The seven-transmembrane G protein-coupled receptors activated by leukotrienes are divided into two subclasses based on their ligand specificity for either leukotriene B4 or the cysteinyl leukotrienes (LTC4, LTD4, and LTE4). These receptors have been designated BLT and CysLT receptors, respectively, and a subdivision into BLT1 and BLT2 receptors and CysLT1 and CysLT2 receptors has been established. However, recent findings have also indicated the existence of putative additional leukotriene receptor subtypes. Furthermore, other ligands interact with the leukotriene receptors. Finally, leukotrienes may also activate other receptor classes, such as purinergic receptors. The aim of this review is to provide an update on the pharmacology, expression patterns, and pathophysiological roles of the leukotriene receptors as well as the therapeutic developments in this area of research.
I. Introduction: Leukotriene Receptors
The seven-transmembrane G protein-coupled receptors (GPCRs1) activated by leukotrienes are divided into two subclasses based on their ligand specificity for either leukotriene B4 (LTB4) or the cysteinyl leukotrienes (LTC4, LTD4, and LTE4). The leukotriene receptors belong to the A5 subfamily of the rhodopsin receptor-like GPCRs (Joost and Methner, 2002). According to the previous report from the International Union of Basic and Clinical Pharmacology's (IUPHAR) Nomenclature Committee for Leukotriene Receptors, these receptors have been designated BLT and CysLT receptors, respectively (Brink et al., 2003). Furthermore, a subdivision into BLT1 and BLT2 receptors and CysLT1 and CysLT2 receptors has been established, as indicated in Table 1 (Brink et al., 2003).
However, several reports have also indicated the existence of additional leukotriene receptor subtypes (Rovati et al., 1997; Bäck, 2002; Norel and Brink, 2004; Rovati and Capra, 2007; Austen et al., 2009). In addition, evidence has emerged that ligands other than leukotrienes interact with the leukotriene receptors and that leukotrienes may also signal through receptors preferentially activated by other endogenous ligands. For example, lipid mediators that are structurally different from leukotrienes have been reported to signal through BLT receptors (Arita et al., 2007; Okuno et al., 2008). As described later, 12(S)-hydroxyheptadeca-5Z,8E,10E-trienoic acid (12-HHT) was identified as a more potent ligand for BLT2R than for LTB4 (Okuno et al., 2008). Furthermore, although evidence that LTE4 signals through the P2Y12 receptor was demonstrated (Nonaka et al., 2005; Paruchuri et al., 2009; Fredman et al., 2010), one of the orphan GPCRs, GPR17, has been postulated to be activated by both cysteinyl leukotrienes and nucleotides (Ciana et al., 2006). Finally, both the CysLT1 receptor (CysLT1R) antagonists in clinical use as antiasthmatic drugs have been demonstrated to inhibit the effects of nucleotides acting at different P2Y receptors. These observations demonstrate that these classic leukotriene receptor antagonists (LTRAs) can inhibit non-CysLT1-mediated proinflammatory reactions, suggesting activities potentially relevant for interpatient variability in response to treatment. The aim of the present review is to provide an extensive update on the pharmacology, expression patterns, and pathophysiological roles of the leukotriene receptors. Furthermore, a comment on the potential therapeutic developments in this area of research will be presented.
A. Brief Historical Background of Leukotrienes and Their Receptors
Ever since the identification of the biochemical structure of the leukotrienes and their association with inflammation (Samuelsson, 1983), the pathophysiological role of these lipid mediators has been explored in many experimental approaches (Brink et al., 2003). Initially, LTB4 was appreciated for its potent chemotactic effects on neutrophil granulocytes (Ford-Hutchinson et al., 1980; Malmsten et al., 1980; Palmblad et al., 1981) and its ability to serve as a complete secretagog for human peripheral blood neutrophils (Serhan et al., 1982). In contrast, initial studies of cysteinyl leukotrienes were mainly focused on their potent bronchoconstrictive effects (Dahlén et al., 1980). However, subsequent studies have extended these findings and associated leukotriene receptor signaling with several pathophysiological processes.
The first leukotriene receptor to be cloned was the human BLT1, and the molecular structure was reported in 1997 (Yokomizo et al., 1997). This molecular identification of the BLT1 receptor (BLT1R) permitted the characterization of functional BLT1R in several leukocyte populations (Kim and Luster, 2007) and also in nonmyeloid cells, such as vascular smooth muscle and endothelial cells (Bäck et al., 2005). Although the LTB4-induced interaction with BLT1R corresponded with several effects observed in those target cells, initial studies had revealed both high- and lower-affinity binding sites for LTB4 specifically in human granulocytes (Goldman and Goetzl, 1984). The molecular explanation for the latter finding was provided in 2000, when a gene with high sequence similarity to BLT1R was identified and encoded a low-affinity LTB4 receptor, which has subsequently been denoted BLT2R (Yokomizo et al., 2000; Brink et al., 2003). Transgenic overexpression of the human BLT1R in mice increased the inflammatory response to LTB4 (Chiang et al., 1999), whereas genetic BLT1R disruption decreased leukocyte chemotaxis and protected against disease development in response to several different proinflammatory stimuli (Haribabu et al., 2000; Tager et al., 2000; Brink et al., 2003) (see section II.E).
The classification of the CysLT receptors was initially based on the differential profiles of guinea pig trachea contractions induced by either LTC4 or LTD4 (Brink et al., 2003). Because a classification based on agonist selectivity is not generally applicable, the names CysLT1 and CysLT2 were introduced when referring to responses being sensitive and resistant (Labat et al., 1992) to the class of CysLT receptor antagonists, which had been developed for the treatment of asthma (Brink et al., 2003). The subsequent cloning of these receptors in 1999 and 2000 (Lynch et al., 1999; Sarau et al., 1999; Heise et al., 2000; Nothacker et al., 2000; Takasaki et al., 2000) largely confirmed these pharmacological characteristics of the CysLT1 and CysLT2 receptors.
B. Chemical Structure of Endogenous Leukotriene Receptor Agonist
Most endogenous ligands for the leukotriene receptors are derived from lipoxygenase metabolism of arachidonic acid and are shown in Fig. 1. The formation of leukotrienes by 5-lipoxygenase (5-LO) involves the oxygenation at carbon number 5 of arachidonic acid to form 5-hydroperoxyeicosatetraenoic acid (HpETE). Subsequently, removal of hydrogen at carbon 10 from 5-HpETE leads to formation of the epoxide intermediate LTA4, which serves as precursor for LT synthesis (Shimizu et al., 1984; Rådmark and Samuelsson, 2009) (Fig. 1). Although detected in the cytosolic or nucleosolic fraction of resting cells, cellular activation leads to 5-LO translocation to the nuclear envelope in a calcium-dependent manner, where colocalization with the 5-LO-activating protein (FLAP) is a prerequisite for LT synthesis (Evans et al., 2008).
The enzyme LTA4 hydrolase is located in the cytosol and stereospecifically adds water to carbon 12 of LTA4, leading to formation of LTB4 (Haeggström, 2004). Several inactivating pathways for LTB4 are known, of which ω-oxidation has been most extensively studied in human leukocytes (for review, see Yokomizo et al., 2001b). In this pathway, LTB4 is converted to 20-hydroxy-LTB4 by LTB4 ω-hydroxylase and then to 20-carboxy-LTB4 as indicated in Fig. 1 (Yokomizo et al., 2001b). The latter metabolites may also act as ligands at the BLT receptors, which will be discussed below.
The other metabolic pathway of LTA4 involves the enzyme LTC4 synthase (LTC4S), a microsomal glutathione transferase, which conjugates LTA4 with glutathione to form LTC4 (Fig. 1) (Austen, 2007). After transportation to the extracellular space by an ATP-dependent transporter (multidrug resistance-associated protein) that recognizes its glutathione moiety, LTC4 is metabolized by γ-glutamyl transpeptidase into LTD4. Subsequently, a serum dipeptidase, which cleaves the peptide bond between the cysteinyl and glycyl residues in the LTD4 side chain, leads to the formation of LTE4 (Fig. 1). Because LTC4, D4, and E4 all contain a cysteinyl group at carbon number 6 (Fig. 1), these LTs are referred to as the cysteinyl leukotrienes. Thus, the receptors for these LTs are termed CysLT receptors (Brink et al., 2003).
In addition to 5-LO metabolites, products of the 12- and 15-lipoxygenase metabolism of arachidonic acid metabolism can act as ligands for the LT receptors (Fig. 1). For example, 12(R)-hydroxyeicosatetraenoic acid (HETE) competes with LTB4 binding at BLT1R (Yokomizo et al., 1997, 2001c), whereas 12(S)-HETE, 12(S)-HpETE, and 15(S)-HETE activate BLT2R (Yokomizo et al., 2001c). The orphan receptor GPR31 has been suggested as a high-affinity receptor for 12(S)-HETE (Guo et al., 2011). Finally, 12-HHT (Fig. 1), which is formed through the cyclooxygenase pathway of arachidonic acid metabolism is a high-affinity BLT2R agonist (Okuno et al., 2008).
II. BLT Receptors
A. BLT Receptor Subtypes
1. BLT1 Receptor.
The open reading frame of the human LTB4R gene encodes the BLT1R. The BLT1R protein consists of 352 aa (NCBI Reference Sequence: NP_858043) (Fig. 2). The BLT1R has also been isolated in other species, with a homology similar to that of the human protein in mice (78%; Huang et al., 1998), in guinea pigs (78%; Masuda et al., 1999), and in rats (80%; Toda et al., 1999).
Although LTB4 seems to be the sole full agonist at BLT1R, binding studies in membrane fractions of BLT1R-transfected cells have shown significant competitions for 20-OH-LTB4, 12-oxo-LTB4 and 12-epi-LTB4, 20-COOH-LTB4, and 12(R)-HETE (Yokomizo et al., 1997, 2001c). However, not all of these ligands seem to induce similar intracellular signaling. For example, studies of the recombinant guinea pig BLT1R have shown that although 20-OH-LTB4 was less potent than LTB4 in the mobilization of intracellular Ca2+ ([Ca2+]i), these agonists were equally potent in inhibiting cAMP formation (Masuda et al., 1999). In addition, the eicosapentaenoic acid metabolite resolvin (Rv) E1 was identified as a partial BLT1R agonist. In particular, RvE1 selectively inhibited adenylate cyclase and attenuated LTB4-induced nuclear factor κB (NF-κB) activation in BLT1- but not in BLT2-transfected cells, whereas in human peripheral blood mononuclear cells RvE1 partially induced calcium mobilization and blocked subsequent stimulation by LTB4 (Arita et al., 2007). These data thus suggest that this mediator, besides activating the receptor ChemR23, could act through inhibition of LTB4-induced signaling. More information on this new genus of specialized proresolving mediators has been published recently (Bannenberg and Serhan, 2010).
Initially, BLT1R expression was reported to be restricted to phagocytic leukocytes (Yokomizo et al., 1997), and subsequent studies identified BLT1R expression in granulocytes, monocytes, and dendritic cells. In addition, lymphocytes also express BLT1R (see section II.D), and functional BLT1R are now known to be expressed in nonmyeloid cells, such as vascular smooth muscle cells (SMCs) (Bäck et al., 2005), as well as endothelial cells (ECs) (Qiu et al., 2006), skeletal muscle satellite cells (Sun et al., 2009), and neural stem cells (Wada et al., 2006).
2. BLT2 Receptor.
In the analysis of human and mouse LTB4R genes, an open reading frame encoding a putative seven-transmembrane-type receptor with sequence similarities to BLT1R was identified (Yokomizo et al., 2000). This gene was shown to encode a receptor protein that exhibited LTB4 binding with a 20-fold higher Kd, and calcium signaling with a 30-fold higher EC50 value compared with BLT1R (Yokomizo et al., 2000). The gene has been designated LTB4R2, and the receptor has been classified as the BLT2 receptor (BLT2R) (Brink et al., 2003). The BLT2R protein consists of 358 aa (NCBI Reference Sequence: NP_062813) and exhibits a 36 to 45% aa identity with the human BLT1R (Kamohara et al., 2000; Tryselius et al., 2000; Yokomizo et al., 2000). The murine BLT2R has a 92% aa homology with the human receptor protein, which is a higher homology compared with that of BLT1R (78% aa identity between murine and human proteins; Iizuka et al., 2005). These data suggest that the BLT2R has been conserved during evolution.
In contrast to the relatively specific binding of LTB4 at BLT1R, several lipoxygenase products in addition to LTB4 have been identified as ligands for BLT2R. These include 12(S)-HETE, 12(S)-HpETE, and 15(S)-HETE (Fig. 1) (Yokomizo et al., 2001c). Furthermore, the thromboxane synthase product 12-HHT formed in activated blood platelets and macrophages from prostaglandin H2 (Fig. 1) is also a natural ligand for BLT2R (Okuno et al., 2008). In the latter study, the EC50 values of 12-HHT and LTB4 for [Ca2+]i mobilization in CHO-BLT2R cells were 19 and 142 nM, respectively, whereas 12-HHT failed to induce calcium mobilization in CHO-BLT1R cells (Okuno et al., 2008). Several binding studies further demonstrated that BLT2R is a high-affinity receptor for 12-HHT and that 12-HHT and LTB4 occupy the same binding site on BLT2R (Okuno et al., 2008). A synthetic selective BLT2R agonist has also been reported and termed compound A (Iizuka et al., 2005).
In CHO cells expressing the human BLT2R, maximal chemotaxis was observed at LTB4 concentrations greater than 1 μM, compared with 10 nM for CHO cells expressing human BLT1R (Yokomizo et al., 2001a; Okuno et al., 2008). Data suggest that coexpression of both human BLT1 and BLT2 receptors makes CHO cells migrate toward both very low and high LTB4 concentrations (Yokomizo et al., 2001a). Other hydroxyeicosanoids acting as BLT2R agonists, such as 12(S)-HETE, 12(R)-HETE, and 12-epi-LTB4, require even higher concentrations (approximately 10 μM) to induce a maximal chemotactic response (Yokomizo et al., 2001c). In contrast, maximal chemotaxis for 12-HHT in CHO cells expressing the human BLT2R is observed at 30 nM, suggesting that this high-affinity BLT2R ligand is coupled to chemotaxis (Okuno et al., 2008). In line with findings in CHO cells, murine BLT2R-expressing 300.19 cells exhibit higher affinity for 12-HHT than for LTB4. 12-HHT-induced intracellular calcium flux with the same efficacy as LTB4 (Mathis et al., 2010). However, in the latter cells, 12-HHT was only a weak agonist for chemotaxis with 3% activity relative to LTB4 (Mathis et al., 2010), suggesting a ligand-specific coupling to chemotaxis. Studies of primary cells have generated variable results as to the role of BLT2R in LTB4-induced chemotaxis (see section II.D).
The most abundant BLT2R expression is found in the spleen, followed by liver, ovary, and leukocytes, and BLT2R expression is also detected in several other organs, suggesting a ubiquitous expression pattern, at least in human tissues (Yokomizo et al., 2000). The tissue distribution of BLT2R is different in mouse tissues, with the highest expression in small intestine followed by colon and skin (Iizuka et al., 2005). Physiological and pathophysiological responses associated with BLT2R signaling include macrophage and mast cell (MC) chemotaxis (Subbarao et al., 2004; Lundeen et al., 2006; Okuno et al., 2008), guinea pig lung parenchyma contraction (Sakata et al., 2004), and mouse models of angiogenesis (Kim et al., 2009), itch-associated scratching (Kim et al., 2008), colitis (Iizuka et al., 2010), and arthritis (Mathis et al., 2010).
3. Genomic Organization and Transcriptional Regulation.
The genes encoding BLT1R and BLT2R are designated LTB4R and LTB4R2, respectively (Table 1), and are located within 10 kilobase pairs of each other in both the human and murine genome (Yokomizo et al., 2000). The genes encoding the BLT receptors form a gene cluster on human chromosome 14q11.2-q12 with the open reading frame of the LTB4R2 gene overlapping one of the promoter regions of the human LTB4R gene (Kato et al., 2000; Yokomizo et al., 2000; Brink et al., 2003). The LTB4R gene consists of three exons, and, in line with other receptors for chemoattractants such as fMet-Leu-Phe and interleukin (IL)-8, the open reading frame is intronless (Kato et al., 2000). One of the LTB4R promoters has been shown to contain consensus sequence for NF-κB and Sp1 binding (Kato et al., 2000). Spontaneous mutations around the LTB4R and LTB4R2 genes have been reported to be associated with cerebrovascular disease (Bevan et al., 2008), but the close proximity of the genes encoding those receptors suggests that those polymorphisms cannot be ascribed to either of the genes alone.
In THP-1 cells, the deletion of a Sp1 binding site in the LTB4R promoter decreases transcriptional activity (Kato et al., 2000). In human monocytes, BLT1R mRNA expression is down-regulated by proinflammatory stimuli, such as lipopolysaccharide (LPS), interferon (IFN)-γ, and tumor necrosis factor (TNF)-α, whereas the anti-inflammatory cytokine IL-10 and dexamethasone up-regulate BLT1R mRNA in these cells (Pettersson et al., 2005). Likewise, in human neutrophil granulocytes, BLT1R mRNA is up-regulated by dexamethasone (Stankova et al., 2002), and BLT1R protein and LTB4 binding are decreased by TNF-α and LPS (O'Flaherty et al., 1991; Pettersson et al., 2003). The latter findings support earlier studies demonstrating decreased LTB4 binding and LTB4-induced chemotaxis in neutrophils derived from endotoxemic rabbits (Hartiala et al., 1985; Goldman et al., 1986). Of interest, the transcriptional regulation of the LTB4R gene may vary between cells of hematopoietic and nonhematopoietic origin (Kato et al., 2000). In contrast to findings in leukocytes, BLT1R expression is up-regulated by LPS, IL-1β, and IFN-γ in vascular SMCs, as demonstrated by studies of cells derived from human, rat, or mouse vessels (Bäck et al., 2005; Heller et al., 2005). Furthermore, vascular injury in vivo significantly up-regulates BLT1R mRNA through NF-κB signaling (Bäck et al., 2005). Likewise, whereas BLT1R is expressed at low levels in human umbilical cord endothelial cells (HUVECs) under resting conditions, either LPS or IL-1β induces BLT1R mRNA and protein levels and renders HUVECs responsive to LTB4 (Qiu et al., 2006). This cell-specific expression of BLT1R may be dependent on epigenetic regulation, for example, through methylation of the promoter region, which has been shown to regulate BLT1R transcription in vitro (Kato et al., 2000). In a recent study, the enhancer in the human LTB4R, termed AE-BLex, was discovered at the intron I-exon-II boundary (Hashidate et al., 2010). AE-BLex possesses two acute myeloid leukemia 1 (AML1, also known as Runx1) recognition sites. The AML1/AE-BLex complex was confirmed in several BLT1R-expressing leukemia cell lines and human peripheral leukocytes. Thus, AML1 enhances BLT1R expression by binding to AE-BLex, which is accessible in leukocytes. The enhancement of the BLT1R expression in leukocytes is due to a loosening of the chromatin structure around AE-BLex, leading to the incremental binding of AML1 (Hashidate et al., 2010).
B. Structure-Function Relationships for BLT Receptors
LTB4 binds near the extracellular surface of BLT1R, and although polar residues within transmembrane domains (TMs) III, V, and VI and extracellular loop 2 are critical for ligand binding, polar residues in TMs II, III, and VII play a central role in transducing the ligand-induced conformational change to activation (Sabirsh et al., 2006; Basu et al., 2007) (Fig. 2). Furthermore, a conformational change of TM IV takes place during ligand binding (Baneres et al., 2003).
Site-directed mutagenesis has revealed several of the residues involved in LTB4 binding to BLT1R (Fig. 2). Although initial studies had suggested that mutating all receptor cysteines (except Cys90 and Cys168) to serine did not alter ligand binding, selective replacement of Cys97 in the TM III domain to serine leads to a 100-fold decrease in the affinity for the agonist (Mesnier and Banères, 2004). Furthermore, residues His94, Tyr102, Val105, and Ile108 in TM III, Asn241 in TM IV, and Arg178 and Glu185 in TM IV have been shown to be necessary for ligand binding (Sabirsh et al., 2006; Basu et al., 2007) (Fig. 2). In addition to the TM domains, the second extracellular loop of BLT1R may play an even more critical role for LTB4 binding as suggested by the complete loss of ligand binding after mutation of Arg156 to alanine (Basu et al., 2007) (Fig. 2). In line with the homology between the BLT1 and BLT2 receptors (see section II.A), most of the binding site residues described above suggest a common binding mode for LTB4 at the two BLT receptor subtypes (Basu et al., 2007). However, the presence of Tyr94 in BLT2R at the place of His94 within BLT1R has been suggested to account for some of the differences in the binding affinities and agonist selectivity between these receptors (Basu et al., 2007).
Structural models for the ligand-free and ligand-bound states of BLT1R have revealed an activation core formed around Asp64, and mutagenesis toward Asn36, Ser100, and Asn281 and a triad of serines, Ser277, Ser278, and Ser279 (Fig. 2), resulted in loss of signaling capacity, whereas mutant receptors retained normal LTB4 binding (Basu et al., 2007). Furthermore, mutating the third intracellular loop, which consists of the cytoplasmic end of TM V (Fig. 2), reduced Gi-dependent signaling associated with a loss of the high-affinity LTB4 binding state (Kuniyeda et al., 2007).
The C-tail containing a putative cytoplasmic helical domain, helix 8, of the BLT receptors has recently received attention. For example, BLT1R with a mutated helix 8 exhibit increased LTB4 binding as well as prolonged [Ca2+]i mobilization and cellular metabolic activation compared with the WT receptor, suggesting that helix 8 might work as a scaffold for G proteins, changing BLT1R to a low-affinity state leading to release of the ligand and signal shutdown (Okuno et al., 2003). In contrast, helix 8 may not be critical for either ligand binding or activation of BLT2R (Yasuda et al., 2009). However, human BLT2Rs lacking helix 8 accumulate in the endoplasmic reticulum, suggesting that helix 8 may be required for BLT2R protein folding and passage through the endoplasmic reticulum to the cell membrane (Yasuda et al., 2009).
Whereas cells expressing a complete C-tail-truncated BLT1R (G291stop, Fig. 2) exhibit increased numbers of binding sites and increased signal transduction compared with wild-type BLT1R (Gaudreau et al., 2004), a partially C-tail-truncated BLT1R mutant (G319stop, Fig. 2) has shown binding characteristics similar to WT BLT1R (Gaudreau et al., 2002). Furthermore, mutating a dileucine motif (Leu304-Leu305), suggested to be involved in a hydrophobic core (Fig. 2), mimicked the increased binding and signaling associated with the complete C-tail truncated receptor (Gaudreau et al., 2004). In addition, this latter mutation prevented LTB4-induced BLT1R internalization (Gaudreau et al., 2004), which corroborated the findings that the complete C-tail-truncated BLT1R lost the capacity to internalize through the involvement of GPCR kinases (GRK) 2 and 6 (Gaudreau et al., 2002; Chen et al., 2004). In addition, substitution of Thr308 within a putative casein kinase 2 site in the full-length BLT1R (Fig. 2) prevented most of GRK6-mediated inhibition of LTB4-induced inositol phosphate production while only partially affecting the LTB4-induced BLT1R phosphorylation (Gaudreau et al., 2002). Through the substitution of all Ser and Thr residues in the C-terminal tail with Ala (to generate a phosphorylation-defective mutant) or with Asp/Glu (to mimic constitutive phosphorylation), it has been shown that BLT1R phosphorylation may be an important mediator of G protein activation, whereas β-arrestin-associated BLT1R internalization seems to be independent of phosphorylation (Jala et al., 2005). For the murine BLT1R, a PKC consensus phosphorylation site has been reported to be located on the second intracellular loop, and a Ser127 substitution within this site prevented mBLT1R desensitization (Mollerup et al., 2007).
C. Intracellular Signaling Pathways and Second-Messenger Systems
The downstream intracellular signaling pathways after BLT1 and BLT2 receptor activation involve increased [Ca2+]i and inhibition of adenylyl cyclase (Brink et al., 2003). The G protein coupling for the BLT receptors may depend on the type of G proteins expressed in the different cells studied. In several cell types expressing endogenous BLT receptors, pertussis toxin (PTX) inhibits the increased [Ca2+]i in response to LTB4, suggesting coupling to Gi proteins (Brink et al., 2003). However, in BLT1R-transfected cells, a significant portion of the LTB4-induced [Ca2+]i response is resistant to PTX, suggesting also the involvement of Gq-protein coupling (Yokomizo et al., 1997; Haribabu et al., 1999; Brink et al., 2003; Sabirsh et al., 2004). These different profiles of PTX sensitivity suggest that BLT1R and BLT2R signal through different subtypes of Gi proteins (Yokomizo et al., 2000).
BLT receptor transduction also activates a number of kinases that phosphorylate downstream signal transduction proteins (Brink et al., 2003). For example, phosphorylation of mitogen-activated protein kinases (MAPKs) is involved in the LTB4-induced proliferation of RAW 264.7 macrophages and bronchial SMCs (Nieves and Moreno, 2006; Watanabe et al., 2009). In addition, extracellular signal-regulated kinases (ERKs) may also be involved in LTB4-induced proliferation, as demonstrated in vascular SMCs (Heller et al., 2005) and in signaling associated with delayed neutrophil apoptosis (Pétrin et al., 2006). The reported effects of different kinase inhibitors may also depend on which LTB4-induced response is studied. For example, wortmannin, a phosphatidylinositol 3-kinase (PI3K) inhibitor has been reported to block LTB4-induced chemotaxis, whereas calcium mobilization remains intact (Haribabu et al., 1999; Sabirsh et al., 2004). Tyrosine kinases have been implicated in LTB4-signaling in neutrophil granulocytes (Dryden et al., 1992) and BLT1-transfected HeLa cells (Sabirsh et al., 2004).
Effects on transcription have been implicated in BLT receptor intracellular signaling. For example, treatment of monocytes with LTB4 increased the transcriptional activation of the IL-6 gene (Brach et al., 1992). In this study, a reporter gene assay identified two restricted regions within the IL-6 promoter. These binding sites for NF-IL-6 and NF-κB conferred inducibility by LTB4. Exogenous addition of LTB4 increased the DNA binding of these transcription factors in monocytes (Brach et al., 1992; Huang et al., 2004) and induced nuclear translocation of NF-κB p65 in murine bone marrow-derived dendritic cells (Toda et al., 2010). The LTB4-induced NF-κB DNA binding activity was abolished by the BLT1R antagonist 1-[3-(4-phenyl-benzyl)-4-hydroxy-chroman-7-yl] cyclopentane carboxylic acid (CP-105,696) (Huang et al., 2004), suggesting that transcriptional alterations by LTB4 may be through BLT1R. One possible mechanism for this transcriptional activation by LTB4 could be through LTB4-induced intracellular H2O2 formation (Gagnon et al., 1989), because the oxidant scavenger N-acetyl-l-cysteine completely blocks LTB4-mediated transcription factor binding in monocytes (Brach et al., 1992). However, another study failed to observe any effects of other antioxidants on IL-6 protein secretion (Rola-Pleszczynski and Stanková, 1992). In the latter study, the partial inhibition of LTB4-induced IL-6 secretion by genistein suggested the involvement of tyrosine phosphorylation events during this process.
D. Receptor Distribution and Cellular Targets
1. Neutrophil Granulocytes.
The expression of both BLT1 and BLT2 receptor proteins in neutrophil granulocytes has, for example, been demonstrated by immunohistochemical analysis of surgical specimens derived from human abdominal aortic aneurysms (Houard et al., 2009). The potent neutrophil chemotactic activity of LTB4 through BLT1R signaling has been well established (Ford-Hutchinson et al., 1980; Malmsten et al., 1980; Palmblad et al., 1981). In neutrophils derived from the synovial fluid of patients with rheumatoid arthritis, BLT1 receptor antagonism inhibits LTB4-induced calcium mobilization and produces a rightward shift of the peak chemotactic response. In contrast, the BLT2R antagonist 1-[5-ethyl-2-hydroxy-4-[[6-methyl-6-(1H-tetrazol-5-yl)heptyl]oxy]phenyl]ethanone (LY255283) only inhibits [Ca2+]i and the chemotactic response to LTB4 at high concentrations, at which this compound may also act as an antagonist at BLT1R (Mathis et al., 2010). Studies in knockout mouse models have also shown that the LTB4-induced chemotaxis and calcium mobilization in zymosan-elicited murine peritoneal neutrophils derived from either WT or BLT2 knockout mice display similar LTB4 dose-response relations, whereas neutrophils from BLT1R knockout and BLT1R-BLT2R knockout mice lack both chemotaxis and alterations of [Ca2+]i in response to LTB4 (Haribabu et al., 2000; Tager et al., 2000; Terawaki et al., 2005; Mathis et al., 2010).
In addition to chemotaxis, LTB4 stimulates neutrophil release of lysozyme (Hafstrom et al., 1981), matrix metalloproteinases (MMP) (Kjeldsen et al., 1992), myeloperoxidase (Terawaki et al., 2005), elastase, α-defensins (Flamand et al., 2007), and azurocidin (Di Gennaro et al., 2009). Initial characterization of the binding of [3H]LTB4 in relation to the functional effects of LTB4 on human neutrophils suggested that whereas high-affinity binding sites mediated chemotaxis, this lysosomal degranulation occurred by LTB4 binding to low-affinity sites (Goldman and Goetzl, 1984). Subsequent results demonstrated that both subsets of LTB4 binding sites contributed to increases in [Ca2+]i induced by LTB4 (Goldman et al., 1985). In line with those findings, the neutrophil secretagog activity of LTB4 has consistently been observed at higher LTB4 concentrations compared with those needed to induce chemotaxis (Serhan et al., 1982). For example, maximal release of α-defensins from neutrophils was observed at 100 nM LTB4 (Flamand et al., 2007), and the release of azurocidin was gradually increased at LTB4 concentrations ranging from 10 nM to 1 μM (Di Gennaro et al., 2009). The latter apparent low-affinity response, however, was inhibited by the selective BLT1R antagonist CP-105,696, but not by the selective BLT2R antagonist LY255238. Despite the low-affinity profile, further support for lysosomal enzyme release being mediated through BLT1R rather than BLT2R signaling has emerged from studies of knockout mice. Although a low-affinity binding site for LTB4 has been characterized in neutrophils derived from BLT1R knockout mice, the myeloperoxidase release in response to LTB4 is also abolished in those cells (Terawaki et al., 2005), hence challenging the concept of lysosomal enzyme release through BLT2R signaling.
In addition to being a major neutrophil chemoattractant and promoting the release of lysosomal enzymes (Hafstrom et al., 1981; Serhan et al., 1982), prolonged exposure of neutrophils to LTB4 delays their constitutive apoptosis in a time- and concentration-dependent manner (Hébert et al., 1996). Furthermore, threshold levels of LTB4 (0.01 nM) induce surface expression of adhesion molecules, such as the β2-integrin CD11b and Mac-1, which have been implicated in LTB4-induced neutrophil adhesion to endothelial cells (Gimbrone et al., 1984; Showell et al., 1998). The antimicrobial neutrophil response may also be facilitated by LTB4 through increased surface expression of Toll-like receptors 7, 8, and 9 (Gaudreault and Gosselin, 2009) and stimulated secretion of the cathelicidin LL-37 (Flamand et al., 2007; Wan et al., 2007).
2. Eosinophil Granulocytes.
Western blot analysis has demonstrated BLT1R expression in eosinophils of both murine and human origin, with up-regulated expression levels in IL-5 transgenic mice and after inhalation of mold allergen (Huang et al., 1998). Consistent with findings in neutrophils, eosinophils derived from mice (Huang et al., 1998) and guinea pigs (Ng et al., 1991; Sun et al., 1991) also display a dose-dependent bell-shaped chemotactic response to LTB4 with an optimum concentration of 1 to 10 nM. However, species differences may exist in terms of LTB4-induced responses in eosinophils, because human cells may be less responsive to LTB4 when chemotaxis is monitored (Sun et al., 1991). Nevertheless, human blood-derived eosinophils and eosinophil-differentiated HL-60 cells exhibit an increase in intracellular calcium concentration upon stimulation with LTB4 (Patry et al., 1996; Murray et al., 2003). Furthermore, in line with findings in IL-5 transgenic mice (Huang et al., 1998), human eosinophils primed with IL-5 exhibit enhanced LTB4-induced chemotaxis (Sehmi et al., 1992). Priming with cytokines, such as IL-5, may therefore be a prerequisite for regulating the expression and responsiveness of BLT receptors on human eosinophils. In guinea pigs, LTB4 induced eosinophil superoxide generation in vitro (Ng et al., 1991). However, in contrast to findings in neutrophils, human eosinophil apoptosis was not altered by LTB4 (Murray et al., 2003).
There are also animal studies that support a role of LTB4 in regulating the recruitment of eosinophils into tissues in vivo. In guinea pigs, cutaneous LTB4 injection induces an accumulation of 111In-marked eosinophils (Faccioli et al., 1991) and oral treatment of sensitized guinea pigs with the selective BLT1R antagonist -[6-(3-hydroxy-1E,5Z-undecadienyl)-2-pyridinyl]-1,5-hexanediol (U75302) decreased eosinophil influx to the lung and BAL after antigen challenge (Richards et al., 1989). However, the studies of OVA sensitization and challenge in BLT1R-deficient mice have generated contradictory results, with both decreased eosinophil accumulation (Terawaki et al., 2005) and no alterations compared with results in WT mice (Miyahara et al., 2005a). In support of a pathophysiological role of LTB4-induced eosinophil activation, BLT1R-deficient mice exhibit decreased leukocyte content in peritoneal exudates compared with those in WT mice in response to thioglycollate-induced acute peritonitis (Tager et al., 2000). Furthermore, oral treatment with the BLT1R antagonist CP-105,696 blocked the recruitment of eosinophils into the spinal cord and completely inhibited the development of paralysis in experimental allergic encephalomyelitis (EAE), a murine model of multiple sclerosis (MS) (Gladue et al., 1996). The potential in vivo effects of either pharmacological or genetic BLT1R targeting also reflect indirect eosinophil activation, for example, through T lymphocyte-dependent responses and Th2 cytokines, such as IL-5 and IL-13 (Miyahara et al., 2005a,b; Terawaki et al., 2005). Finally, findings that blood eosinophil counts are unaltered in BLT1R knockout mice compared with those in WT mice have suggested that BLT1R is not necessary for granulocytopoiesis (Terawaki et al., 2005).
3. Basophil Granulocytes.
There is presently a paucity of information regarding the expression of BLT receptors on basophil granulocytes. In human blood-derived basophils, LTB4 induces a slight chemotactic response, which is more pronounced than that observed for PAF but less than that observed with C5a (Tanimoto et al., 1992). The receptor involved in the latter response, however, is presently unknown.
In human blood-derived monocytes, LTB4 induces a dose-dependent increase in [Ca2+]i with an ED50 of approximately 1.0 nM, which is inhibited by PTX (Rediske et al., 1993). Furthermore, BLT1R protein has been demonstrated on human blood-derived monocytes by fluorescence-activated cell sorting (Friedrich et al., 2003; Pettersson et al., 2003). An analysis of monocyte subpopulations suggested that BLT1R protein is more abundantly expressed in monocytes with high CD14 expression compared with CD16+ monocytes (Pettersson et al., 2005).
In addition to chemotaxis, BLT1R signaling in human monocytes also induced an up-regulation of integrin expression, involved in monocyte adhesion to endothelial cells in vitro and in vivo (Vaddi and Newton, 1994; Friedrich et al., 2003). Furthermore, LTB4-induced activation of human monocytes has been associated with the release of IL-1β (Rola-Pleszczynski and Lemaire, 1985), IL-6 (Brach et al., 1992; Rola-Pleszczynski and Stanková, 1992), and monocyte chemoattractant protein 1 (MCP-1) (Huang et al., 2004).
BLT1R expression has also been demonstrated on peritoneal and alveolar macrophages and in RAW 264.7 cells (Huang et al., 1998; Toda et al., 1999). In the latter macrophage cell line, LTB4 induces increased proliferation and [3H]thymidine uptake, suggesting that LTB4 may also have mitogenic effects on macrophages (Nieves and Moreno, 2006).
Early studies had shown that LTB4 enhanced macrophage bacterial phagocytosis in vitro (Demitsu et al., 1989). Subsequent results have demonstrated that this effect required the cells to be opsonized with IgG and that LTB4 enhanced only Fc receptor (FcR)-mediated phagocytosis (Mancuso and Peters-Golden, 2000), whereas the phagocytosis of apoptotic cells (efferocytosis) was not affected by LTB4 (Canetti et al., 2003). The signaling pathways involved in LTB4-enhanced FcR-dependent phagocytosis include Gαi3 protein (Peres et al., 2007), PKC activation (Mancuso and Peters-Golden, 2000), and the tyrosine kinase Syk (Canetti et al., 2003). It is noteworthy that a recent report demonstrated that FcR engagement results in tyrosine phosphorylation of BLT1R by the Src family of kinases, which leads to a molecular complex within plasma membrane lipid rafts comprising the FcR, BLT1R, Gαi3 protein, and Src (Serezani et al., 2009). Because disruption of lipid rafts abolished the LTB4-enhanced phagocytosis in the latter study, the formation of a complex between BLT1R and immunoreceptors may represent a major mechanism in macrophage phagocytosis.
Several of the LTB4-induced responses in monocytes/macrophages, such as MCP-1 secretion, intregrin-dependent monocyte arrest, and macrophage phagocytosis enhancement, are blocked by selective BLT1R antagonists (Friedrich et al., 2003; Huang et al., 2004). However, BLT2 receptor mRNA has also been detected in human monocytes (Yokomizo et al., 2001a). Furthermore, BLT1 and BLT2 receptor expression has been demonstrated immunohistochemically in macrophage-rich areas within human atherosclerotic lesions (Bäck et al., 2005) and abdominal aortic aneurysms (Houard et al., 2009), as well as by in situ hybridization of macrophages in synovial tissues from patients with rheumatoid arthritis (Hashimoto et al., 2003). Although an initial study demonstrated that macrophages derived from BLT1R knockout mice did not alter [Ca]i after LTB4 stimulation (Tager et al., 2003), a subsequent study reported that these cells display chemotaxis toward LTB4 albeit at 100-fold higher concentrations than that in WT macrophages (Subbarao et al., 2004). The latter observation suggests that the low-affinity BLT2 receptor may be involved in LTB4-induced monocyte/macrophage chemotaxis. Based on results obtained in transfected CHO cells, in which expression of one BLT receptor leads to chemotaxis toward a narrow range of LTB4 concentrations, coexpression of human BLT1 and BLT2 receptors makes cells migrate toward a wider range of LTB4 concentrations (Yokomizo et al., 2001a).
5. Dendritic Cells.
Dendritic cells (DCs) are antigen-presenting cells that control adaptive immunity through direct interaction with T lymphocytes. Initially, human monocyte-derived DCs were reported to express mRNA for BLT2R but not BLT1R and the LTB4-induced DC chemotaxis was reported to be inhibited by the BLT2R antagonist LY255283 (Shin et al., 2006). However, a subsequent study revealed BLT1R protein expression by flow cytometric analysis of human DCs (Toda et al., 2010). The latter finding was in line with the demonstration of BLT1R but not BLT2R mRNA in murine bone marrow-derived DCs (BMDCs) (Toda et al., 2010). The latter cells exhibit functional responses to LTB4, such as increased [Ca2+]i and chemotaxis (Del Prete et al., 2007; Toda et al., 2010), not observed in BMDCs derived from either BLT1R or BLT1R and BLT2R knockout mice (Del Prete et al., 2007; Toda et al., 2010). Other responses associated with LTB4 stimulation of murine DCs include induction of the chemokine receptor CCR7 (Del Prete et al., 2007), IL-12 secretion, and NF-κB translocation (Toda et al., 2010). Furthermore, murine DCs lacking both BLT1R and BLT2R exhibited decreased in vivo migration to draining lymph nodes after either paw injection (Del Prete et al., 2007) or intratracheal instillation (Miyahara et al., 2008). In vivo models have also implicated a pathophysiological role for DC BLT receptors by demonstrating decreased airway hyperreactivity after transfer of OVA-pulsed BLT1R-deficient BMDCs compared with wild-type BMDCs (Miyahara et al., 2008) and decreased skin contact hypersensitivity in BLT1R and BLT2R knockout mice (Del Prete et al., 2007). Finally, the OVA-dependent cytokine production in mouse splenocytes after adoptive transfer of BLT1R-deficient BMDCs exhibited decreased production of the Th1 cytokine IFN-γ and elevated levels of the Th2 cytokines IL-4 and IL-5 (Toda et al., 2010).
6. T Lymphocytes.
The chemotactic response of human blood-derived human T lymphocytes was initially reported to exhibit a peak concentration of LTB4 at 10−8 M (i.e., a lower potency compared with granulocytes) (Leppert et al., 1995). The latter observation was supported by the binding of radioactive LTB4 to the CD4+8+3low Tsup-1 cell line, which was characterized by a Kd of approximately 200 nM and a rank order of competitive potencies of 20-OH-LTB4, 6-trans-LTB4, which were both similar to that of the low-affinity subset of human neutrophil LTB4-binding sites (Leppert et al., 1995). Subsequent studies of peripheral blood cells have supported those initial observations by the identification of dominant levels of BLT2R mRNA compared with BLT1R mRNA (Yokomizo et al., 2001a). Moreover, in peripheral blood, LTB4 bound only to 11% of T lymphocytes, representing mainly cells of the CD8+ subset (Payan et al., 1984). The latter observation was subsequently supported by a flow cytometric demonstration of BLT1 and CD8 coexpression in a small fraction of peripheral blood cells, whereas CD4+ lymphocytes were BLT1R-negative (Pettersson et al., 2003). However, decreased BLT2R expression has been observed after T-lymphocyte stimulation (Yokomizo et al., 2001a), and animal studies indicate that BLT1R activation may be the major signaling pathway mediating LTB4-induced T lymphocyte recruitment in vivo (Goodarzi et al., 2003; Tager et al., 2003; Miyahara et al., 2005a). Furthermore, morphological analyses have associated adventitial vascular inflammation with an accumulation of BLT1R-expressing T lymphocytes (Houard et al., 2009).
BLT1R mRNA expression is induced in antigen-exposed murine CD8+ T-effector cells, associated with an increased chemotactic response; effects inhibited either by the BLT1R antagonist CP-105,696 or in cells derived from BLT1R knockout mice (Goodarzi et al., 2003; Ott et al., 2003). These findings suggest that the ability to use BLT1R for migration into inflamed tissues is a feature specific to fully differentiated CD8+ effector cells (Goodarzi et al., 2003). Likewise, BLT1R mRNA is induced in CD4+ murine lymphocytes during either Th1 or Th2 polarization (Tager et al., 2003). A subsequent study of human blood-derived T lymphocytes demonstrated that the small fraction of BLT1R-positive cells also were enriched for the activation markers CD38 and HLA-DR and expressed the effector cytokines IFN-γ and IL-4 (Islam et al., 2006). Furthermore, in vitro activation of human T lymphocytes by alloactivated DCs increased BLT1R surface expression and enhanced LTB4-induced chemotaxis (Islam et al., 2006). In further support of an up-regulation of BLT1R in activated T lymphocytes, the number of circulating CD8+ cells positive for BLT1R was increased during acute Epstein-Barr virus infection (Islam et al., 2006). Finally, bronchoalveolar lavage derived from patients with lung transplant graft dysfunction (Medoff et al., 2005) and asthma (Islam et al., 2006) contained an increased proportion of BLT1R-positive T lymphocytes.
LTB4 stimulation also enhanced the production of IL-1, IL-2, IL-5, and IFN-γ in human blood-derived T lymphocytes in vitro (Rola-Pleszczynski et al., 1986; Yamaoka and Kolb, 1993) and the production of MMP-2, -3, and -9 in CD4+8+3low Tsup-1 cells (Leppert et al., 1995). Furthermore, LTB4 inhibited the differentiation of murine naive CD4+ T lymphocytes into CD25+ Foxp3+ T regulatory cells and increased IL-17 expression (Chen et al., 2009a), providing the first suggestion that LTB4 may counteract the generation of immunosuppressive T regulatory cells and promote the differentiation into Th17 cells. In support of a role for LTB4 in the development of Th17 immune responses, lymph node cells derived from BLT1R knockout mice after EAE induction exhibit decreased IL-17 production (Kihara et al., 2010).
The role of BLTR signaling in T lymphocyte recruitment has also received support from in vivo studies. The pulmonary inflammation and airway hyperresponsiveness after OVA sensitization and challenge are reduced in BLT1R knockout mice and restored after reconstitution with BLT1R-expressing T lymphocytes (Tager et al., 2003; Miyahara et al., 2005a). In addition, mice lacking CD8+ T lymphocytes are protected from airway hyperresponsiveness induced by allergen sensitization and challenge (Miyahara et al., 2005b). Whereas reinstallation of CD8+ cells restored airway hyperresponsiveness in the latter mice, adoptive transfer of BLT1R-deficient CD8+ cells maintained the protective effect (Miyahara et al., 2005b). In other models, such as complete Freund's adjuvant-induced peritonitis, adoptive transfer of fluorescence-marked T lymphocytes derived from both BLT1R-deficient and WT mice resulted in an accumulation of 1.6- to 1.9-fold higher numbers of WT cells in peritoneal exudates (Goodarzi et al., 2003).
7. B Lymphocytes.
BLT1R protein and BLT1R and BLT2R mRNA have been detected in CD19+ human B lymphocytes (Yokomizo et al., 2001a; Pettersson et al., 2003; Runarsson et al., 2005). In cells derived from either blood or tonsillectomy tissues, LTB4 enhances the effects of other stimulating factors on CD23 expression, proliferation, and differentiation of resting, but not activated, B lymphocytes (Yamaoka et al., 1989, 1994; Dugas et al., 1990). The latter findings were subsequently supported by the demonstration that Epstein-Barr virus infection of B lymphocytes is associated with a loss of BLT1R expression and a lack of LTB4-induced effects on the proliferation of infected cells (Liu et al., 2008). BLT1R protein expression has also been demonstrated in B lymphocytes derived from patients with B-cell chronic lymphocytic leukemia (Runarsson et al., 2005). In this study, LT synthesis inhibitors decreased DNA synthesis and antigen expression in B-cell chronic lymphocytic leukemia lymphocytes, an effect that was reversed by exogenously added LTB4 (Runarsson et al., 2005).
8. Mast Cells.
Experiments in human umbilical cord blood-derived mast cells have demonstrated that LTB4 is a chemoattractant for the immature c-kit+ MC populations, whereas the mature c-kithigh MCs are unresponsive (Weller et al., 2005). The chemotactic effects induced by LTB4 on murine bone marrow-derived mast cells (BMMCs) cultured in the presence of IL-3 may also depend on the maturity of the MC, because cells cultured for 2 to 6 weeks migrate in response to LTB4 (Kitaura et al., 2005; Weller et al., 2005; Lundeen et al., 2006), whereas after 10 weeks of culture, the cells are unresponsive to this agonist (Weller et al., 2005). The latter finding correlated with a decrease in BLT1R mRNA. For example, 2-week-old BMMCs expressed 4-fold higher levels than 6-week-old cells and 10-fold more than 10-week-old cells (Weller et al., 2005). A subsequent study also described BLT2R mRNA in BMMC after 4 weeks of culture and indicated that either the BLT1R antagonist U75302 or the BLT2 receptor antagonist LY255283 inhibited LTB4-induced BMMC chemotaxis, although no additive effect was observed after combination of the two receptor antagonists (Lundeen et al., 2006). In addition, antigen stimulation increased BLT2R expression in murine BMMCs (Cho et al., 2010a). In support of a functional murine MC BLT2 receptor, the BLT2 receptor agonists 12-HHT and 12(S)-HETE have also been reported to induce BMMC migration (Lundeen et al., 2006; Okuno et al., 2008). The [Ca2+]i induced by 12-HHT was abolished in BMMCs derived from BLT2R knockout mice, whereas BMMCs from both WT and BLT1R knockout mice responded to LTB4 (Mathis et al., 2010). However, LTB4-induced migration was abolished in BMMCs derived from BLT1R-deficient mice, whereas 12-HHT-induced migration was abolished in BMMCs derived from BLT2R-deficient mice (Okuno et al., 2008). Finally, a human MC line, HMC-1, has been reported to express both BLT1 and BLT2 receptor mRNA and migrate in response to LTB4 (Lundeen et al., 2006). An unexpected finding was that flow cytometry and immunostaining suggested an intracellular localization of BLT1R protein in these cells (Lundeen et al., 2006), but the functional significance of this receptor localization remains to be established.
9. Vascular Smooth Muscle Cells.
The LTB4-induced vasoconstriction, demonstrated in human pulmonary artery (Sakata et al., 2004), in guinea pig pulmonary artery and aorta (Bäck et al., 2004; Sakata et al., 2004), and in rat basilar artery (Trandafir et al., 2005), has led to the hypothesis that nonmyeloid cells, such as SMCs, also express BLT receptors. Immunohistochemical analysis of human arteries has revealed BLT1R expression in the muscular layers of carotid atherosclerotic endarterectomies, as well as in human nonatheroclerotic mammary arteries (Bäck et al., 2005). Moreover, in human coronary artery SMCs, 8-fold higher BLT1R mRNA levels compared with BLT2R mRNA has been demonstrated by RT-PCR (Heller et al., 2005), supporting the observation that the BLT1R subtype may be the dominant BLT receptor subtype in vascular SMCs. The expression of BLT1R proteins on human coronary artery SMCs has also been demonstrated by Western blot (Bäck et al., 2005) and flow cytometry (Heller et al., 2005). In contrast to this apparent constitutive BLT1R expression in human coronary artery SMCs, murine aortic SMCs express BLT1R mRNA only after 24 h of treatment with TNF-α and IFN-γ (Heller et al., 2005).
In human coronary artery SMCs, LTB4 induced an approximately 4-fold increase in whole-cell currents measured by patch clamp, and the activation of SMC by LTB4 induces migration and proliferation (Bäck et al., 2005). Furthermore, LTB4 stimulated the release of MMP-2 (Hlawaty et al., 2009; Seo et al., 2010) and activation of integrin signaling pathways in vascular SMCs (Moraes et al., 2010). The pathophysiological importance of LTB4-induced activation of SMCs in the context of atherosclerosis has been reinforced by the inhibitory effects of BLT receptor antagonists on the development of intimal hyperplasia after vascular injury (Kondo et al., 1998; Bäck et al., 2005; Hlawaty et al., 2009). Furthermore, atherosclerotic lesions of apoE(−/−) mice display reduced SMC content after BLT1R gene disruption (Heller et al., 2005).
10. Bronchial Smooth Muscle Cells.
In line with findings in human vessels, immunohistochemical staining of human bronchi revealed colocalization of both BLT1R and BLT2R protein with smooth muscle α-actin-positive airway SMCs (Watanabe et al., 2009). Furthermore, both receptors were also identified on human airway SMCs in culture by RT-PCR, Western blot analysis, and flow cytometry. In the latter cells, LTB4 induced proliferation and migration, which was inhibited by the selective BLT1R antagonist U75302 (Watanabe et al., 2009).
11. Endothelial Cells.
In human vessels, endothelial expression of BLT1R has been demonstrated by immunohistochemical analysis and was only observed in atherosclerotic and not in healthy arteries, suggesting an induction during vascular inflammation (Bäck et al., 2005). The latter notion of BLT1R immunoregulation has received support from studies in HUVECs demonstrating BLT1R up-regulation by LPS and IL-1β (Table 2), whereas BLT2R was induced by TNF-α (Qiu et al., 2006).
LTB4-induced endothelial activation has also been implicated in increased leukocyte adherence (Hoover et al., 1984). However, the in vitro neutrophil adherence observed after LTB4 stimulation of HUVECs is weak (Lindström et al., 1990; Palmblad et al., 1994), and the role of endothelial BLT receptors in leukocyte adhesion is today not fully characterized. Papayianni et al. (1996) demonstrated that LTB4 had no effect on P-selectin expression in HUVECs. Another study reported no effect on monocyte arrest under physiological flow conditions after EC stimulation with LTB4 (Friedrich et al., 2003). These observations provided evidence for a direct effect of LTB4 on leukocytes rather than on ECs. However, other investigators demonstrated that the endothelium released vasoactive factors via BLT receptor activation (Bäck et al., 2004; Qiu et al., 2006). These reports provide evidence for a role of LTB4 in regulating endothelial function. In addition, vascular endothelial growth factor (VEGF) up-regulated the expression of BLT1R and BLT2R mRNA and protein in HUVECs (Kim et al., 2009). It is noteworthy that VEGF also increased the release of 12(S)-HETE, and the VEGF-induced blood vessel formation in vivo and in vitro was inhibited by either the BLT2R antagonist LY255283 or BLT2R knock down by small interfering RNA (Kim et al., 2009). These findings suggest a role for 12(S)-HETE signaling through endothelial BLT2R in angiogenesis. Taken together, the data indicate that the ligands for the BLT receptors affect ECs but that the interaction between the ligands and the endothelium in adhesion experiments is presently difficult to observe, possibly being masked by the LTB4-induced effects on leukocytes.
E. BLT Receptor Functional Analysis through Altered Gene Expression
1. BLT Receptor Transgenic Models.
Transgenic mice expressing the human BLT1R under the CD11b promoter to obtain leukocyte specific expression exhibit an increased granulocyte infiltration in response to topical LTB4 application and ischemia reperfusion (Chiang et al., 1999; Brink et al., 2003). In addition, transgenic BLT2R overexpressed in mice have been studied in a model of angiogenesis and shown to exhibit increased blood vessel formation in response to LTB4 and 12(S)-HETE (Kim et al., 2009).
2. BLT1 Receptor Knockout.
Mice lacking BLT1R were generated in 2000 (Haribabu et al., 2000; Tager et al., 2000; Brink et al., 2003). Although these mice developed normally and had no apparent hematopoietic abnormalities, decreased leukocyte chemotaxis (Haribabu et al., 2000; Tager et al., 2003) and integrin-mediated leukocyte arrest in postcapillary venules (Tager et al., 2003) in response to LTB4 compared with WT mice were reported. In addition, BLT1R deficiency conferred protection against inflammatory responses in different disease models.
After induction of acute peritonitis, BLT1R-deficient mice exhibit decreased leukocyte content in peritoneal exudates compared with that in WT mice (Haribabu et al., 2000; Tager et al., 2000; Goodarzi et al., 2003) Other in vivo studies of endotoxin provocation have demonstrated decreased portal venule leukocyte adherence in response to systemic LPS in BLT1R knockout mice compared with that in WT mice (Ito et al., 2008), whereas the leukocyte infiltration after local intraocular LPS injection was not altered by BLT1R disruption (Smith et al., 2004). In addition to models of acute inflammation, BLT1R knockout mice have also revealed decreased inflammation in more chronic inflammatory responses. For example, in contrast to the lack of effects on endotoxin-induced uveitis, BLT1R-deficient mice were protected from experimental autoimmune uveitis induced by immunization of interphotoreceptor retinoid-binding protein (Liao et al., 2006).
Mice lacking apolipoprotein E exhibit hyperlipidemia and develop spontaneous atherosclerosis. Apolipoprotein E and BLT1R double knockout mice display smaller lesions compared with their apoE(−/−) littermates (Subbarao et al., 2004; Heller et al., 2005; Bäck, 2008b). The latter double knockout mice are also protected from abdominal aortic aneurysm development induced by angiotensin II infusion (Ahluwalia et al., 2007).
After OVA sensitization, BLT1R knockout mice exhibit decreased airway responsiveness associated with a decreased leukocyte accumulation (Tager et al., 2003; Miyahara et al., 2005a; Terawaki et al., 2005), which is restored after adoptive transfer of BLT1R-expressing T lymphocytes (Miyahara et al., 2005a; Tager et al., 2003) (see section II.D). In the EAE model, BLT1R-deficient mice develop less severe clinical signs of disease and a significantly delayed onset of disease than WT mice, associated with decreased inflammatory infiltration in the spinal cord (Kihara et al., 2010). Finally, BLT1R knockout mice are protected in models of both collagen- and K/BxN serum-induced arthritis (Kim et al., 2006a; Shao et al., 2006; Mathis et al., 2010), discussed in detail below. In this context, BLT1R knockout mice displayed a decrease in bone resorption induced by either LPS or ovariectomy (Hikiji et al., 2009).
3. BLT2 Receptor Knockout.
In recent studies, BLT2R knockout mice were generated (Iizuka et al., 2010; Mathis et al., 2010). Because of to the close proximity of the BLT1R gene promoter, the BLT2R gene disruption was generated through insertion into the open reading frame of the BLT2 gene to avoid alterations of BLT1R expression (Iizuka et al., 2010; Mathis et al., 2010). The BLT2R-deficient mice were viable, developed normally, and displayed no overt behavioral or morphological defects (Mathis et al., 2010). An unexpected finding was that opposing phenotypic changes were observed in two different mouse models of disease. BLT2R knockout mice displayed attenuated inflammatory arthritis in response to K/BxN serum transfer (Mathis et al., 2010). The latter effect was associated with decreased inflammatory cell influx and could be reproduced by bone marrow transplantation, suggesting that leukocyte BLT2R is necessary for full arthritis development (Mathis et al., 2010). However, the protection against K/BxN serum-induced arthritis conferred by the lack of BLT1R is more pronounced compared with that observed in BLT2R-deficient mice (Kim et al., 2006a; Mathis et al., 2010). Nevertheless, the transfer of WT neutrophils into BLT1R knockout mice promoted the entry of endogenous BLT1R-deficient neutrophils into the joints of these mice (Kim et al., 2006a) through the production of IL-1 (Chou et al., 2010). Taken together, these data suggested that although the BLT1R may be necessary for the initiation of autoantibody-induced arthritis, BLT2R may play a possible role at later stages of disease, when local LTB4 concentrations are higher in the joint (Mathis et al., 2010). In contrast to the findings of decreased arthritis in BLT2R knockout mice, another study demonstrated that this gene disruption induced more severe colitis in response to dextran sodium sulfate (DSS) compared with that in either WT or BLT1R knockout mice, which was accompanied by increased expression of inflammatory cytokines, chemokines, and MMPs (Iizuka et al., 2010). However, in the latter study, BLT2R deficiency was associated with a dysfunctional barrier function in colonic epithelial cells rather than direct effects on leukocytes (Iizuka et al., 2010). Although the effects of DSS-induced colitis have not been fully explored in BLT1R knockout mice, those mice exhibit weight loss similar to that in WT mice, suggesting no protective effects. The notion that BLT2R agonists other than LTB4 may be involved in the exacerbated colitis observed after BLT2R knockout (Iizuka et al., 2010) is supported by the fact that HHT is produced at 5- to 6-fold higher levels compared with LTB4 in inflamed colonic biopsy specimens (Zijlstra et al., 1992, 1993) (see section II.E).
4. BLT1 and BLT2 Receptor Double Knockout.
Double knockout of BLT1R and BLT2R has been generated through direct targeting of the genes encoding both receptors (Shao et al., 2006). In line with findings after disruption of any of the BLT receptor subtypes alone (see above), these mice were also viable, developed normally, and displayed no overt behavioral or morphological defects and no alterations of leukocyte subpopulation counts (Shao et al., 2006). After collagen-induced arthritis, both BLT1R knockout and BLT1R-BLT2R knockout mice respond in a similar fashion with less clinical and histological signs of arthritis, as well as smaller synovial inflammatory cell infiltration compared with WT mice, without altering the antibody response (Shao et al., 2006).
F. Potential Therapeutic Applications
The cell-type specific BLTR signaling discussed above (see section II.D) may imply differential effects of LTB4 in the development of different diseases. For example, LTB4 may mainly exert its effects on neutrophil granulocytes in models of arthritis (Kim et al., 2006a) abdominal aortic aneurysms (Houard et al., 2009), and cerebral ischemia/reperfusion (Barone et al., 1992), whereas models of MS have implicated a major role of LTB4-induced effects on eosinophil granulocytes (Gladue et al., 1996). In contrast, T lymphocytes may be the main effector cells in LTB4-induced airway hyperresponsiveness (Miyahara et al., 2005a). Furthermore, in other disease models, LTB4 signaling may affect several cell types, such as in atherosclerosis, in which macrophages, vascular SMCs, and ECs express BLT1 receptors (Bäck et al., 2005). The roles of LTs and related lipid mediators in various diseases have been extensively reviewed (Shimizu, 2009).
The uptake and modification of lipids in the vascular wall induce a local inflammatory reaction, eventually developing into an atherosclerotic lesion. Generation of LTB4 has been described in ex vivo stimulated atherosclerotic vessels, and intraluminal LT formation has been demonstrated during coronary balloon angioplasty (Brezinski et al., 1992; Bäck, 2008b). In addition, several of the BLT receptor-expressing immune cells discussed above are present in the atherosclerotic lesion (Bäck, 2008b). Although studies of targeted 5-LO have generated contradictory results, BLT1R knock out (Subbarao et al., 2004; Heller et al., 2005) and the BLT1R antagonist CP-105,696 (Aiello et al., 2002) reduces atherosclerosis in hyperlipidemic mice (Bäck, 2008b). In addition to effects on the inflammatory response, LTB4 signaling through BLT1R expressed in vascular SMCs may also be involved in the development of atherosclerosis and intimal hyperplasia after vascular interventions (Bäck et al., 2005; Heller et al., 2005; Hlawaty et al., 2009). Clinical trials have been initiated to evaluate the effects of anti-LTs on biomarkers and atherosclerosis morphology in patients with coronary heart disease (Bäck, 2009a).
2. Aortic Abdominal Aneurysms.
Circulating neutrophils derived from patients undergoing surgery for abdominal aortic aneurysms produce increased levels of LTB4 (Gadaleta et al., 1994), and the expression levels of LTB4 synthesizing enzymes has in addition been demonstrated in neutrophils infiltrating the intraluminal thrombus covering the aneurysm (Houard et al., 2009). This local LTB4 production has also been demonstrated to transduce the major part of the chemotactic activity derived from the intraluminal thrombus (Houard et al., 2009). In addition to neutrophils, BLTR-positive macrophages and T lymphocytes may be involved in the adventitial inflammation, which is part of the pathogenesis of abdominal aortic aneurysms (Houard et al., 2009). The importance of BLT1R signaling has also received support from animal studies. Either genetic or pharmacological targeting of BLT1R signaling reduces the incidence of experimental abdominal aortic aneurysms induced by angiotensin infusion in apoE knockout mice (Ahluwalia et al., 2007; Kristo et al., 2010).
3. Cerebrovascular Disease.
Leukotrienes are produced during cerebral ischemia, and LT synthesis inhibitors limit the damage after experimental cerebral ischemia and reperfusion (Bäck, 2009b). Although most studies on the receptors involved in this protective effect have focused on CysLTRs and more recently GPR17 (see section III.H), there are also studies supporting a role for BLTR signaling in cerebrovascular disease. In a study in rats, temporary middle cerebral artery occlusion followed by reperfusion time dependently increased LTB4 binding in the cerebral cortex, which was paralleled by increases in myeloperoxidase activity (Barone et al., 1992). The results of the latter study suggested increased LTB4 binding to receptors located on accumulating neutrophils in ischemic brain tissue (Barone et al., 1992).
In human cerebrovascular disease, macrophage 5-LO expression and LTB4 production have been reported to correlate with either clinical or radiological signs of cerebral ischemia in patients undergoing vascular surgery for carotid artery stenosis (Cipollone et al., 2005). Furthermore, BLT1R and BLT2R are expressed within carotid artery atherosclerotic lesions (Bäck et al., 2005). Finally, and as stated above (see section II.A), haplotypes within the LTB4R and LTB4R2 gene complex conferred a 2.3-fold increased risk for ischemic stroke in two case-control studies (Bevan et al., 2008).
4. Multiple Sclerosis.
MS is an inflammatory disorder of the central nervous system associated with blood-brain barrier breakdown, inflammatory cell accumulation, and myelin degradation (Mirshafiey and Jadidi-Niaragh, 2010). As discussed above (see section II.D), the beneficial effects on EAE after either genetic or pharmacological BLT1R targeting (Fretland et al., 1991; Gladue et al., 1996; Kihara et al., 2010) suggest that that BLT1R signaling may potentially affect both the onset and severity of MS. In support of this, increased levels of LTB4 have been detected in cerebrospinal fluid from patients with MS (Neu et al., 1992, 2002). Of interest, 5-LO was identified as one of the most up-regulated genes by microarray analysis in both human MS lesions and in brains from mice after EAE induction (Whitney et al., 2001). In the brain parenchyma around MS lesions, 5-LO expression colocalized with macrophages (Whitney et al., 2001). However, in contrast to findings in BLT1R knockout mice (Kihara et al., 2010), 5-LO-deficient mice exhibit an exacerbated EAE compared with WT mice (Emerson and LeVine, 2004). These opposing effects of targeting either LT synthesis or LT receptors are in line with findings in atherosclerosis (see above) and point to the fact that targeting 5-LO will inhibit both proinflammatory mediators (i.e., LTs) and anti-inflammatory mediators (i.e., lipoxins), as has been demonstrated in 5-LO-deficient mice (Bafica et al., 2005). This, in turn, suggests that endogenous anti-inflammatory mediators derived from 5-LO metabolism may be involved in regulating these disease models. In addition, studies in mice lacking 12/15-LO have also shown exacerbated EAE (Emerson and LeVine, 2004) or a defect in both corneal reepithelization and neutrophil recruitment that correlates with a reduction in endogenous lipoxin formation (Gronert et al., 2005), raising the notion that lipoxins, for which the biosynthesis involves both 5-LO and 12/15-LO (Serhan et al., 2008), may be beneficial in MS. These findings suggest that BLT receptor antagonism may represent an advantage compared with 5-LO inhibition in MS.
BLTR expression has been demonstrated in synovial tissues derived from patients with rheumatoid arthritis (Hashimoto et al., 2003). The beneficial effects in models of rheumatoid arthritis after genetic disruption of BLT1R and/or BLT2R in C57BL6 mice (Kim et al., 2006a; Shao et al., 2006; Mathis et al., 2010) were discussed above (see section II.E. Pharmacological treatment with the BLT1R antagonist CP-105,696 reduced the histological signs of collagen-induced arthritis in the more arthritis-prone DBA/1J mice (Griffiths et al., 1995). In humans, oral administration of the BLT receptor antagonist amebulant (Birke et al., 2001) inhibits Mac-1 expression on neutrophils (Alten et al., 2004), but clinical trials in patients with arthritis have not shown any statistically significant beneficial effects (Díaz-González et al., 2007).
6. Pulmonary Inflammation.
Although most studies of LTs in asthma have focused on CysLT receptor signaling (see section III.A), several studies point to a participation of LTB4 in chronic airway inflammation. For example, whereas CysLT1R antagonists generally have small effects on the increased bronchial responsiveness in asthma, clinical trials of either 5-LO inhibitors or FLAP antagonists have been more effective, supporting an involvement of LTB4 (Dahlén, 2006). In addition, neutrophil recruitment to the airway is thought to be a major component of continuing inflammation and progression of chronic obstructive pulmonary disease (COPD), and the 46% of the neutrophil chemotactic activity in sputum derived from patients with COPD is inhibited by a BLT receptor antagonist (Woolhouse et al., 2002). Increased concentrations of LTB4 have also been reported in BAL (Wenzel et al., 1995) and exhaled breath condensates from patients with asthma (Csoma et al., 2002). In further support of a role of BLT1R signaling in asthma, the BLT1R antagonist CP-105,696 decreased airway hyperresponsiveness in a primate model of asthma (Turner et al., 1996). As discussed above (section II.E), BLT1R knockout mice exhibited decreased airway responsiveness after OVA sensitization and challenge, associated with decreased pulmonary inflammation and mucus secretion compared with that in WT mice (Miyahara et al., 2005a; Terawaki et al., 2005). In addition, BLT2R signaling may be involved in the response to OVA sensitization and challenge, as suggested by the reduced airway inflammation and reduced airway hyperresponsiveness after BLT2R targeting through in vivo administration of either the BLT2R antagonist LY255283 or an antisense BLT2R (Cho et al., 2010b).
In a crossover study of 12 atopic asthmatic individuals, the BLT receptor antagonist LY293111, 2-[2-propyl-3-[3-[2-ethyl-4-(4-fluorophenyl)-5-hydroxyphenoxy]-propoxy]phenoxyl]benzoic acid (LY293111) did not alter either lung function or airway reactivity after allergen challenge but significantly reduced the number of neutrophils in BAL derived from treated patients compared with placebo (Evans et al., 1996).
7. Inflammatory Bowel Disease.
Colonic mucosa derived from patients with inflammatory bowel disease (IBD) exhibit increased mRNA levels of LT-forming enzymes and enhanced LTB4 production (Sharon and Stenson, 1984; Jupp et al., 2007). However, studies performed with ex vivo stimulated human colonic mucosa have revealed that lower amounts of LTB4 are released compared with other arachidonic acid metabolites (Zijlstra et al., 1992, 1993). For example, 15-HETE is the dominant product formed after addition of exogenous arachidonic acid to human biopsy specimens obtained at coloscopy (Zijlstra et al., 1992, 1993). Furthermore, HHT, which may act as a high-affinity BLT2R ligand (see section II.A and Fig. 1), is produced at 5- to 6-fold higher concentration compared with LTB4 (Zijlstra et al., 1992, 1993). Nevertheless, the exact role of these ligands and the two BLTR subtypes in IBD remains to be established. Whereas different BLTR antagonists inhibit colonic inflammation and neutrophil infiltration in animal models of IBD (Fretland et al., 1990, 1991), BLT2R-deficient mice exhibit exacerbated colonic inflammation (Iizuka et al., 2010).
LTB4 levels are increased in different human cancer tissues, such as prostate (Larré et al., 2008) and oral (el-Hakim et al., 1990) cancers. Furthermore, immunohistochemical analysis has revealed expression of BLT receptors in human pancreatic and colon cancers (Hennig et al., 2002; Ihara et al., 2007). As outlined above, LTB4 signals through pathways associated with cell proliferation, such as MAPK, ERK, and PI3K, which has also been implicated in tumor growth (Tong et al., 2005; Ihara et al., 2007). In support of the latter suggestion, in vitro studies have shown that LTB4 promotes and that BLT1R antagonists inhibit the proliferation of cultured human cancer cell lines (Earashi et al., 1995; Bortuzzo et al., 1996; Ihara et al., 2007). In addition to direct effects on tumor cells, the BLT2 receptor-associated angiogenesis discussed above (see section II.A) (Kim et al., 2009) may potentially also be involved in LTB4-induced cancer growth (Wang and Dubois, 2010a).
Administration of the BLT1R antagonist LY293111 after transplantation of human cancer cells to athymic mice resulted in reduced tumor growth and reduced incidence of metastases (Hennig et al., 2004, 2005). However, results from clinical trials using the latter antagonist in combination with gemcitabine did not improve the prognosis for patients with either pancreatic or lung cancer (Wang and Dubois, 2010a).
G. Other Receptors Involved in BLT Receptor Signaling
1. Peroxisome Proliferator-Activated Receptors.
Peroxisome proliferator-activated receptors (PPARs) are ligand-activated transcription factors that belong to the nuclear hormone receptor superfamily (Michalik et al., 2006). PPARs activate gene transcription associated with metabolism and inflammation, in response to synthetic PPAR ligands developed for the treatment of diabetes and dyslipidemia (Michalik et al., 2006). In addition, LTB4 has been identified among the endogenous ligands for PPARα. In transfected HeLa cells, LTB4 activated PPARα reporter gene transcription in the micromolar range and bound a fusion protein containing the ligand-binding domain of PPARα, with a Kd of 90 nM (Devchand et al., 1996). Subsequent studies using different methods have confirmed LTB4 binding to and activation of PPARα (Krey et al., 1997; Lin et al., 1999; Narala et al., 2010), although negative results have also been reported (Forman et al., 1997). In addition to results obtained in vitro, a recent study has also demonstrated that mice lacking LTB4 biosynthesis through 5-LO knockout exhibited reduced PPARα activation in response to LPS administration in vivo (Narala et al., 2010). The latter results therefore support the observation that endogenous intracellular LTB4 formation can reach sufficient concentrations for PPARα activation under inflammatory conditions (Narala et al., 2010).
2. Vanilloid Transient Receptor Potential V1 Receptor.
The TRPV1 receptor is a ligand-gated, nonselective cation channel, which belongs to a family of TRP channels present exclusively in small sensory neurons and associated with pain. Although capsaicin and other vanilloids are the classic agonists, endogenous ligands derived from lipoxygenase metabolism of arachidonic acid and anandamide have also been reported to activate TRPV1 receptors (Hwang et al., 2000; McHugh et al., 2006). In inside-out patches of cultured dorsal root ganglion, TRPV1 receptors are activated by lipoxygenase products such as 12(S)-HpETE, 15(S)-HpETE, 5(S)-HETE, and LTB4, whereas LTC4 is without effect (Hwang et al., 2000). The EC50 for LTB4-induced TRPV1 activation was 11 μM, which was 10-fold higher than that obtained for capsaicin in the latter study (Hwang et al., 2000) and substantially lower compared with BLT1R-mediated LTB4-induced signaling. In line with these findings, LTB4 also increased [Ca2+]i in TRPV1-transfected CHO cells, albeit with significantly lower Emax and a longer exposure time required to achieve maximum effects compared with capsaicin (McHugh et al., 2006). Indeed, very recently it has also been demonstrated in vivo that both peripheral and spinal administration of RvE1 or RvD1 in mice potently reduces inflammatory pain behaviors, without affecting basal pain perception. These actions are transduced through RvE1-mediated inhibition of TRPV- and TNF-α-induced excitatory postsynaptic current increases in spinal dorsal horn neurons (Xu et al., 2010).
III. CysLT Receptors
A. CysLT Receptor Subtypes
1. From Cloning CysLT Receptors to Recent Molecular Advances.
The CysLTRs eluded gene cloning with conventional approaches for many years. Eventually, in 1999 two separate groups cloned the first CysLTR (Lynch et al., 1999; Sarau et al., 1999), and in the following year the second receptor subtype was cloned by three different groups (Heise et al., 2000; Nothacker et al., 2000; Takasaki et al., 2000). Hydrophobicity analysis of the deduced primary amino acid sequences demonstrated that both receptors possess seven TM helices organized in a serpentine topology, confirming early studies that provided evidence that cysteinyl LTs were acting through a GPCRs and confirmed pharmacological evidence for CysLTR heterogeneity based on the functional as well as on ligand-binding data (for review of early reports, see Capra, 2004; Capra et al., 2007; Evans, 2003). To date well documented results have been obtained, establishing the existence of two receptor subtypes referred to as CysLT1 and CysLT2 (see section III.G). This nomenclature was based on the observations that the CysLT1 receptor was sensitive to inhibition by classic antagonists, whereas the effects mediated by the CysLT2 receptor were not inhibited by these antagonists (Labat et al., 1992; Brink et al., 2003). These receptors belong to the rhodopsin family of the GPCR gene superfamily and, in particular, to the purine receptor cluster (within the δ group) of phylogenetically related receptors. In addition to a number of orphan receptors, this group includes receptors that respond to purinergic or pyrimidinergic nucleotides, proteases, and PAF (Fredriksson et al., 2003; Kroeze et al., 2003). Unlike the monoamine or neuropeptide receptors, the receptors belonging to the purine cluster have no clear homologs in invertebrates, suggesting a relatively recent evolutionary origin (Adams et al., 2000). Human CysLT1 and CysLT2 receptors share only 38% aa identity, with very low homology in the extreme carboxyl termini. Of interest, human CysLTRs have higher homology with the purinoceptor P2Y1/2/6 (32–30% aa identity) or GPR17 (34–32% aa identity) than with the other known LT receptors, the BLT1/2 receptors (≤27% aa identity).
The human CysLT1 gene/chromosome location has been mapped on the long arms of chromosome X (Xq13-Xq21), a region lacking asthmatic disease markers (Lynch et al., 1999), whereas CysLT2 on human chromosome 13 (13q14) (Heise et al., 2000; Takasaki et al., 2000) is near a marker that has been associated with atopic asthma (Cookson, 1999; Kimura et al., 1999).
a. CysLT1 Receptor.
The open reading frame of the human CysLT1 encodes a protein of 337 aa (NCBI Reference Sequence: NP_006630) with a calculated molecular mass of 38 kDa, observed to migrate at a molecular mass of approximately 42 kDa as a monomeric form (Bautz et al., 2001; Figueroa et al., 2001; Ohd et al., 2003; Hasegawa et al., 2010), although oligomers were often observed (see section III.G). However, other authors observed monomeric migration at different molecular masses, possibly reflecting differences in the cell type, protein/cell maturation, or experimental conditions: 30 to 36 kDa (Sjöström et al., 2001, 2002) and 38 to 40 kDa (Capra et al., 2004). Human CysLT1R possesses four potential N-glycosylation sites, one in the extracellular N-tail, two in the second extracellular loop, and one in the third extracellular loop, besides many potential protein kinase A and C phosphorylation sites (see section III.D), mostly located in the third intracellular loop and carboxyl terminal (Lynch et al., 1999). The rank order of potency for the cysteinyl LTs is LTD4 > LTC4 when tested in different functional assays, such as stimulation of calcium-activated chloride conductance in Xenopus laevis oocytes or fluorescence calcium imaging in COS-7, HEK-293, or CHO human CysLT1-transfected cells. LTE4 is the less potent agonist acting as a partial agonist. As expected, the LTD4 functional response is potently inhibited by the selective CysLT1R antagonists, 3-[[3-[2-(7-chloroquinolin-2-yl)vinyl]phenyl]-(2-dimethylcarbamoylethylsulfanyl)methylsulfanyl] propionic acid (MK-571), zafirlukast, pranlukast, montelukast, and pobilukast (Brink et al., 2003; Evans, 2003; Capra, 2004). Saturation radioligand binding experiments on membranes from COS-7 cells expressing the human CysLT1R and using [3H]LTD4 as a labeled ligand demonstrated a single Kd of 0.3 nM (Lynch et al., 1999) or high- and low-affinity binding sites with Kd values of 0.06 and 6.2 nM (Capra et al., 2005). Bmax has been reported to be approximately 50 fmol/mg membrane protein in both reports.
The genomic organization of the human CYSLTR1 gene has been obtained by 5′- and 3′-rapid amplification of cDNA ends method, modified to ensure the amplification of only full-length transcripts (Woszczek et al., 2005). The gene consists of five exons that are variably spliced and a single promoter region TATA-less with multiple transcription start sites. Four different CysLT1R transcripts were identified with a dominant and wide expression of the transcript I, containing exons 1, 4, and 5, with the strongest presence in blood leukocytes, spleen, thymus, lung, and heart (Woszczek et al., 2005). Successive articles reported that CYSLTR1 contains three exons with the entire open reading frame located in exon 3 and that five splice variants were detected in human leukocytes (Zhang et al., 2006), whereas in human airway SMCs and peripheral blood mononuclear cells (MNCs) three transcripts and, possibly, two putative promoters were found (Duroudier et al., 2007). In summary, multiple splice variants of CysLT1R exist, and the transcript expression patterns seem to differ from tissues and cell types. At present, the role of alternative splice variants of CysLT1R is not known. Woszczek et al. (2005) also reported in their study that the promoter region contains several binding sites for transcription factors, such as activator protein-1 (AP-1) and GATA and that CysLT1R expression is functionally regulated at the transcriptional level by IL-4 through a signal transducer and activator of transcription (STAT)-6 response element localized to the proximal promoter region, confirming former studies by Thivierge et al. (2001) (see section III.B).
b. CysLT2 Receptor.
The open reading frame of human CysLT2R encodes a protein of 346 aa (NCBI Reference Sequence: NP_065110), which seems to migrate at a molecular mass of 58 kDa in basophil lysates (Gauvreau et al., 2005) and approximately 50 kDa in platelet (Hasegawa et al., 2010) or at approximately 40 to 42 kDa in HUVECs and COS-7 cells transfected with human CysLT2 cDNA (Carnini et al., 2011). Human CysLT2R possesses four potential N-glycosylation sites, three of which in the extracellular N-tail, as well as many potential protein kinase A and C phosphorylation sites mostly located in the third intracellular loop and carboxyl terminal (Heise et al., 2000). Ca2+ mobilization assay in HUVECs confirmed early observation in recombinant systems that LTD4 and LTC4 are equipotent agonists with an identical EC50 of ∼35 nM, and lack of sensitivity of the response to the classic CysLT1R antagonists, montelukast, zafirlukast, and pranlukast, whereas LTE4 and 6(R)-(4-carboxyphenylthio)-5(S)-hydroxy-7(E),9(E), 11(Z),14(Z)-eicosatetraenoic acid (BAY U9773 behave as partial agonists (Lötzer et al., 2003). Equilibrium binding studies in intact HUVECs using [3H]LTC4 as labeled ligand revealed the presence of a high-affinity binding site with Kd of 29 pM and Bmax of 32 fmol/108 cells (Carnini et al., 2011). In CysLT2-transfected COS-7 cell membranes high- and low-affinity binding sites with Kd values of 0.4 and 50 nM or a single site of 4.8 nM were obtained in two independent saturation experiments using [3H]LTD4 as labeled ligand (Heise et al., 2000).
The genomic organization of the human CYSLTR2 has also been published recently. Similarly to CYSLTR1, the gene has a TATA-less promoter with multiple transcription start sites. Six variably spliced exons have been identified and eight different CysLT2R transcripts were also identified in endothelial and monocytic cells (Woszczek et al., 2007). It is noteworthy that IFN-γ increased CysLT2R mRNA expression and calcium signaling in ECs. However, there were no significant changes in gene reporter and mRNA t1/2 assays in response to this cytokine, suggesting transcriptional control of CysLT2R mRNA up-regulation by IFN-γ response motifs localized outside of the cloned CysLT2R promoter region. Stimulation of ECs by cysteinyl LTs induced mRNA and protein expression of early growth response genes 1, 2, and 3 and cycloxygenase-2 (COX-2) (Woszczek et al., 2007).
2. Receptor Expression Patterns with Functional Significance.
The pathophysiological role of cysteinyl LTs in asthma is well documented (Drazen, 2003; Holgate et al., 2003; Sampson et al., 2003; Arm, 2004; Capra et al., 2007; Hallstrand and Henderson, 2010), and results obtained from localization studies are consistent with the antibronchoconstrictive and anti-inflammatory activities of CysLT1R antagonists (Kemp, 2003; Riccioni et al., 2004; Currie et al., 2005; Capra et al., 2006; Dahlén, 2006; Ducharme et al., 2006; Montuschi et al., 2007; del Giudice et al., 2009). However, the finding of CysLTR expression in other tissues will certainly encourage the discovery of new functions for cysteinyl LTs in other physiological and pathological conditions (Capra et al., 2007).
a. CysLT1 Receptor.
Initial immunohistochemical analysis confirmed the presence of CysLT1R protein in a series of cells of particular relevance to asthma and atopy such as monocytes and eosinophils but also in pregranulocytic CD34+ cells and in subsets of B lymphocytes (Figueroa et al., 2001). Monocytes/macrophages isolated from peripheral blood MNCs express CysLT1R mRNA either in large excess compared with CysLT2R transcript levels (Lötzer et al., 2003) or, exclusively, such as in U937 (Capra et al., 2005), a human leukemic monocyte/macrophage cell line. This correlates with the finding that cysteinyl LTs contribute to inflammatory reactions by induction of MCP-1 via the CCR2B receptor in THP-1 cells (Ichiyama et al., 2005; Woszczek et al., 2005; Hashimoto et al., 2009). This reported induction by MCP-1 was also observed in another monocytic leukemia cell line and was further supported by the enhanced activation of chemotactic activity observed in human monocytes (Woszczek et al., 2008a). Likewise, in THP-1 cells, montelukast and zafirlukast significantly down-regulated the chemotaxis induced by MCP-1 and p38 MAPK expression (Hung et al., 2006). These data suggest a dominant functional pathway for the interaction between cysteinyl LTs and MCP-1 in a variety of MNCs, including human monocytes. In animal models, inhibition of MCP-1 produces antiatherogenic effects in vivo in a rabbit carotid balloon injury model (Ge et al., 2009).
The expression of CysLTRs in eosinophils was not unexpected (Bandeira-Melo and Weller, 2003), because the contribution of cysteinyl LTs to their accumulation within asthmatic airways has been well documented (Gauvreau et al., 2001; Ohshima et al., 2002; Nagata and Saito, 2003; Saito et al., 2004). In a recent study, the CysLT1R has been associated with cytokine transduction signals for the up-regulation of eosinophilopoiesis by IL-13 and eotaxin in murine bone marrow (Queto et al., 2010). These data support early reports demonstrating the inhibitory action of CysLT1R antagonists on eosinophil activation and migration (Virchow et al., 2001; Fregonese et al., 2002; Suzuki et al., 2003; Ueda et al., 2003; Saito et al., 2004; Nagata et al., 2005), as well as on adhesion (Fregonese et al., 2002; Nagata et al., 2002; Kushiya et al., 2006; Meliton et al., 2007; Profita et al., 2008). There is considerable information available to demonstrate the ability of CysLTR antagonists to reduce airway eosinophilia and eosinophil cationic protein (ECP) in animals (Underwood et al., 1996; Ihaku et al., 1999) and in humans (Pizzichini et al., 1999; Obase et al., 2002; Steinke et al., 2003; Strauch et al., 2003; Laitinen et al., 2005; Kopriva et al., 2006).
CysLT1Rs are also found to be expressed in B lymphocytes and CD34+ hematopoietic progenitor cells (Figueroa et al., 2001). Indeed, early observations demonstrated that LTD4-stimulate chemotaxis and transendothelial migration of CD34+ hematopoietic progenitor cells (Bautz et al., 2001; Mohle et al., 2003) as well as their proliferation (Braccioni et al., 2002; Parameswaran et al., 2004; Boehmler et al., 2009) and that these activities were suppressed by different CysLT1 receptor antagonists. Altogether these data indicate a physiological role for cysteinyl LTs as autocrine regulators of hematopoiesis corroborated by the expression of LTC4S in immature myeloid cells (Tornhamre et al., 2003).
Basophils have also been shown to accumulate in the airways of subjects with atopic asthma (Gauvreau et al., 2001), and these authors documented that human basophils express variable levels of functional CysLT1R. Of interest, they found that prolonged stimulation of cell cultures with LTD4 reduced the frequency of CD95 Fas receptor expression, an effect that, albeit modest, was reversed by the CysLT1R antagonist zafirlukast (Gauvreau et al., 2005). These results seem to suggest that the cysteinyl LT/CysLT1R system might have a direct effect on basophil accumulation in allergic tissues.
MCs are a primary source of cysteinyl LTs (Austen, 2005) after allergic or nonallergic stimulation and, accordingly to their classification as autacoids, CysLTRs are widely expressed in these cells (Mellor et al., 2001; Sjöström et al., 2002) and are responsible for their proliferation and activation (Jiang et al., 2006, 2009; Kaneko et al., 2009). In agreement with this finding, montelukast has been found to significantly reduce the number of MCs in the inflamed paws of mice with collagen-induced arthritis (Shiota et al., 2006).
The expression of both CysLT1R and CysLT2R has also been reported in human platelets by RT-PCR, Western blotting, and flow cytometry (Hasegawa et al., 2010). In this report, the authors demonstrate that cysteinyl LTs induced a release of RANTES and that this effect was inhibited by pranlukast, suggesting a novel role for platelets in allergic inflammation (Hasegawa et al., 2010).
In agreement with a role of cysteinyl LTs in asthma, CysLT1Rs are expressed by a variety of airway mucosal inflammatory cells and the numbers of inflammatory CysLT1R-expressing cells (eosinophils, neutrophils, MCs, macrophages, and B lymphocytes but not T lymphocytes) significantly increase in subjects with stable asthma and patients hospitalized for asthma exacerbation compared with control subjects (Zhu et al., 2005). Of interest, a strong positive correlation exists between this observation and the augmented numbers of CD45+ progenitors (cells expressing the pan-leukocyte marker CD45 antigen) (Zhu et al., 2005). In addition, human fibroblasts have been shown to express both CysLT1 (James et al., 2006; Vannella et al., 2007) and CysLT2 receptors (Vannella et al., 2007). These cells are also known to produce cysteinyl LTs, suggesting a regulation via an autocrine/paracrine secretion of these lipid mediators. Indeed, exogenous administration of LTD4 but not of LTC4 was also able to induce proliferation of both murine and human fibrocytes, and CysLT1R antagonists blocked this mitogenic effect (Vannella et al., 2007). In contrast, there are two reports that evaluated the contribution of cysteinyl LTs to fibroblast-mediated fibrosis and remodeling in chronic hyperplastic eosinophilic sinusitis (Steinke et al., 2004) or in healthy subjects (Yoshisue et al., 2007) and found that fibroblasts did not substantially express CysLT1R or CysLT2R, even after priming with IL-4 or IL-13 (Steinke et al., 2004). However, cysteinyl LTs have a broader capability to synergize with epidermal growth factor receptor (EGFR) to increased thymidine incorporation and cell proliferation, effects that, in agreement with the absence of CysLTRs, are not inhibited by specific CysLT1R antagonists (MK-571 and montelukast) or by the dual antagonist BAY U9773, pointing to a different mechanism/receptor (Yoshisue et al., 2007).
In human nasal mucosa, increased expression of CysLT1R has been found both at gene and protein levels in blood vessels and in the interstitial vascular ECs, as well as in eosinophils, MCs, macrophages, and neutrophils (Shirasaki et al., 2002, 2006) and in nasal polyps (Chao et al., 2006; Pérez-Novo et al., 2006). Because subjects with aspirin-induced asthma (AIA) have greater airway hyperresponsiveness to the effects of inhaled cysteinyl LTs than those with aspirin-tolerant asthma (ATA) (Arm et al., 1989), Sousa and colleagues (Sousa et al., 2002; Corrigan et al., 2005) hypothesize that this could be due to the elevated expression of CysLT1R on inflammatory cells. Their study on nasal biopsy specimens from patients with chronic rhinosinusitis and nasal polyps revealed that aspirin-sensitive rhinosinusitis is indeed characterized by increased numbers of nasal inflammatory leukocytes expressing the CysLT1R. These results have been confirmed (Ozen et al., 2007) and are consistent with two other reports in which the augmented receptor expression was observed for the majority of eosinophils and subsets of MNCs obtained by nasal lavage (Figueroa et al., 2003) or in sputum after allergen challenges in patients with seasonal allergic rhinitis (AR) (Boulay et al., 2010). Finally, CysLT1 but not CysLT2 mRNA has been demonstrated in nasal C-fiber neurons, where LTD4 directly increased the excitability of capsaicin-sensitive guinea pig nasal trigeminal neurons. These observations suggest a novel mechanism for the actions of cysteinyl LTs and offer an explanation for the observed beneficial use of CysLT1R antagonists in treating nasal allergen-induced neuronal symptoms (Taylor-Clark et al., 2008).
RT-PCR has allowed the demonstration that human saphenous veins express both a functional CysLT1R, which mediates contractile effects of cysteinyl LTs, but also a CysLT2R, which, on the contrary, does not seem to be implicated in contraction (Mechiche et al., 2004) and whose functional role remains to be determined (Allen et al., 1992). Human pulmonary veins also express both receptors and although activation of the former induced the release of a contractile factor only partially blocked by CysLT1R antagonists, the activation of the CysLT2R released nitric oxide (NO) (Ortiz et al., 1995). In addition, human pulmonary artery vascular SMCs seem to express at least two different receptors, a CysLT1R and a novel CysLTR subtype, both responsible for vasoconstriction (Walch et al., 2002). Thus, the mechanical effects of LTD4 on human pulmonary vasculature are complex and involve both direct and indirect mechanisms mediated via at least two, and possibly more, types of cysteinyl LT receptors (Walch et al., 2000).
In the gastrointestinal system, CysLT1R expression has been documented in the small intestine and colon (Lynch et al., 1999; Sarau et al., 1999). Subsequently, the presence of the CysLT1R subtype was reported in a nontransformed epithelial intestinal cell line (Ohd et al., 2003), corroborating previous findings for LTD4-induced signaling in these cells (Grönroos et al., 1995; Thodeti et al., 2000, 2001). In addition, CysLT1R functions were characterized in colorectal carcinoma cells (Ohd et al., 2003), and low expression of CysLT1R and high expression of CysLT2R correlated with high differentiation and good prognosis (Magnusson et al., 2007, 2010). The same results have been also extended to breast cancer cells and patient survival (Magnusson et al., 2011).
b. CysLT2 Receptor.
Localization studies have identified a distinctive expression pattern for human CysLT2R, despite some overlapping with CysLT1R. Indeed, expression in heart, particularly in Purkinje fiber cells, myocytes, and fibroblasts derived from atrium and ventricle, brain, and adrenal glands is specific to CysLT2R (Heise et al., 2000; Nothacker et al., 2000; Takasaki et al., 2000).
Several authors indicated that HUVECs almost exclusively express CysLT2R (Mita et al., 2001b; Lötzer et al., 2003), which is responsible for the calcium mobilization and contraction evoked in these cells by cysteinyl LTs, as well as by the selective CysLT2R agonist BAY U9773 (Sjöström et al., 2003; Carnini et al., 2011), for which activation results in a proinflammatory EC phenotype (Uzonyi et al., 2006). ECs are strategically located at the interface with blood circulation where they become exposed to neutrophil- and platelet-derived LTs, a setting that has potential implications for cardiovascular diseases (CVDs). Moreover, previous studies demonstrated that cysteinyl LTs can trigger several functional responses in ECs, such as PAF accumulation and neutrophil adhesion (McIntyre et al., 1986) and secretion of von Willebrand factor and of P-selectin surface expression (Datta et al., 1995). In light of these expression data, the responses data can be attributed to the CysLT2R and, accordingly, are not inhibited by selective CysLT1R antagonists (Pedersen et al., 1997).
Kamohara et al. (2001) were the first to report the presence of a functional CysLT2R on human coronary artery SMCs. These authors demonstrated that LTC4 enhanced [Ca2+]i, an effect that was not blocked by CysLT1R antagonists but blocked by the calcium channel blocker nicardipine. Further significance of cysteinyl LTs involvement in atherosclerosis arises from the observation that cysteinyl LTs induce contractions of human atherosclerotic coronary arteries, whereas nonatherosclerotic arteries are unresponsive (Allen et al., 1998). Taken together, these results suggest that the activation of CysLT2R can induce profound effects in cardiac as well as in hemodynamic and microcirculatory pathophysiology and that this receptor subtype represents an interesting pharmacological target in the future for CVDs.
Although little expression for CysLT1R has been reported in brain (Lynch et al., 1999; Sarau et al., 1999), apparent localization has been documented by immunohistochemical analysis in human brains with traumatic injury or tumors (Zhang et al., 2004). In contrast, CysLT2R mRNA is highly expressed in the central nervous system, with particular concentration in hypothalamus, thalamus, putamen, pituitary, and medulla (Heise et al., 2000) and in small, but not large, vessels in mouse brain (Moos et al., 2008). A number of articles have reported the involvement of either CysLT1 (Fang et al., 2006) or CysLT2 (Fang et al., 2007) or both receptors (Sheng et al., 2006; Wang et al., 2006; Huang et al., 2008) in the inflammatory process subsequent to brain vascular insults (vascular ischemia or oxygen deprivation). Spatiotemporal expression of CysLT2R mRNA in rat brain was observed after focal cerebral ischemia induced by middle cerebral artery occlusion, suggesting that CysLT2R may be related to the acute neuronal injury and late astrocyte proliferation in the ischemic brain (Fang et al., 2007).
In the immune system, moderate expression of CysLT2 mRNA was seen in spleen, lymph nodes, and peripheral blood leukocytes (Heise et al., 2000). A comparative study on the levels of CysLTR subtype expression in human peripheral blood leukocytes (eosinophils, neutrophils, monocytes, and T lymphocytes) indicated a significantly high expression for CysLT2 mRNA in eosinophils (Mita et al., 2001b), suggesting unidentified roles for this receptor in these cells. Mellor et al. (2003) reported that human MCs constitutively express the CysLT2R, whose proposed function seems to be the production of IL-8. CysLT2R expression was also reported in basophils (Gauvreau et al., 2005), but a functional role for this protein was not identified, despite flow cytometric analysis, which revealed an expression level equivalent to that of the CysLT1R. No CysLT2R expression was found in either undifferentiated or differentiated promyelocytic HL-60 and U937 cells (Nothacker et al., 2000; Capra et al., 2005).
CysLT2R was expressed in the majority of eosinophils and in subsets of MCs and MNCs, but not in neutrophils, obtained by nasal lavage of patients with seasonal AR (Figueroa et al., 2003). Sousa et al. (2002) and Corrigan et al. (2005) extended a previous study on CysLT1R expression in the nasal mucosa of subjects with chronic rhinosinusitis, both AIA and ATA, and found that the distribution of CysLT2R differs from that of CysLT1, with a predominance of CysLT2R on glands and epithelium. They also showed that nasal mucosal inflammatory leukocytes showed no evidence of up-regulation of CysLT2R in subjects with rhinosinusitis, whether AIA or ATA, compared with control subjects (Corrigan et al., 2005). However, CysLT2R up-regulation has been demonstrated in nasal polyp tissue and expression was correlated with eosinophilic inflammation (Pérez-Novo et al., 2006). In human lung, the CysLT2 mRNA signal was high in interstitial macrophages and weak in SMCs (Heise et al., 2000), in which, on the contrary, CysLT1 mRNA expression was elevated. Of interest, RT-PCR revealed that A549 cells, a human lung adenocarcinoma-derived line with alveolar epithelial cell properties, express mRNA for CysLT2R but not CysLT1R (Sloniewsky et al., 2004).
The adrenal gland may represent a novel tissue for future studies on cysteinyl LT functions and CysLT2R role in modulating the endocrine system because strong expression was detected in medullary pheochromocytes (Lynch et al., 1999; Nothacker et al., 2000).
3. Localization of CysLT Receptors.
The mechanism of action of cysteinyl LTs is thought to be primarily dependent on their specific plasma membrane receptors that, as discussed above, belong to the superfamily of GPCRs. However, a few reports indicate that CysLTR could also be localized to a nuclear compartment, suggesting major unanticipated roles for these receptors in cell signaling and function. Bandeira-Melo et al. (2002b) provided the first evidence for an intracrine CysLTR-induced signaling of eosinophil vesicular transport-mediated IL-4 secretion, which is PTX-sensitive. Furthermore, CysLT1R was found to be located in the outer nuclear membrane in colon cancer cells and may apparently be translocated into the nucleus after prolonged exposure to the agonist in nontransformed intestinal epithelial cells (Nielsen et al., 2005a). These authors also demonstrated that a nuclear localization sequence (localized in the C tail) is essential for the receptor translocation to the nucleus, as already demonstrated for other GPCRs (Boivin et al., 2008) as well as for other proteins that normally target the nucleus, such as 5-LO (Hanaka et al., 2002). Finally, CysLT1R expression in myofibroblasts derived from human heart valves exhibits a perinuclear pattern, and LTC4 increased nuclear [Ca2+] in those cells (Nagy et al., 2011).
Because other lipid mediators have been demonstrated to exert biological functions including transcriptional regulation of COX-2 and inducible NO synthase through their nuclear GPCRs (Marrache et al., 2005), one can speculate that the roles for nuclear receptors may be different from those on plasma membrane and that intracellular GPCRs may constitute a distinctive mode of action for gene regulation (Boivin et al., 2008). A recent report has shown that CysLT1R and CysLT2R, but also the purinergic P2Y12 receptor, are expressed on eosinophil granule membranes and that cysteinyl LTs stimulated isolated eosinophil granules to secrete ECP. Of interest, although montelukast inhibited ECP release, the P2Y12 receptor antagonist 2,2-dimethyl-propionic acid 3-(2-chloro-6-methylaminopurin-9-yl)-2-(2,2-dimethyl-propionyloxymethyl)-propyl ester (MRS 2395) also inhibited cysteinyl LT-induced ECP release, adding another piece of evidence to the CysLT-P2Y receptor interaction puzzle (Neves et al., 2010) (see sections III.D and III.H).
B. CysLT Receptor Expression and Immunoregulation
Human CysLT1R expression was up-regulated through augmented transcriptional activity by priming cells with Th2-like cytokines (i.e., IL-5 in eosinophil-differentiated HL-60 cells) (Thivierge et al., 2000) and IL-4 and IL-13 in human monocytes and monocyte-derived macrophages (Thivierge et al., 2001; Shirasaki et al., 2007). The regulation of human CysLT1R expression by IL-13 has also been reported in the literature for lung fibroblasts (Chibana et al., 2003) and airway SMCs together with TGF-β and IFN-γ, whereas, surprisingly, IL-4 was found to be ineffective (Espinosa et al., 2003). In the latter cells, up-regulation of CysLT1R by IFN-γ has been demonstrated, and this correlates with increases in LTD4-induced responses (Amrani et al., 2001). HUVECs express CysLT1R at a very low level (Sjöström et al., 2001), unless they are subjected to prolonged treatment with IL-1β (Gronert et al., 2001). The up-regulation by IL-1β and IFN-γ seems intriguing because these cytokines are generally considered to be counterregulatory, inhibiting both the production and activities of the proallergic Th2-like cytokines. Use of IFN-γ is indeed one of the theoretical approaches to obtain IgE immunomodulation in asthma, albeit adverse effects suggest limited use (Stokes and Casale, 2004). However, there are also investigators who suggest that IFN-γ is likely to be proinflammatory in many aspects of chronic allergic inflammation, evolving our understanding of classic Th2 diseases (Liu, 2000).
Priming human MCs with IL-4 did not result in the up-regulation of CysLT1R, either at the transcript or the protein level, but rather resulted in the appearance of an apparently new CysLTR also activated by the pyrimidinergic ligand UDP (Mellor et al., 2001). In a subsequent report, the same authors also demonstrated that human MCs primed with IL-4 responded to UDP, LTC4, and LTD4 by producing IL-5, TNF-α, and especially large quantities of macrophage inflammatory protein (MIP)-1β that were inhibited by the CysLT1R antagonist MK-571. Nevertheless, human MCs constitutively also express the CysLT2R, and IL-4 consistently up-regulated surface expression on a fraction of cells (Mellor et al., 2003). Whether IL-4 priming up-regulates a known CysLTR or induces the expression of a new receptor subtype (see section III.G) remains to be established.
A 2- to 3-fold enhancement of CysLT1R expression was observed in B lymphocytes after exposure to a combination of activating anti-CD40 antibody and IL-4. This enhancement, in turn, induced an increase in responsiveness to LTD4 in terms of Ca2+ flux and up-regulation of IgE and IgG production, which was totally prevented by the selective CysLT1R antagonist montelukast (Lamoureux et al., 2006). Therefore, Early et al. (2007) showed that IL-4, but not IL-13, was able to significantly induce mRNA and protein concentrations for CysLT1R from T and B lymphocytes.
CysLTRs can also be immunoregulated by cytokines that play an essential role in down-modulating adaptive and innate immune responses. IL-10, for example, down-regulated mRNA of CysLT1 and CysLT2 receptors in a time- and concentration-dependent fashion. In addition, cysteinyl LT-induced activation and chemotaxis of human monocytes and monocyte-derived immature DCs, measured by cytosolic Ca2+ flux and immediate-early gene expression, was potently decreased by IL-10 and by the CysLT1 antagonist MK-571 (Woszczek et al., 2008b). Of interest, maturation of DCs with LPS (Thivierge et al., 2006), a classic Toll-like receptor 4 agonist, or zymosan (Thivierge et al., 2009), a Toll-like receptor 2 agonist, down-regulated CysLT1R mRNA levels and protein expression and reduced functional responsiveness to LTD4. Indeed, the effect of zymosan was, at least partially, dependent on endogenous production of PGE2 and IL-10 (Thivierge et al., 2009). Therefore, montelukast prevents the decrease of IL-10 and inhibits NF-κB activation in inflammatory airway of “asthmatic” guinea pigs (Wu et al., 2006).
Transcript levels of CysLT2R seem to be up-regulated by the Th2-like cytokine IL-4 in HUVECs (Lötzer et al., 2003) but suppressed by Th1-like (proinflammatory) cytokines, such as TNF-α (Lötzer et al., 2003; Sjöström et al., 2003), LPS, or IL-1β (Sjöström et al., 2003) in a rapid and partially reversible manner. However, IFN-γ has been found to induce CysLT2R expression and to enhance the responsiveness to cysteinyl LTs of human ECs (Woszczek et al., 2007), as well as of monocytes, T cells, and B lymphocytes (Early et al., 2007). These data confirm early observations by Fujii et al. (2005), who postulated that CysLT2R might modulate exacerbations of asthma, as they observed that CysLT2R expression on eosinophils was up-regulated by IFN-γ and increased during asthma exacerbation, especially in nonatopic subjects. In addition, another proinflammatory cytokine, IL-18, has also been recently postulated to up-regulate CysT2R expression in HUVECs at the early stage of administration, accelerating cell apoptosis (Zhou et al., 2009).
Besides in HUVECs, CysLT2R expression was also significantly increased in monocytes and eosinophils after IL-4 priming (Early et al., 2007) or IL-13 treatment (Shirasaki et al., 2007). At variance with data reported previously for CysLT1R, CysLT2R expression increased after DC maturation induced by LPS, suggesting that these cells may differentially respond to cysteinyl LTs, depending on their maturational stimuli (Thivierge et al., 2006).
These observations suggest that the expression of the CysLTRs may be functionally up-regulated, mainly by Th2-like cytokines (consistent with the classic view of the pathobiological mechanisms of asthma and other allergic diseases), but also by a classic Th1 cytokine (i.e., IFN-γ) at least in some cells (Table 3). This finding raises the possibility that constitutive and inducible receptors may or may not behave in a similar way, prompting a detailed pharmacological examination of receptors expressed before and after priming.
C. Intracellular Signaling Pathways and Second-Messenger Systems
1. CysLT1 Receptor.
In consideration of the crucial bronchoconstrictor activity of cysteinyl LTs (Hallstrand and Henderson, 2010) and their role in asthma (Capra et al., 2007), intracellular Ca2+ mobilization and phosphatidylinositol (PI) metabolism were the obvious signal transduction systems to investigate (for a review of early reports, see Brink et al., 2003; Rovati and Capra, 2007). Upon cloning, LTD4-induced functional responses in oocytes (Lynch et al., 1999) or HEK-293 cells (Sarau et al., 1999) indicated that, at least in these systems, CysLT1R is very weakly, if not at all, coupled to a PTX-insensitive G protein (Gq/11).
However, an initial study demonstrated in circulating MNCs that cysteinyl LTs modulate Ca2+ responses through a PTX-sensitive G protein (Gi/o) (Baud et al., 1987). In further reports, investigators have shown that the Ca2+ modulation may be through two distinct G proteins, one PTX-sensitive and one insensitive, in monocyte/macrophage U937 cells (Pollock and Creba, 1990; Capra et al., 2003) or in a human epithelial cell line (Sjölander et al., 1990; Adolfsson et al., 1996). Subsequently, CysLT1R-dependent actin reorganization was shown to be coupled with a PTX-sensitive G protein and to Rho in an intestinal epithelial cell line (Grönroos et al., 1996; Massoumi and Sjölander, 1998; Massoumi et al., 2002) or in bronchial SMCs (Saegusa et al., 2001). Thus, these data confirm CysLT1R promiscuity in G protein coupling in constitutive systems, in good agreement with the finding that LTD4 activates distinct signaling pathways differently coupled to G proteins in MNCs (Saussy et al., 1989; Skoglund and Claesson, 1991; Hoshino et al., 1998; Capra et al., 2003) or intestinal epithelial cells (Grönroos et al., 1998; Paruchuri and Sjölander, 2003; Nielsen et al., 2005b). Among these pathways several groups have demonstrated the activation of MAPK by CysLT1R in THP-1 cells (Hoshino et al., 1998), renal mesangial cells (McMahon et al., 2000), intestinal epithelial cells (Paruchuri et al., 2002), monocyte/macrophage U937 cells (Capra et al., 2004), astrocytes (Ciccarelli et al., 2004), colon cancer cells (Parhamifar et al., 2005), MCs (Jiang et al., 2006), and airway SMCs (Ravasi et al., 2006). Cysteinyl LTs have been shown to induce proliferation in a variety of cells, such as human hematopoietic cell lines (Snyder et al., 1989; Braccioni et al., 2002; Parameswaran et al., 2004), airway epithelial cells (Leikauf et al., 1990), vascular (Porreca et al., 1995, 1996; Kaetsu et al., 2007) and airway SMCs (Cohen et al., 1995; Panettieri et al., 1998; Espinosa et al., 2003; Bossé et al., 2008), glomerular mesangial cells (Kelefiotis et al., 1995), intestinal epithelial cells (Paruchuri and Sjölander, 2003), astrocytes (Fang et al., 2006; Huang et al., 2008) and fibrocytes (Vannella et al., 2007).
There are many possible pathways through which GPCRs may induce ERK1/2 activation, one of which requires the transactivation of a growth factor receptor. Indeed, a first report suggested that LTD4 synergizes with the insulin growth factor axis to induce airway SMC proliferation (Cohen et al., 1995), involving proteolysis of airway SMC-produced inhibitory insulin growth factor-binding proteins by LTD4-induced MMP-1 (Rajah et al., 1996). In the same cellular system, LTD4 has been demonstrated to induce phosphorylation of apoptosis signal-regulating kinase 1 (Kumasawa et al., 2005), a kinase upstream of c-Jun NH2-terminal kinase (JNK) and p38 MAPK, which in turn regulates transcription factor AP-1, an essential step for regulation of cell proliferation and differentiation. In renal mesangial cells, LTD4-induced proliferation requires ERK1/2 and p38 activation and is dependent on PI3K and PKC (McMahon et al., 2000). LTD4 also transactivates the platelet-derived growth factor receptor β, which is a process associated with c-Src recruitment and Ras activation, an effect insensitive to PTX and apparently related to CysLT1R activation (McMahon et al., 2002). LTD4-induced airway SMC proliferation was also demonstrated to require transactivation of the EGFR through generation of reactive oxygen species (Ravasi et al., 2006). It is noteworthy that very recently LTD4 has been demonstrated to transcriptionally activate VEGF production via CysLT1R, with the involvement of JNK, ERK, the AP-1 complex, and Sp1 (Poulin et al., 2011). Taken together, these findings suggest that cysteinyl LTs may be important in the process of airway remodeling and potentially provide a previously unknown benefit of using LTRAs in the prevention or treatment of chronic asthma, albeit long-term studies aimed to determine its effects on airway remodeling are still lacking. Finally, in human MCs, LTD4 enhanced proliferation in a CysLT1- and ERK-dependent manner, which in turn required transactivation of c-kit (Jiang et al., 2006).
However, in THP-1 cells LTD4 has been postulated to activate MAPK through a PKC-Raf-1-dependent pathway (Hoshino et al., 1998), whereas in differentiated U937 cells ERK1/2 activation involves a Ras-GTP-dependent pathway, phospholipase C, and Ca2+-dependent tyrosine kinase(s) (Capra et al., 2004). In addition, LTD4 has been shown to induce proliferation and migration of mouse embryonic stem cells via a mechanism involving STAT-3, PI3K, glycogen synthase kinase-3β/β-catenin phosphorylation and calcineurin expression (Kim et al., 2010).
In intestinal epithelial cells, LTD4 has been shown to increase cell survival and proliferation (Ohd et al., 2000; Paruchuri and Sjölander, 2003), activating MAPK through a Ras-independent but PKCε-dependent pathway (Paruchuri et al., 2002), and to increase cell mobility via a PI3K/Rac signaling pathway (Paruchuri et al., 2005). Furthermore, the antiapoptotic effect involves the prevention of caspase 8 activation and Bid cleavage (Wikström et al., 2003a), as well as COX-2 transcription and Bcl-2 up-regulation mediated through a PTX-sensitive G protein and the ERK-1/2 pathway (Wikström et al., 2003b). Indeed, LTD4 exposure, through PI3K-dependent phosphorylation of glycogen synthase kinase-3β induced a β-catenin translocation to the nucleus, where there is an elevation of the promoter activity of the TCF/LEF family of transcription factors, and to the mitochondria, where the action is associated with the cell survival protein Bcl-2 (Mezhybovska et al., 2006). CysLT1R involvement in colon cancer cell proliferation was also shown via endogenous production of cysteinyl LTs that, through cytosolic phospholipase A2 activation (Parhamifar et al., 2005), mediates an autocrine survival and proliferation signal in nontumor- and tumor-derived epithelial cells (Paruchuri et al., 2006). To the best of our knowledge, there is only one report suggesting that cysteinyl LTs may, on the contrary, inhibit the growth of a human cell line (i.e., the mammary cancer MCF-7 cells) (Przylipiak et al., 1998).
Proliferation is not the only effect induced by MAPK activation. For example, in monocytes, LTD4 induced p38 phosphorylation (Woszczek et al., 2008a), a pathway involved in the regulation of immediate-early gene expression. In a human epithelial cell line, CysLT1R activation has been shown to lead to either STAT-1 phosphorylation through PKC and ERK1/2 activation, causing enhanced intercellular cell adhesion molecule-1 expression and eosinophil adhesion (Profita et al., 2008), or to up-regulate mucin gene MUC2 transcription via a signaling pathway involving PKC and NF-κB (Suzuki et al., 2008). In this respect, NF-κB, a transcription factor largely involved in inflammation and in regulating the immune response, has been demonstrated to be involved in the CysLT1R transduction pathway by different groups. In isolated lung MNCs, for example, LTD4 activated NF-κB and induced production of RANTES (Kawano et al., 2003), whereas in THP-1 and in human DCs CysLT1R engagement induced AP-1- and NF-κB-dependent IL-8 expression (Thompson et al., 2006), as well as increased MCP-1 (Hashimoto et al., 2009). In partial agreement with these data LTD4 has been demonstrated to induce AP-1, but not NF-κB signaling in intestinal epithelial cells (Bengtsson et al., 2008). Therefore, montelukast inhibited NF-κB activation in THP-1 cells in a dose-related manner (Maeba et al., 2005), and TNF-α-stimulated IL-8 expression through changes in NF-κB p65-associated histone acetyltransferase activity in differentiated U937 cells (Tahan et al., 2008).
However, three distinct studies have reported that LTD4 did not induce NF-κB nuclear translocation. In contrast, pranlukast inhibited NF-κB activation (Ichiyama et al., 2003), TNF-α production (Tomari et al., 2003), or MUC2 gene transcription (Ishinaga et al., 2005), effects that were suggested to be independent from CysLT1R antagonism. A different study also reported that pranlukast inhibited IL-5 production in various cells irrespective of their CysLT1R mRNA expression (Fukushima et al., 2005), again suggesting that the compound may have other activities beyond CysLT1R antagonism. Furthermore, montelukast and zafirlukast have been demonstrated to inhibit the effects of nucleotides acting at different P2Y receptors (P2Y1,2,4,6) (Mamedova et al., 2005), although in a noncompetitive manner. These observations have recently been confirmed by another report demonstrating that montelukast and zafirlukast, acting in a concentration-dependent manner, can inhibit non–CysLT1R-mediated proinflammatory reactions in human monocytes while inhibiting UDP-induced Ca2+ mobilization (Woszczek et al., 2010). Taken together, these data demonstrate non-CysLTR-related activities for these classic LTRAs. Thus, the therapeutic effects of these compounds must be considered with some reserve.
Another interesting cellular function that emerged to be modulated by CysLT1R is the up-regulation of β integrins. Massoumi and Sjölander (2001) first demonstrated that in intestinal epithelial cells LTD4 induced a Src-dependent rapid tyrosine phosphorylation of vinculin, as well as PKC-dependent up-regulation of active β1 integrins on the cell surface and a consequent enhanced adhesion of cells to collagen IV. Furthermore, in Caco-2 cells, LTD4 controlled adhesive properties and migration by up-regulating COX-2 and stimulating PGE2-induced expression of α2β1 integrins (Massoumi et al., 2003). In another study, LTD4 was demonstrated to rapidly induce focal adhesion kinase-related tyrosine kinase phosphorylation and significantly up-regulated α4β1 and α5β1 integrin-dependent adhesion of both primitive and committed hematopoietic stem and progenitor cell (Boehmler et al., 2009).
In some systems, CysLT1R activation can contribute to the propagation of the inflammatory reaction by the release of various inflammatory mediators and cytokines. Thus, although cytokines may regulate CysLTR expression, cysteinyl LTs may, in turn, induce their release in an amplifying circuit of inflammation. For example, in U937 cells, LTD4 triggered a rapid release of arachidonic acid metabolites into the culture medium, an effect that was suppressed by the CysLTR antagonist 2(S)-hydroxyl-3(R)-carboxyethylthio)-3-[2-(8-phenyloctyl) phenyl] propanoic acid (SKF 104353), by the topoisomerase I inhibitor camptothecin, and by staurosporine (Mattern et al., 1990). Several different groups have reported CysLT1R-induced release of Th2 cytokines, such as IL-4 from cord blood-derived eosinophils (Bandeira-Melo et al., 2002a,b), IL-5 from MNCs (Nabe et al., 2002; Frieri et al., 2003; Faith et al., 2008), chemokines, such as RANTES, from MNCs (Kawano et al., 2003) or platelets (Hasegawa et al., 2010), IL-8 from THP-1 cells (Thompson et al., 2006), and MIP-1α and MIP-1β (Ichiyama et al., 2009) or IL-11 (Lee et al., 2007a) from epithelial cells. Of interest, LTC4 and LTD4, but not LTE4, enhanced TNF-α-induced MMP-9 production in human MNCs (Ichiyama et al., 2007), an effect that, although inhibited by montelukast, does not seem to be CysLT1R-dependent in eosinophils (Langlois et al., 2006). Finally, a combination of LTD4 and IL-5 or granulocyte macrophage–colony-stimulating factor synergistically induced TGF-β1 expression in eosinophils (Kato et al., 2005), whereas LTC4 induced TGF-β1 production in airway epithelial cells (Perng et al., 2006) or in a time- and concentration-dependent manner in CysLT1R-transfected HEK-293 cells (Bossé et al., 2008).
Thus, montelukast and/or pranlukast significantly reduced in OVA-treated mice the increased IL-4 and IL-13 (Henderson et al., 2002; Lee et al., 2007a) in BAL fluid, IL-11 and TGF-β1 in BAL and lung tissues (Lee et al., 2007a), IL-4, IL-5, IL-13, and eotaxin in BAL, and TGF-β gene expression in GATA-3-overexpressing mice (Kiwamoto et al., 2011), thereby preventing airway inflammation, bronchial hyperresponsiveness, and airway fibrosis. Of interest, LTD4 in HEK-293 cells stably transfected with the CysLT1R can transcriptionally activate furin production with consequent maturation of furin, a proprotein convertase involved in the maturation/activation of several substrates implicated in the remodeling processes (Thompson et al., 2008b).
Treatment with montelukast decreased the amount of IL-4 but increased the amount of INF-γ mRNA expression in the lungs of IL-2-treated OVA-sensitized rats (Nag et al., 2002), decreased tissue inflammation, decreased total cell count and number of eosinophils and lymphocytes in BAL and inhibited the increase in endothelin-1 and INF-γ after intratracheal Sephadex provocation in rats (Finsnes et al., 2002). Furthermore, montelukast significantly inhibited LPS-induced IL-6, TNF-α, and MCP-1 production in the peripheral blood MNCs of control subjects and patients with asthma (Maeba et al., 2005), whereas pranlukast decreased the levels of many cytokines and chemokines (IL-4, IL-5, IL-1β, TNF-α, RANTES, and IL-8) in nasal mucosa, leading to improvement in patients with nasal symptoms (Ueda et al., 2003).
2. CysLT2 Receptor.
Obtaining detailed information about signal transduction systems involved in CysLT2R activation was hampered by the absence of selective antagonists, at least until a short time ago (Huang et al., 2008; Wunder et al., 2010). Thus, most of the data still rely on either recombinant systems or have been obtained in ECs that predominantly express CysLT2R. As in the case of the CysLT1R, the CysLT2R also seems to be mainly coupled to PI hydrolysis and intracellular Ca2+ mobilization through a PTX-insensitive G protein (Gq/11) in recombinant systems (Heise et al., 2000). CysLT2R activation in HUVECs causes an increase in [Ca2+]i (Lötzer et al., 2003), confirming early reports in which LTC4 and LTD4 were demonstrated to induce a rapid [Ca2+]i transient that was inhibited by the receptor antagonist pobilukast (Heimbürger and Palmblad, 1996), which, despite being considered a CysLT1R antagonist, is apparently a dual CysLT1/CysLT2R antagonist (G. E. Rovati, unpublished observations). Furthermore, in the same cells cysteinyl LTs induce myosin light-chain kinase activation, stress fiber formation, and EC contraction that are totally PTX-insensitive (Carnini et al., 2011). In contrast, in human MCs under conditions in which CysLT1R is blocked by MK-571 pretreatment, CysLT2R stimulation elicited secretion of IL-8 through p38 activation, an effect completely inhibited by PTX (Mellor et al., 2003).
As mentioned previously, in ECs of human vasculature CysLT2R induced relaxation (Allen et al., 1992), an effect that was successively linked to the formation of NO, at least in human pulmonary arteries and veins (Ortiz et al., 1995). In addition to direct action on the vascular tone, cysteinyl LT-induced activation of ECs may also lead to changes in the transcriptional activity. In HUVECs, LTD4 induces endothelial P-selectin expression (Pedersen et al., 1997), strongly stimulates expression of MIP-2 (Zhao et al., 2004), and together with thrombin up-regulates 37 early inducible genes, among which the most strongly induced are early growth response 1 (EGR1), nuclear receptor subfamily 4 group A transcription factors, E-selectin, CXC ligand 2, IL-8, a disintegrin-like and metalloprotease (reprolysin type) with thrombospondin type 1 motif 1 (ADAMTS1), Down syndrome critical region gene 1 (DSCR1), tissue factor, and COX-2 (Uzonyi et al., 2006). IL-8 induction has been studied in more detail, demonstrating that in HEK-293 cells stably transfected with CysLT2R LTC4 transcriptionally activates its production through induction of NF-κB and AP-1 transcription factors (Thompson et al., 2008a). However, at variance with the results in HUVECs, a recent article suggested that in EA.hy926 cells, a human EC line, LTD4 induced the phosphorylation of ERK1/2, but not that of p38 or JNK and that these effects were blocked by the CysLT1R antagonist montelukast or by the dual antagonist BAY U9773, but not by a novel CysLT2R antagonist referred to as Bay cysLT2 (Yuan et al., 2009).
D. Desensitization and Cross-Talk of CysLT Receptors
Early reports showed that in rat basophilic leukemia cells (Vegesna et al., 1988) and in differentiated U937 cells (Winkler et al., 1988; Pollock and Creba, 1990) treatment with phorbol 12-myristate 13-acetate reduced LTD4-induced PI metabolism and [Ca2+]i mobilization, whereas inhibitors of PKC selectively enhanced LTD4-induced [Ca2+]i transients (Winkler et al., 1990). These observations suggested for the first time that PKC might play a role in determining the responsiveness of CysLT1R. Indeed, native human CysLT1R has been demonstrated to be the target for unidirectional extracellular nucleotide-mediated heterologous desensitization (Capra et al., 2005). In fact, ATP/UDP-induced CysLT1R desensitization, which was sensitive to PKC inhibition, did not cause receptor internalization and induced a faster recovery of CysLT1R functionality with respect to LTD4-induced homologous desensitization. The latter, which was not sensitive to PKC activation by phorbol 12-myristate 13-acetate, induced a fast receptor trafficking and down-regulation, and has been shown to most likely depend on GRK2 activation. Of interest, activation of the CysLT1R by LTD4 had no effect on P2Y receptor responses (Capra et al., 2005).
In addition, heterologous desensitization of CysLT1R is mediated not only by Gq/i-coupled receptors but also by Gs-coupled receptors [i.e., β2-adrenoreceptors (β2AR), histamine H1/2 and prostaglandin EP2/4 receptors] in differentiated U937 cells and human monocytes, this time through activation of PKA (Capra et al., 2010). Of interest, heterologous desensitization seems to affect mostly the Gi-mediated signaling of the CysLT1R (Capra et al., 2005, 2010).
However, Naik et al. (2005) reported that agonist-induced internalization of recombinant human CysLT1R expressed in HEK-293 cells is GRK/arrestin-independent and significantly PKC-dependent. They also demonstrated that mutation of putative PKC phosphorylation sites in the C-tail of the receptor reduced internalization and increased the signaling, significantly attenuating the effects of PKC inhibition. These findings may reflect the use of recombinant systems that produce varying results, depending on cell type and G protein availability, especially when dealing with GPCRs (Kenakin, 1997). In addition, both PKCα and PKCε have been shown to be activated and involved in the regulation of the LTD4-induced Ca2+ signal in human SMCs (Accomazzo et al., 2001) and intestinal epithelial cells (Thodeti et al., 2001). Therefore, in human airway SMCs, PKC inhibition augmented LTD4-stimulated contraction and increased PI hydrolysis and Ca2+ flux, confirming that PKC can regulate CysLT1R responses (Deshpande et al., 2007). Such an integrated network presumably tunes the strength and duration of the inflammatory signals, thereby leading to fine regulation of the physiological responses.
Although a phylogenetic relationship exists between P2YR and CysLTR on the basis of sequence alignment followed by evolutionary analyses (Fredriksson et al., 2003; Kroeze et al., 2003), functional relationships between cysteinyl LT/CysLTR and purine/P2YR systems are still a matter of debate. Mellor et al. (2001, 2002) first suggested that UDP is also able to activate CysLTRs and that UDP and LTD4 mediated ERK activation and cytoprotection, implying a cysteinyl LT-mediated autocrine loop (Jiang et al., 2009). However, UDP was not able to compete for binding sites labeled by [3H]LTD4 in COS-7 cells transiently expressing hCysLT1R. These data, together with the lack of cross-desensitization between CysLT1R and P2Y receptors, seem to exclude the involvement of a CysLTR/P2YR heterodimer (Capra et al., 2005). These observations have been recently confirmed by another group using receptor desensitization and small interfering RNA experiments in monocytes and CysLT1R-transfected HEK-293 cells. These investigators reported that LTD4 and UDP were acting on separate receptors (Woszczek et al., 2010). On the other hand, Nonaka et al. (2005) first reported that LTE4 acts as an agonist at the P2Y12 receptor, data recently confirmed by a different laboratory (Fredman et al., 2010), which seemed to corroborate the demonstration of LTE4-induced pulmonary inflammation mediated through the P2Y12 receptor (Paruchuri et al., 2009). In conclusion, a number of investigations demonstrated a close interplay between these two receptor systems, a topic that may have an impact on the development of therapeutic drugs (see section III.H).
The purinergic receptor, however, is not the only receptor system shown to interact with the CysLTRs. A number of studies have reported that LTRAs used as add-on therapy decrease the need for β2-agonists (Nathan et al., 1998; Noonan et al., 1998; Jarvis and Markham, 2000; Virchow et al., 2000; Keam et al., 2003; Vaquerizo et al., 2003; Deykin et al., 2007; Borderias et al., 2007) and that a positive association exists between a CysLTR antagonist and a β2-agonist in controlling asthma symptoms, particularly in patients homozygous for the glycine-16 β2AR polymorphism, which predisposes them to agonist-induced down-regulation and desensitization of the β2AR (Lipworth et al., 2000). The response to β2-agonist agonists is reduced in asthmatic airways, and this desensitization has been postulated to be due in part to inflammatory mediators such as cysteinyl LTs (Song et al., 1998; Milanese et al., 2005). Indeed, LTD4, through CysLT1R activation, induced β2AR desensitization in human airway SMCs in vitro, an effect again mediated by the PKC pathway (Rovati et al., 2006). Another interesting aspect of CysLTR cross-talk is the growth factor receptor transactivation, an aspect already discussed in section III.C.
E. CysLT Receptor Functional Analysis through Altered Gene Expression
1. Gene Disruption.
In 2001, a short isoform and a long isoform of mCysLT1R were identified in the mouse (Maekawa et al., 2001; Martin et al., 2001). Comparison of the human and the mouse CysLT1R cDNAs demonstrated that human transcript II contains sequences equivalent to the mouse short isoform (Woszczek et al., 2005). Both cDNAs, the human transcript II and the mouse short isoform, are less preferentially expressed (Woszczek et al., 2005). At the same time, the mCysLT2R was cloned and two 5′-untranslated region splice variants were identified, with the short form lacking exon 3 as the predominant transcript (Hui et al., 2001). A pharmacological difference was also noted between mouse and human CysLT2R. Pranlukast, a specific inhibitor for hCysLT1R, antagonized mCysLT2R responses at concentrations of 10 μM, as determined by [Ca2+]i elevation and receptor-induced promoter activation (Ogasawara et al., 2002).
Soon thereafter, Maekawa et al. (2002) generated CysLT1R-deficient mice by targeted gene disruption; this resulted in loss of both isoforms of the receptor. Peritoneal macrophages from wild-type mice, which expressed both CysLT1R and CysLT2R, responded strongly to LTD4 and slightly to LTC4 in a Ca2+ mobilization assay, whereas those from mutant mice responded to neither ligand. Furthermore, the same authors observed that plasma protein extravasation, but not neutrophil infiltration, was significantly suppressed in CysLT1R-null mice after zymosan-induced peritonitis or IgE-mediated passive cutaneous anaphylaxis, demonstrating that in acute inflammation the microvasculature responds through the CysLT1R to cysteinyl LTs generated in vivo by either monocytes/macrophages or MCs. The finding that CysLT1R-null mice were not protected against chronic injury induced by bleomycin suggested a proinflammatory role for the CysLT2R in the chronic response. This finding was corroborated by the observation that bleomycin-induced pulmonary inflammation with fibrosis is significantly attenuated in LTC4S-null mice but markedly aggravated in CysLT1R-null mice (Beller et al., 2004a). The same group of researchers examined this issue by generating CysLT2R-deficient mice (Beller et al., 2004b) and found that CysLT2R is responsible for mediating the contribution of the cysteinyl LTs to bleomycin-induced chronic pulmonary inflammation and fibrosis. In conclusion, these authors unexpectedly found that CysLT1R does not mediate the signal for chronic inflammation and fibrosis. They propose that CysLT1R is responsible for bronchoconstriction, for microvasculature leakage in acute inflammation, and for a counteracting role in chronic inflammation, whereas the CysLT2R is the receptor providing the signal for chronic inflammation. Furthermore, CysLT2R-targeted gene disruption revealed a significant reduction in vascular permeability (Beller et al., 2004b). These data have been further confirmed in transgenic mouse models (see below).
GPR17-deficient mice sensitized and challenged with dust mite exhibited markedly increased pulmonary inflammatory and serum IgE responses compared with those of wild-type mice. In contrast, a reduced response to this intervention was observed in mice lacking CysLT1R. These findings reveal a constitutive negative regulation of CysLT1R functions by GPR17 in both the antigen presentation and downstream phases of allergic pulmonary inflammation (Maekawa et al., 2010), confirming the in vitro observation of an interplay between GPR17 and CysLT1R (Maekawa et al., 2009) (see section III.H).
2. CysLT Receptor Transgenic Models.
A transgenic mouse specifically overexpressing the human CysLT1R in SMCs via the α-actin promoter was generated to explore the importance of increased cysteinyl LT signaling in airway function. This transgenic model overcame the problem of mice and isolated murine tracheal rings having a minimal bronchoconstrictor response to direct LT challenge compared with that of guinea pigs or humans (Dahlén et al., 1980; Ashida et al., 1987; Martin et al., 1988). Mice were sensitized by intraperitoneal injections with Aspergillus fumigatus to obtain a model of allergic airway inflammation. The authors observed that in transgenic mice allergen-induced airway hyperresponsiveness was significantly enhanced after LTD4 challenge, thus indicating that overexpression of human CysLT1R on SMCs leads to smooth muscle hyperresponsiveness (Yang et al., 2004).
Hui et al. (2004) have generated mice in which the human CysLT2R is overexpressed in ECs and reported the effects on endothelial integrity and blood pressure regulation mediated by this receptor subtype in vivo. These researchers observed that, in CysLT2R transgenic mice, the permeability response to exogenous LTC4 and to endogenous cysteinyl LTs evoked by passive cutaneous anaphylaxis was augmented and that this enhanced vascular permeability is controlled via Ca2+ signaling and takes place via a transendothelial vesicle transport mechanism as opposed to a paracellular route (Moos et al., 2008). Furthermore, the rapid, systemic pressor response to intravenous LTC4 was diminished coincidentally with augmented production of NO (Hui et al., 2004). The same group also observed that overexpression of the hCysLT2R in mice aggravates myocardial ischemia-reperfusion injury by increasing endothelial permeability and exacerbating inflammatory gene expression, leading to accelerated left ventricular remodeling, induction of peri-infarct zone cellular apoptosis, and impaired cardiac performance (Jiang et al., 2008).
F. Pharmacogenetics of CysLT System
The characterization of the genes encoding the cysteinyl LT system, including the CysLTR1 and CysLTR2 genes, has advanced the study of CysLTR pharmacology and the GPCR pharmacogenomics of inflammatory disease (Thompson et al., 2005). Furthermore, many of the genes encoding proteins involved in the synthesis and signaling pathways of the cysteinyl LTs are polymorphic and are of interest in genetic studies of asthma (Silverman et al., 2001; Palmer et al., 2002; Israel, 2005).
1. CysLT1 Receptor Pharmacogenetics.
The CysLT1R is a candidate gene in atopic asthma because drugs that act as high-affinity antagonist ligands of the CysLT1R, such as montelukast, zafirlukast, and pranlukast, have been shown to have a beneficial therapeutic effect in some cases of atopy that result in asthma (for reviews on LTRAs in therapy, see Bousquet et al., 2009; Diamant et al., 2009; Dunn and Goa, 2001; Keam et al., 2003; Wahn and Dass, 2008), allergic rhinitis (AR), or rhinosinusitis (Grayson and Korenblat, 2007) (see section III.I.1). Moreover, approximately 20% of patients report no change in symptoms during treatment with these agents (Drazen et al., 1999, 2000; Israel, 2005). This finding suggests that in some patients with asthma, the pathway regulating airway cysteinyl LT activity may not be activaed with the clinical symptoms of asthma resulting from the actions of other mediators, which would lead to the clinical phenotype of resistance to LTs modifiers (Asano et al., 2002; Lima et al., 2006; Telleria et al., 2008; Tantisira et al., 2009; Langmack and Martin, 2010).
Although the CysLT1R gene may harbor inactivating mutations in some populations, there are many more variants of the CysLT2R (Thompson et al., 2003). For example, in studies of the people living on the isolated south Atlantic island of Tristan da Cunha, two variants of the CysTL1R have been identified, an I206S variant and an activating G300S variant (Thompson et al., 2007). Although the G300S variant was found to be a mildly activating in response to LTD4 in vitro, it was nonetheless associated with atopy in the Tristan da Cunha population.
Furthermore, a number of SNPs in the promoter region of the CYSLTR1 gene have been found by different groups. Some SNPs were not associated with either the development of asthma/rhinitis (Choi et al., 2004; Zhang et al., 2006) or the response to treatment (Lee et al., 2007b), whereas others, in particular the −927T > C polymorphism (Arriba-Mendez et al., 2006; Hao et al., 2006; Sanz et al., 2006; Al-Shemari et al., 2008; Bizzintino et al., 2009), have been postulated to cause differences in transcriptional activity that may be predictors of disease risk (Duroudier et al., 2007; Hong et al., 2009), or treatment response (Kim et al., 2007b) or be associated with AIA (Kim et al., 2006b, 2007a).
As mentioned before, Lynch et al. (1999) mapped the CysLT1 gene to the long arm of the human X chromosome. Therefore, Ohd et al. (2003) found a significantly higher frequency of CysLT1R staining in tissues from male patients affected by colorectal adenocarcinoma, suggesting the presence of allele-dependent dysregulation or sex-specific transcriptional subordination. Although there are four SNP variants in the promoter region of the CYSLTR1 gene, two of the haplotypes, termed CAAC and TGGC, constitute more than 95% of the alleles noted in small population samples. There was no predisposition in patients with asthma or rhinitis to have one haplotype or the other. However, among women with asthma who were homozygous for the CAAC haplotype, there was a significantly higher expression of the CysLT1R transcript I than in women who had the CAAC/TCGC complex genotype; transcript II levels were also significantly lower than those in healthy women homozygous for the CAAC haplotype. These data are concordant with previous findings that TCG and CAA haplotypes in the promoter region are associated with different transcriptional activity (Zhang et al., 2006). Of interest, in a recent report of the British 1958 birth cohort, data suggested that the CYSLTR1 gene promoter polymorphisms might influence the risk of atopy in the female white population with suggestive evidence of heterozygote vigor (Duroudier et al., 2009).
2. CysLT2 Receptor Pharmacogenetics.
The CysLT2R is also a candidate gene for asthma and atopy because it maps to a region of chromosome 13q14 that has been linked to asthma (Kimura et al., 1999) and atopy (Daniels et al., 1996). Specific CysLT2R antagonists have only recently been reported (Huang et al., 2008; Wunder et al., 2010), and thus the pharmacological importance of CysLT2R antagonism in asthma or atopy is yet to be established. However, the development of drugs that target the CysLT2R may be useful not only for asthma, given that approximately 20% of patients fail to respond to CysLT1R agents, particularly when the CysLT2R is also polymorphic, but also for cardiovascular diseases (Funk, 2005; Moos and Funk, 2008; Riccioni et al., 2008).
The Tristan da Cunha study has identified an atopy associated with the M201V variant, which, unlike the G300S CysLT1 variant, is partially inactivating in terms of [Ca2+]i signaling (Thompson et al., 2003) and has been demonstrated to have a reduced affinity for LTC4 and particularly for LTD4, an impairment of IL-8 transactivation as well as NF-κB and JNK phosphorylation (Brochu-Bourque et al., 2011). This association of the CysLT2R M201V variant with airway disease has been independently replicated in a large outbred population (Pillai et al., 2004). Transmission disequilibrium testing in 137 Japanese asthmatic families revealed that the −1220A>C polymorphism of the CYSLTR2 gene is associated with the development of asthma (Fukai et al., 2004), whereas four other sequence variants (either in the 5′- or 3′-flanking region or downstream of the gene) are associated with AIA (Park et al., 2005) and have been linked to altered CysLT2R expression (Shin et al., 2009). It is noteworthy that two other CYSLTR2 gene variants, in association with an ALOX5 polymorphism, have been shown to increase the response to montelukast (Klotsman et al., 2007).
The fact that CysLT1R and CysLT2R are both polymorphic in some individuals suggests that if they are coexpressed on inflammatory cells, they may functionally alter cysteinyl LT signaling. If, like many GPCRs, the CysLT1R and CysLT2R form functional dimers with unique pharmacological properties (Jiang et al., 2007), the CysLT2R may also play a role in the CysLT1R pharmacology (see section III.G). Paradoxically, in the inbred Tristan da Cunha population the activating CysLT1R G300S variant and the inactivating CysLT2R M201V variant seem to interact to confer risk for atopy. In this sample, all double heterozygotes were atopic. General linear modeling indicated that the CysLT1R and CysLT2R genotypes were independently associated with atopy and asthma phenotypes. The gene-gene interaction, therefore, presents as being consistent with an additive effect that is highly associated with atopy (Thompson et al., 2007).
G. Homodimers or Heterodimers?
Some experimental evidence indicates the possibility that CysLTR might exist as homodimers and/or heterodimers, on the basis of the observation of dimeric and oligomeric forms of CysLT1R in Western blots and the punctate appearance of the immunohistochemical signal in peripheral blood leukocytes (Figueroa et al., 2001) or U937 cells (Capra et al., 2005). In human MCs, which can express both receptor subtypes, Mellor et al. (2003) observed that under conditions in which CysLT1R is blocked, IL-5 generation results only from stimulation with the selective CysLT2R agonist BAY U9773 and not with cysteinyl LTs, leading them to speculate that this could arise from stimulation of a CysLT1/CysLT2 heterodimer at a site inaccessible to interference from MK-571. Finally, Beller et al., (2004b) postulated, on the basis of the magnitude of the attenuation in IgE-dependent, MC-mediated passive cutaneous anaphylaxis in CysLT1R/CysLT2R-null mice, that the effect observed in wild-type littermates was mediated through CysLT1/CysLT2 heterodimers. These speculations have been now confirmed by assessment of the formation of a CysLT1/CysLT2 heterodimer at the nuclear envelope of human MCs. Negative regulation of CysLT1R-induced mitogenic signaling responses of MCs by CysLT2R demonstrates physiologically relevant functions for GPCR heterodimers on primary cells central to inflammation (Jiang et al., 2007). It is noteworthy that recently it has been demonstrated that CysLT1R formed heterodimers with its counter-receptor CysLT2R under basal conditions and that LTD4 triggers reduced dimerization of CysLTRs in intestinal epithelial cells (Parhamifar et al., 2010).
These observations are of potential interest, considering that GPCR oligomerization influences CysLTR pharmacology and function and offers new horizons for the study of important aspects of cysteinyl LT biology and possibly for the development of new drugs (Dalrymple et al., 2008; Kenakin, 2002; Prinster et al., 2005) (see section III.H).
H. Other Receptors Related to CysLT Receptors
Over the years, data were reported in the literature to suggest the existence of additional CysLTR subtypes in human tissues (Brink et al., 2003). This proposal was based on classic pharmacodynamic observations on agonist and antagonist potencies and as well as on receptor pharmacological profiles (Rovati et al., 1997; Bäck, 2002; Norel and Brink, 2004; Rovati and Capra, 2007; Austen et al., 2009).
In an intriguing report Mellor et al. (2001) suggested that in MCs both the CysLT1R and a yet-unidentified elusive receptor up-regulated by treatment with the proinflammatory cytokine IL-4 were responsive to both cysteinyl LTs and UDP. Pharmacological studies performed with classic CysLT1R or the dual CysLT1/CysLT2R antagonists excluded the possibility that this additional receptor was the CysLT2R subtype. Subsequently, the hypothesis that the hCysLT1R was responding to UDP was also excluded (Capra et al., 2005; Woszczek et al., 2010), whereas different CysLT1R antagonists have been demonstrated to inhibit P2Y receptor signaling (Mamedova et al., 2005; Woszczek et al., 2010).
Screening for orphan GPCRs at an intermediate phylogenetic position between P2Y and CysLT receptor families revealed that the heterologous expression of GPR17 in 1321N1 cells resulted in the generation of an apparently specific and concentration-dependent response to both cysteinyl LTs and extracellular nucleotides (Ciana et al., 2006). In line with previous expression data (Bläsius et al., 1998), both human and rat GPR17 was highly present in brain and in other organs typically undergoing ischemic damage, such as kidney and heart. These data are consistent with the demonstration that both the CysLT1 and P2Y12/13 receptor antagonists (montelukast and cangrelor, respectively) or in vivo GPR17 receptor knockdown protected against brain damage in the permanent middle cerebral artery occlusion model of ischemic damage in the rat (Ciana et al., 2006). However, GPR17 has been demonstrated to colocalize and dimerize with CysLT1R on the cell surface of human peripheral blood monocytes and was a negative regulator of the CysLT1R both in vitro (Maekawa et al., 2009) and in vivo (Maekawa et al., 2010), without inducing a specific response, at least when expressed in a recombinant system. A recent article confirmed that GPR17 is indeed activated by uracil nucleotides (Benned-Jensen and Rosenkilde, 2010), although, at least in isolation, it may not respond to cysteinyl LTs (G. E. Rovati, unpublished observations). In addition, in a different native system (differentiated PC12 cells) knock down of GPR17 by small interfering RNA abolished LTD4-induced effect on cell viability (Daniele et al., 2010). In this respect, it might be important to note that P2Y6 receptors also require an intact cysteinyl LT synthesis and signaling system to induce survival and activation of MCs (Jiang et al., 2009). Thus, CysLT1R and P2Y6R, which are coexpressed on many cell types of innate immunity, seem to reciprocally amplify one another's function through endogenous ligands. In conclusion, whether GPR17 is really a dualistic receptor (a single protomer responding to two different classes of ligands) or may form heterodimers with a P2Y receptor is a matter certainly deserving further investigation.
2. Leukotriene E4 Receptor(s).
Although LTE4 is only a weak agonist for the defined CysLT1R and CysLT2R, this ligand is the most stable and abundant cysteinyl LT in vivo. However, the literature on LTE4-induced effects is conflicting, and there seems to be tissue- and model-related differences that remain to be defined. In addition, data from studies in the presence and absence of conversion inhibitors also produced different results (Brink et al., 2003). Early reports indicated greater potency for LTE4 compared with that for the other cysteinyl LTs in constricting guinea pig trachea in vitro (Lewis et al., 1980; Lee et al., 1984) and comparable activity in eliciting a cutaneous wheal and flare response in humans (Soter et al., 1983). Furthermore, patients with AIA responded with enhanced bronchoconstriction to inhalation of LTE4 relative to their responses to LTC4 or histamine (Christie et al., 1993), and inhaled LTE4 caused more tissue and airway eosinophilia than LTD4 (Gauvreau et al., 2001). In contrast, LTE4 failed to activate a CysLTR in human pulmonary vessels (Walch et al., 2002). In 2008, the first report that LTE4 mediated prostaglandin D2 production, which was sensitive to the CysLT1R antagonist MK-571 but resistant to knock down of this receptor, was published (Paruchuri et al., 2008). Furthermore, this effect, induced through COX-2 induction and ERK, p90RSK, and cAMP-regulated-binding protein phosphorylation, was not mediated by LTD4 but was sensitive to PPAR-γ knockdown, suggesting the presence of a LTE4-selective receptor in human MCs (Paruchuri et al., 2008). Almost simultaneously it was demonstrated that intradermal injection of LTE4 into the ear of mice deficient in both CysLT1R/CysLT2R elicited a vascular leak that exceeded the response to intradermal injection of LTC4 or LTD4 and that this response was inhibited by pretreatment of the mice with PTX or a Rho kinase inhibitor (Maekawa et al., 2008). This putative mouse LTE4-specific receptor, however, was not MK-571-sensitive, adding complexity to LTE4 puzzle (Austen et al., 2009). One last piece of the puzzle has been revealed by the same group on the basis of a study in mice lacking both CysLT1R/CysLT2R: the ADP-reactive P2Y12 receptor was required for LTE4-mediated pulmonary inflammation. Platelet P2Y12 receptor expression permitted LTE4-induced activation of ERK in CHO cells and allowed chemokine and prostaglandin D2 production by LAD2 cells, a human MC line. Of interest, competition experiments between radiolabeled ADP and unlabeled LTE4 in this last system seemed to suggest that P2Y12 must complex with another receptor to recognize LTE4 (Paruchuri et al., 2009). Thus, in the light of the progress on heterodimer pharmacology of GPCRs and their possible physiological significance (Prinster et al., 2005; Gurevich and Gurevich, 2008; Milligan, 2009), the possibility that some of these additional CysLTR subtypes might be the result of the formation of heterodimers with a different pharmacological profile, rather than representing new distinct proteins, should always considered (see section III.H). In all previous cases, these data add complexity to the already established “cross-talk” between the purinergic and the LT receptor systems (Capra et al., 2005; Jiang et al., 2009), suggesting these as possible additional means by which these two receptor systems interact with each other.
I. New Potential Therapeutic Applications of Leukotriene Receptor Antagonists
1. Allergic Rhinitis.
Many of the cells involved in the pathophysiology of AR produce and release cysteinyl LTs, and a connection between the upper airway disease, rhinitis, and lower airway disease, asthma, has been delineated with the aid of important studies indicating that they essentially represent components of a single inflammatory airway disease (Blaiss, 2005; Thomas, 2006). Molecular and clinical studies also substantiated a role for cysteinyl LTs and their receptors in AR (Peters-Golden and Henderson, 2005; Peters-Golden et al., 2006). Furthermore, beyond the existing connection between AR and asthma, AR and chronic rhinosinusitis seem to be related both epidemiologically and symptomatically. In fact, some studies have demonstrated possible involvement of the cysteinyl LT system in chronic rhinosinusitis also (Arango et al., 2002; Higashi et al., 2004), and both CysLT1 and CysLT2 receptors were expressed in various inflammatory cells of patients with active seasonal AR (Figueroa et al., 2003) or rhinosinusitis (Sousa et al., 2002) (see section III.A).
Given the shared pathophysiologies and the frequent coexistence, a common therapeutic approach would seem logical (Greenberger, 2003; Stelmach and Cukier, 2006). Indeed, evidence has begun to accumulate in support of the utility of LTRAs in AR (Wilson et al., 2004; Rodrigo and Yañez, 2006) or rhinosinusitis (Grayson and Korenblat, 2007; Scadding et al., 2008).
Montelukast, which significantly reduced daytime nasal, eye and throat symptoms, nighttime symptoms, and symptoms affecting quality of life (QL) in adult patients with seasonal AR (Grainger and Drake-Lee, 2006; Nayak and Langdon, 2007; Bousquet et al., 2009), has also been demonstrated to be safe in children (Bisgaard et al., 2009). Similar results have also been obtained with zafirlukast (Piatti et al., 2003) and pranlukast (Ueda et al., 2003; Okubo and Baba, 2008).
Other studies indicate that LTRAs can alleviate the signs and symptoms of AR when used either as monotherapy (van Adelsberg et al., 2003; Lee et al., 2004b; Ho and Tan, 2007; Wahn and Dass, 2008; Diamant et al., 2009; Cingi and Ozlugedik, 2010) or in combination with an antihistamine in adults (Meltzer et al., 2000; Nayak et al., 2002; Ciprandi et al., 2004; Kurowski et al., 2004; Ciebiada et al., 2006; Lagos and Marshall, 2007; Li et al., 2009), as well as in children (Phan et al., 2009). Results from meta-analyses indicate that antihistamines and LTRAs are equally effective in improving symptoms of AR and QL, but that both drugs are less effective than intranasal corticosteroids (Pullerits et al., 2002; Saengpanich et al., 2003; Di Lorenzo et al., 2004b; Wilson et al., 2004; Rodrigo and Yañez, 2006; Van Hoecke et al., 2007). However, in a retrospective observational study, addition of montelukast to current corticosteroid therapy improved long-term asthma control and resulted in substantial reductions in asthma-related resource use by patients with mild or moderate persistent asthma and concomitant seasonal AR (Borderias et al., 2007), confirming previous results on the efficacy of montelukast in patients with AR and asthma (Nayak, 2004; Philip et al., 2004; Price et al., 2006; Virchow and Bachert, 2006; Storms, 2007). Of interest, systemic medication such as montelukast, as expected, provided better relief for symptoms distant from the nasal cavity, whereas the antihistamine reduced rhinorrhea more than either montelukast or budesonide (Sardana et al., 2010). Both second-generation antihistamines and montelukast have been found to be relatively safe and effective in children also (Phan et al., 2009). Montelukast received approval from the U.S. Food and Drug Administration in January 2003 for the treatment of seasonal AR and in 2005 for the treatment of perennial AR.
2. Aspirin/Nonsteroidal Anti-Inflammatory Drug Intolerance.
A variable percentage (2–20%) of adults with asthma exhibit an AIA phenotype (Jawien, 2002), which is the second most frequent untoward allergic reaction to drugs (de Weck et al., 2006). AIA consists of the precipitation of asthma and rhinitis attacks in the airways or urticaria and angioedema in the skin in response to aspirin and other nonsteroidal anti-inflammatory drugs (NSAIDs) that inhibit COX-1 (Szczeklik and Stevenson, 2003); selective inhibition of COX-2 does not provoke such responses (Stevenson and Simon, 2001; Gyllfors et al., 2003; Stevenson and Szczeklik, 2006). Within 3 h of ingestion of aspirin or NSAIDs, individuals with AIA develop bronchoconstriction, often accompanied by rhinorrhea, conjunctival irritation, and scarlet flush (Obase et al., 2005). Compared with normal individuals or subjects with aspirin-tolerant asthma, patients with AIA have increased baseline levels of cysteinyl LTs in urine (Christie et al., 1991; Higashi et al., 2004; Micheletto et al., 2006) exhaled air (Antczak et al., 2002), and in saliva and induced sputum (Gaber et al., 2008), levels that are further enhanced by aspirin challenge (Ortolani et al., 1987; Sladek and Szczeklik, 1993; Szczeklik et al., 1996) and that might be derived, at least in part, from MCs (Sladek and Szczeklik, 1993; O'Sullivan et al., 1996; Mita et al., 2001a). Although a definite explanation for this syndrome is not yet available, a large body of evidence supports the possibility that patients with AIA are particularly dependent on the bronchoprotective and anti-inflammatory properties of prostaglandin E2 (Stevenson and Szczeklik, 2006) and that inhibition of COX-1 in these patients triggers MC activation and release of LTs as well as histamine and tryptase (Picado et al., 1992). In addition to the already discussed high levels of cysteinyl LTs and the localization of CysLT1R in nasal mucosa and leukocytes of patients with AIA (Sousa et al., 2002), the clinical relevance of the cysteinyl LTs in AIA has been supported by studies demonstrating that patients with AIA respond favorably to clinical treatment with LT modifiers (Dahlén et al., 1998, 2002; Israel, 2000; Mastalerz et al., 2002a,b Obase et al., 2002; Lee et al., 2004a; Micheletto et al., 2004; Park et al., 2010).
The possible association between CysLTR1 and CysLTR2 gene polymorphisms and AIA has already been discussed (see section III.F). CysLTR1 gene polymorphism, which is often associated with increased CysLT1R expression, has been found to be associated with AIA in males (Kim et al., 2006b, 2007a). Considering that CysLTR2 gene polymorphisms, which, on the contrary, seem to affect CysLT2R mRNA transcription and stability (Shin et al., 2009), have also been found to be associated with AIA, particularly when combined with the LTC4S polymorphisms −444A>C (Park et al., 2005), one might envisage a protective role for CysLT2R in AIA. These data seem to support the finding that CysLT2R, interacting with CysLT1R, down-modulates cysteinyl LT-dependent responses in MCs (Jiang et al., 2007).
3. Other Atopic Diseases.
Beyond asthma and rhinitis, there are other atopic diseases in which cysteinyl LTs are considered to play a role and that normally affect adults and children with relevant human and economic burden, such as atopic dermatitis (AD) and urticaria. Although the role of cysteinyl LTs in the inflammation of AD is unclear, several reports in the literature, but not all (Veien et al., 2005), detail improvements in patients (adult and children) with moderate to severe AD with the use of LTRAs at the doses generally recommended for asthma treatment (Capra et al., 2006; Leonardi et al., 2007; Broshtilova and Gantcheva, 2010). In support, there is also the evidence that elevated levels of LTE4 have been found in the urine of patients with AD (Adamek-Guzik et al., 2002; Øymar and Aksnes, 2005). Chronic urticaria (CU) may manifest as an idiopathic reaction or as a reaction to a known cause, such as cold, pressure, food additives, or NSAIDs; however, this does not occur with selective COX-2 inhibitors (Zembowicz et al., 2003). Of interest, the prevalence of aspirin sensitivity among patients with CU is estimated to be between 20 and 30% (Grattan, 2003), and recently the LTC4S −444A>C polymorphism has been suggested to be a risk factor for aspirin-induced urticaria (Sánchez-Borges et al., 2009). Thus, these patients may be more likely to respond to LTRA therapy (Asero, 2000; Pacor et al., 2001). Indeed, although some authors found LTRAs to be ineffective in CU (Reimers et al., 2002; Di Lorenzo et al., 2004a) or even to induce a paradoxical exacerbation (Minciullo et al., 2004; Tedeschi, 2009), montelukast has been demonstrated to be effective in patients for whom treatment with histamine H1-receptor antagonists had proved ineffective (Erbagci, 2002; Bagenstose et al., 2004). These data, therefore, suggest that LTRAs may have benefit for patients with conditions not sufficiently controlled with histamine H1-receptor antagonists (Sanada et al., 2005; Wan, 2009).
4. Cardiovascular Diseases.
Biomedical research in the LT field has been markedly polarized to asthma and allergic disorders and has largely overlooked other diseases that are also based on the existence of an inflammatory process with increased vascular permeability and edema, such as CVD (Capra et al., 2007). Although blocking the inflammatory component of any human disease is a long-standing and established concept, the use of LT modifiers (LTRAs and LT synthesis inhibitors) in treating the inflammatory component of CVDs has, surprisingly, been seriously contemplated only in the past few years (Funk, 2005; Peters-Golden and Henderson, 2007; Bäck, 2008b). However, a growing body of evidence suggests a major role for the LTs generated by the 5-LO pathway in the pathogenesis and progression of CVDs (Lötzer et al., 2005; Bäck, 2009a), particularly atherosclerosis, myocardial infarct, stroke, aortic aneurysms, and intimal hyperplasia, as LT signaling is emerging as a crucial component in vascular inflammation (Bäck et al., 2007). Cysteinyl LTs have a unique pharmacological profile, characterized by potent constrictor actions on the microvasculature and thus regulation of blood pressure, enhanced permeability of the postcapillary venules, reduced coronary blood flow, reduced myocardial contractility, and cardiac output without affecting the heart rate (for review, see Brink et al., 2003 and Capra et al., 2007).
Both CysLTR subtypes are expressed in diseased human arteries (Allen et al., 1998; Spanbroek et al., 2003). However, the role of CysLT1R in CVDs is rather controversial. Findings suggesting a role for cysteinyl LTs in the extension of ischemic damage and in cardiac dysfunction during reperfusion (Toki et al., 1988; Hock et al., 1992) are evenly balanced by results suggesting that these autacoids have no (Hahn et al., 1992) or little (Ito et al., 1989) contribution to the progression of myocardial ischemia-reperfusion injury. On the other hand, the LTRAs investigated in these studies belong to the first generation of molecules and may have not displayed enough potency to compete with the endogenous ligands, which are likely to be present with very high local bioavailability at sites where inflammatory cell accumulation and myocardial injury take place. In fact, the newer potent and commercially available CysLT1R antagonist, montelukast, has been found to inhibit atherosclerotic lesion size and intimal hyperplasia (Kaetsu et al., 2007; Jawien et al., 2008) and to reduce vascular reactive oxygen species production, to significantly improve EC function, and to ameliorate atherosclerotic plaque generation in a mouse model in vivo (Mueller et al., 2008). Furthermore, in a rabbit carotid balloon injury model, it significantly reduced the neointima, decreased macrophage content, increased SMC content, and inhibited expression of MCP-1 without having influence on plasma lipids, indicating antiatherogenic effects unrelated to plasma lipid modulation (Ge et al., 2009). A randomized controlled trial of placebo versus either montelukast or theophylline in asthmatic individuals reported significantly lower C-reactive protein in montelukast-treated patients (Allayee et al., 2007).
As regards a role for the CysLT2R in CVD, the early, endothelium-targeted overexpression of the human CysLT2R in mice aggravates myocardial ischemia-reperfusion injury (Hui et al., 2004), leading to accelerated left ventricular remodeling and impaired cardiac performance (Jiang et al., 2008). Despite expression data suggesting that CysLT2R is the main isoform of the receptor for cysteinyl LTs in the normal brain (see section III.A), selective CysLT1R antagonists might have a protective effect in focal cerebral ischemia (Yu et al., 2005a,b; Ciana et al., 2006; Qian et al., 2006), possibly decreasing blood-brain barrier permeability (Biber et al., 2009). However, the dual CysLT1/CysLT2 receptor antagonist BAY U9773 has been proved to be more potent and effective than a selective CysLT1R antagonist in preventing the brain permeability alteration induced by neutrophil activation (Di Gennaro et al., 2004).
The first report indicated that in vivo inhibition of GPR17 by either montelukast or antisense technology dramatically reduced ischemic damage in a rat model of focal cerebral ischemia (Ciana et al., 2006). Subsequently, it was suggested that GPR17 might orchestrate oligodendrocyte maturation after brain or spinal cord injury at a later stage, indicating a role in fostering brain repair (Lecca et al., 2008; Ceruti et al., 2009, 2011). The latter findings were supported by in vitro experiments showing that GPR17 activation stimulated the maturation of preoligodendrocytes (Lecca et al., 2008; Fumagalli et al., 2011). In contrast, GPR17 transgenic mice exhibited blocked differentiation of neural progenitor cells into oligodendrocytes and also inhibited terminal differentiation of primary oligodendrocyte precursor cells, supporting GPR17 as a negative regulator of oligodendrocyte differentiation and myelination (Chen et al., 2009b). Fumagalli et al. (2011) recently suggested that one possible explanation for the apparent differences between those studies may be that the overexpression of GPR17 occurred with the CNPase promoter in the latter study (Chen et al., 2009b), inducing GPR17 expression at a rather advanced maturation stage. Nevertheless, GPR17-deficient mice exhibit early onset of central nervous system myelination (Chen et al., 2009b), and the exact role of GPR17 signaling in oligodendrocyte maturation and function remains to be determined. Thus, whether GPR17 or indeed cysteinyl LTs are friends or foes in ischemic brain injury still needs to be established (Bäck, 2008a).
5. CysLT in Other Diseases.
A number of observations document that 5-LO and its products may be relevant in different diseases either because of an increase in cysteinyl LT production/CysLTR expression or because the use of LTRAs has proved to be efficacious. For example, cysteinyl LTs have been found in the sputum and urine of patients with cystic fibrosis (CF) (Sampson et al., 1990; Spencer et al., 1992), a disease in which the inflammatory process contributes to progressive lung tissue damage, and their concentrations increase during disease exacerbations together with oxidative stress (Reid et al., 2007). Few studies have been conducted to test the hypothesis that LTRAs have the potential to ameliorate CF lung disease. A pilot study indicated that zafirlukast may benefit adult patients with CF (Conway et al., 2003), whereas montelukast treatment significantly decreased serum (Schmitt-Grohé et al., 2002) and sputum levels of eosinophil cationic protein, IL-8, myeloperoxidase, and increased serum and sputum levels of IL-10 in children (Stelmach et al., 2005). These observations suggest that LTRAs may have measurable anti-inflammatory properties and potential to ameliorate CF lung disease, particularly with long-term use (Schmitt-Grohé et al., 2007). Chronic inflammation also plays a crucial role also in COPD. However, pulmonary inflammation in COPD is, to some extent, different from that in asthma, because other inflammatory cells (i.e., neutrophils, macrophages, and CD8+ T lymphocytes) are implicated (Gan et al., 2004). In addition, the profile of exhaled eicosanoids may be different from that previously reported in asthma: exhaled PGE2 is selectively increased in COPD, whereas LTE4 is increased in asthma (Montuschi et al., 2003), but not in COPD (Gaki et al., 2007). There are only a few studies of the clinical effects of LTRAs in COPD. For example, lung function improvement has been observed with zafirlukast in smokers with COPD (Cazzola et al., 2000; Nannini and Flores, 2003) Furthermore, a retrospective study of patients with moderate to severe COPD, reported that montelukast use improved patient's functional scores (shortness of breath, sputum production, wheezing, and nocturnal symptoms), use of drugs (oral and inhaled corticosteroids, inhaled bronchodilators, and supplemental oxygen), number of visits to the emergency department as well as the number and duration of hospitalizations for acute exacerbations (Rubinstein et al., 2004), and QL (Celik et al., 2005). However, further clinical studies are required to judge the real impact of LTRAs in COPD therapy or at least in a subset of patients with COPD (Usery et al., 2008). Given the significant neutrophilic inflammation in COPD, it is likely that LTB4 has a role in the disease (see also section II.F), and thus BLT antagonists and/or 5-LO/FLAP inhibitors may have greater potential as additional therapy in COPD.
Cysteinyl LTs have been implicated in different liver diseases, such as cholestasis, hepatic inflammation, portal hypertension, and hepatorenal syndrome (Keppler et al., 1988; Farzaneh-Far and Moore, 2003), and their excretion is enhanced in patients with liver cirrhosis and hepatorenal syndrome (Huber et al., 1989). Furthermore, cysteinyl LTs are generated in the liver during the reperfusion period and may contribute to the development of hepatic edema and exert cytotoxicity (Takamatsu et al., 2004). Although inhibition of FLAP abrogated experimental liver injury (Titos et al., 2005) and decreased hepatic inflammation and fibrosis in mice (Horrillo et al., 2007), montelukast was also found to be effective in prevention of liver and intestine injury by reducing apoptosis and oxidative stress in a hepatic ischemia-reperfusion injury model (Daglar et al., 2009) and to improve hepatic fibrosis in cholestatic rats (El-Swefy and Hassanen, 2009).
Finally, the presence of receptors that respond to LTD4 in human detrusor myocytes (Bouchelouche et al., 2001a) and evidence for increased urinary LTE4 in patients with interstitial cystitis and detrusor mastocytosis (Bouchelouche et al., 2001b) may suggest a role for cysteinyl LTs as proinflammatory mediators in this disease. Montelukast treatment resulted in significant improvement in urinary frequency and pain (Bouchelouche et al., 2001c), as well in remission of an eosinophilic cystitis (Sterrett et al., 2006), suggesting a role for LTRAs in managing interstitial cystitis.
LTs are elevated in a number of cancers and some preclinical cancer models have shown efficacy of early LT synthesis inhibitors (Rioux and Castonguay, 1998; Steele et al., 1999). In addition, some reports have advanced the concept that GPCRs are mediators of cell growth by demonstrating their potential to activate MAPKs, particularly ERK1/2 (Marinissen and Gutkind, 2001). Several mitogenic pathways might link GPCRs to the nucleus, some of them requiring the capacity of GPCRs to transactivate a growth factor receptor, such as the EGFR (Daub et al., 1996). Indeed, as already discussed in section III.C.1, CysLT1R activation is able to induce MAPK phosphorylation as well as growth factor receptor transactivation, and, thus, to stimulate proliferation (Wang and Dubois, 2010b). Furthermore, consistent data linking CysLT1Rs directly to cancer and block of apoptosis are available for epithelial colon cancer cells (Ohd et al., 2003; Nielsen et al., 2005a; Parhamifar et al., 2005; Magnusson et al., 2007, 2010), whereas little information is available on CysLT1R expression in brain tumors, such as astrocytoma, ganglioglioma, and metastatic adenocarcinoma (Zhang et al., 2004), in prostate cancer (Matsuyama et al., 2007) or in breast cancer (Magnusson et al., 2011). Taken together, these findings encourage the study of pharmacological treatment of selected forms of cancer with the LTRAs already marketed for asthma.
IV. Nomenclature for Leukotriene Receptors
Based on the above-cited studies, the IUPHAR Nomenclature Committee for Leukotriene Receptors concludes that the previously established nomenclature for leukotriene receptors (Brink et al., 2003) remains pertinent. Nevertheless, the present nomenclature may present some limitations, which need to be taken into consideration.
Although 12-HHT is an endogenous ligand for BLT2R with higher affinity than LTB4, the classification of this receptor as the low-affinity receptor for LTB4 can be motivated by the important amino acid identity with the high-affinity BLT1R and the close genomic localization of the LTB4R and LTB4R2 genes. However, considering that 12-HHT is produced in platelets more abundantly than LTB4, the nomenclature of BLT2R may be reconsidered in the future.
Second, the interaction between purinergic and leukotriene signaling represents an interesting cross-talk in which leukotrienes may activate purinergic receptors and nucleotides may signal through CysLT receptors. Although there is a phylogenetic relationship between purinergic receptors and leukotriene receptors, the exact interactions between these receptor classes remain to be established. Likewise, classifying the recently deorphanized GPR17 or the P2Y12 receptor into the leukotriene receptor class seems to be premature at this time. Finally, although functional studies and in vivo experiments indicate that not all leukotriene-induced responses fit the current receptor classification, the molecular structure of the putative leukotriene receptor(s) is presently unknown.
In conclusion, there is no doubt that leukotriene signaling is complex and may involve several receptor classes. The majority of the known leukotriene-induced effects can be assigned to the currently identified receptor subtypes (Table 1). It is hoped that future studies will extend our current understanding of putative additional leukotriene receptors.
We acknowledge Dr. Charles Brink for his critical reading and helpful advice on the content of this review.
Wrote or contributed to the writing of the manuscript: Bäck, Dahlén, Drazen, Evans, Serhan, Shimizu, Yokomizo, and Rovati.
This article is available online at http://pharmrev.aspetjournals.org.
- amino acids
- atopic dermatitis
- aspirin-induced asthma
- acute myeloid leukemia 1
- activator protein-1
- allergic rhinitis
- aspirin-tolerant asthma
- broncho-alveolar lavage
- BAY U9773
- 6(R)-(4-carboxyphenylthio)-5(S)-hydroxy-7(E),9(E),11(Z),14(Z)-eicosatetraenoic acid
- BLT receptor(s)
- BLT1 receptor(s)
- BLT2 receptor(s)
- bone marrow-derived dendritic cell
- bone marrow-derived mast cell
- cystic fibrosis
- Chinese hamster ovary
- chronic obstructive pulmonary disease
- 1-[3-(4-phenyl-benzyl)-4-hydroxy-chroman-7-yl] cyclopentane carboxylic acid
- chronic urticaria
- cardiovascular disease
- cysteinyl leukotriene receptor(s)
- CysLT1 receptor(s)
- CysLT2 receptor(s)
- dendritic cell
- dextran sodium sulfate
- allergic encephalomyelitis
- endothelial cell
- eosinophil cationic protein
- epidermal growth factor receptor
- extracellular signal-regulated kinases
- Fc receptor
- 5-lipoxygenase activating protein
- G protein-coupled receptor
- GPCR kinases
- human embryonic kidney
- hydroxyeicosatetraenoic acid
- hydroxyheptadecatrienoic acid
- hydroperoxyeicosatetraenoic acid
- human umbilical cord endothelial cell
- inflammatory bowel disease
- insulin growth factor
- International Union of Basic and Clinical Pharmacology
- c-Jun NH2-terminal kinase
- LTC4 synthase
- leukotriene receptor antagonist
- 2-[2-propyl-3-[3-[2-ethyl-4-(4-fluorophenyl)-5-hydroxyphenoxy]-propoxy]phenoxyl]benzoic acid
- mitogen-activated protein kinase
- mast cell
- monocyte chemoattractant protein
- macrophage inflammatory protein
- 3-[[3-[2-(7-chloroquinolin-2-yl)vinyl]phenyl]-(2-dimethylcarbamoylethylsulfanyl)methylsulfanyl] propionic acid
- matrix metalloproteinases
- mononuclear cell
- MRS 2395
- 2,2-dimethyl-propionic acid 3-(2-chloro-6-methylaminopurin-9-yl)-2-(2,2-dimethyl-propionyloxymethyl)-propyl ester
- multiple sclerosis
- National Center for Biotechnology Information
- nuclear factor κB
- nitric oxide
- nonsteroidal anti-inflammatory drug
- pyrimidinergic nucleotide
- platelet-activating factor
- prostaglandin E2
- phosphatidylinositol 3-kinase
- protein kinase
- peroxisome proliferator-activated receptor
- pertussis toxin
- quality of life
- regulated on activation normal T cell expressed and secreted
- reverse transcriptase-polymerase chain reaction
- SKF 104353
- 2(S)-hydroxyl-3(R)-carboxyethylthio)-3-[2-(8-phenyloctyl) phenyl] propanoic acid
- smooth muscle cell
- single nucleotide polymorphism
- signal transducer and activator of transcription
- T helper
- transmembrane domain
- tumor necrosis factor
- transient receptor potential
- vascular endothelial growth factor
- wild type.
- © 2011 by The American Society for Pharmacology and Experimental Therapeutics