inexpensive and rapid screen to design suitable analogs that might function in vivo. These analogs were generated by total organic synthesis and were tested for their ability to inhibit neutrophil infiltration and changes in vascular permeability in vivo in several murine models. Methylated 15(R/S)-LXA4 is an analog of both the aspirin-triggered 15-epi-LXA4 and native LXA4. Likewise, 16-phenoxy-LXA4, which has a phenoxy group at the C-16 position, is an analog of native LXA4 that prevents enzymatic inactivation with recombinant 15-prostaglandin dehydrogenase in vitro (Serhan et al., 1995). These analogs have been shown to act by competition for the LXA4 receptor (Chiang et al., 2000).
When applied topically to mouse ears, the lipoxin-stable analogs inhibit both neutrophil infiltration and vascular permeability changes in a concentration-dependent fashion (Takano et al., 1997, 1998). The inhibition of vascular permeability changes paralleled inhibition of neutrophil infiltration with both the ATL and lipoxin analogs. In addition, the impact of LXB4 analogs that resist enzymatic inactivation was also tested in vitro (Maddox et al., 1998). Of the lipoxin-stable analogs tested in this inflammatory model, inhibitory actions of 15(R/S)-methyl-LXA4 on neutrophil infiltration and vascular permeability changes were significantly greater than those of topically applied native LXA4. In addition, a 16-para-fluoro derivative of 16-phenoxy-LXA4 was also prepared to assess whether fluorination of the phenoxy ring could enhance potency. Results indicate that 16-phenoxy-LXA4 was also potent and retained activity at levels comparable with those of 16-parafluorophenoxy-LXA4. Both LXB4 analogs inhibited neutrophil infiltration and vascular permeability. These also proved to be potent inhibitors of neutrophil infiltration in the dorsal air pouch (Serhan, 2001).
(V) Conclusion
Lipoxins and ATL appear to be the first recognized members of a new class of endogenous mediator that are anti-inflammatory or serve for the "pro-resolution" of inflammation. PGE2 can and may display anti-inflammatory properties in certain settings, but in most cases, it enhances inflammation in vivo. This is likely the result of numerous receptor isoforms and differential coupled mechanisms for PGE2 and its diverse role in human physiology. Since the integrated response of the host is essential to health and disease, it is important to achieve a more complete understanding of the molecular and cellular events governing the formation and actions of endogenous mediators of resolution that appear to control the magnitude and duration of inflammation. In view of the present body of evidence, it is not surprising that a protective action for inhibition of COX-2 was found in cardiovascular disease. Characterizing useful experimental systems with clinically relevant endpoints will also take a multidisciplinary approach and require a shift in our current thinking about inflammation and the role of lipid mediators.
Neutrophils play an important role in the mediation of inflammation. These cells, previously regarded as solely generators of enzymes and thus contributors of enzymatic killing mechanisms against pathogens, are gaining more significance in terms of their role at various stages of inflammation. In addition to the recent findings on the generation of lipid mediators of inflammation, persistent activity of neutrophils also contributes to the tissue destruction. Through this mechanism, the protective role of these cells turns into a deleterious action targeting the host itself. Periodontal disease is an important model in which various aspects of lipid mediators can be studied. It also represents a local phenomenon in which neutrophil-mediated tissue injury is shown to be a major contributing factor for progression of the disease. Therefore, studies aimed to elucidate the pathogenetic mechanisms in periodontal disease will likely add to our understanding of various disease sequelae as well. Within this context, lipoxins and lipid mediators not only present an exciting area of research, but also have potential for the development of novel treatment strategies.
Arachidonic acid metabolism and generation of eicosanoids. Phospholipase A2 (PLA2) catalyzes phosphatidyl choline into arachidonic acid when an agonist stimulates a G-protein-coupled receptor on the cell surface. Arachidonic acid is the key molecule for generation of lipid mediators of inflammation. Prostanoids (prostaglandins, thromboxanes) and prostacyclins are produced through cyclooxygenase (COX) activity, while leukotrienes are generated by the action of 5-lipoxygenase (LO). The lipoxygenase:lipoxygenase (LO:LO) pathway—activity of a combination of two LO, 5, 12, or 15 on the same arachidonic acid molecule—produces lipoxins. Aspirin can trigger the production of epimeric lipoxins through acetylation of COX-2, yielding 15-epi-H(p)ETEs, which are then converted to lipoxins by transcellular biosynthesis using 5-LO from neutrophils.
Footnotes
Acknowledgements
This manuscript is supported partially by USPHS grant #DE13499.