Leukotriene pathway genetics and pharmacogenetics in allergy

Authors


Dr Nathalie P Duroudier
Division of Therapeutics and Molecular Medicine
D-Floor, South Block
Queen’s Medical Centre
Nottingham NG7 2UH
UK

Abstract

Leukotrienes (LT) are biologically active lipid mediators known to be involved in allergic inflammation. Leukotrienes have been shown to mediate diverse features of allergic conditions including inflammatory cell chemotaxis/activation and smooth muscle contraction. Cysteinyl leukotrienes (LTC4, LTD4 and, LTE4) and the dihydroxy leukotriene LTB4 are generated by a series of enzymes/proteins constituting the LT synthetic pathway or 5-lipoxygenase (5-LO) pathway. Their function is mediated by interacting with multiple receptors. Leukotriene receptor antagonists (LTRA) and LT synthesis inhibitors (LTSI) have shown clinical efficacy in asthma and more recently in allergic rhinitis. Despite growing knowledge of leukotriene biology, the molecular regulation of these inflammatory mediators remains to be fully understood. Genes encoding enzymes of the 5-LO pathway (i.e. ALOX5, LTC4S and LTA4H) and encoding for LT receptors (CYSLTR1/2 and LTB4R1/2) provide excellent candidates for disease susceptibility and severity; however, their role remains unclear. Preliminary data also suggest that 5-LO pathway/receptor gene polymorphism can predict patient responses to LTSI and LTRA; however, the exact mechanisms require elucidation. The aim of this review was to summarize the recent advances in the knowledge of these important mediators, focusing on genetic and pharmacogenetic aspects in the context of allergic phenotypes.

Leukotrienes (LT) are a family of potent eicosanoid lipid mediators with central importance in disease processes such as inflammation and proliferation most notably observed in allergic conditions like asthma (1). There are two general classes of LT according to the presence of a cysteine residue in their amino acid chain: the cysteinyl leukotrienes (CysLTs) including LTC4, LTD4 and LTE4 and the dihydroxyleukotriene LTB4 (2).

The 5-lipoxygenase pathway

Leukotrienes are produced in a multi-step enzyme pathway called the 5-lipoxygenase (5-LO) pathway, which is active in leucocytes such as neutrophils, eosinophils, mast cells and monocytes (Fig. 1). Arachidonic acid is the precursor for LT synthesis and is hydrolysed from the plasma membrane by cytosolic phospholipase A2 (and other isoforms) (3, 4) in a calcium-dependent process (5). Upon activation, the rate-limiting enzyme 5-LO oxygenates first free arachidonic acid into the unstable intermediate 5-hydroperoxyeicosatetraenoic acid (5-HPETE) which is then either hydrolysed to 5-hydroxyeicosatetraenoic acid (5-HETE) or transformed into the unstable epoxide leukotriene A4 (LTA4) by forming a conjugated triene system through dehydration (6). 5-LO translocates from either the nucleus (in macrophages) or cyotosol (in neutrophils) to the nuclear envelope in response to cell activation (7, 8). This movement occurs as 5-LO is dependent on a 5-LO activating protein (FLAP) for its function. FLAP remains associated with the nuclear membrane (9) and acts as a ‘transfer protein’ presenting arachidonic acid to 5-LO, a system which allows the favourable conversion of 5-HPETE to LTA4 compared to 5-HETE (10). Both 5-LO and FLAP are expressed in cells of myeloid lineage (11) restricting the pathway to these cell types. LTA4 can be further metabolised to the cysteinyl leukotrienes (CysLTs) or LTB4. The specific glutathione S-transferase leukotriene C4 synthase (LTC4S) conjugates LTA4 to form LTC4 which can then be rapidly converted to LTD4 by a gamma-glutamyl transpeptidase and to LTE4 by a dipeptidase once exported out the cell by membrane transport proteins such as multi-drug related protein 1 (MRP1) (12–15). A specific zinc metallohydrolase, LTA4 hydrolase (LTA4H) is responsible for the conversion of LTA4 to LTB4 (16). CysLTs and LTB4 act on different G-protein coupled receptors (GPCRs). Each bind to at least two different receptors: CysLTs to CYSLTR1 and CYSLTR2 and LTB4 to LTB4R1 and LTB4R2 (or BLT1 and BLT2) (17).

Figure 1.

 The 5-lipoxygenase pathway. See text for description. 5-LO, 5-lipoxygenase; aa, arachidonic acid; CYSLTR, cysteinyl leukotriene receptor; DP, dipeptidase; FLAP, 5-lipoxygenase activating protein; γ-GT, γ-glutamyl transpeptidase; LTBR, leukotriene B4 receptor; LTA4, LTB4, LTC4, LTD4, LTE4, leukotriene A4, B4, C4, D4, E4; LTA4H, LTA4 hydrolase; LTC4S, LTC4 synthase; LTRA, leukotriene receptor antagonists; LTSI, leukotriene synthesis inhibitor; MRP1, multi-drug resistant protein 1; PLA2, phospholipase A2; PPARα, peroxisome proliferator activatived receptor alpha.

Regulation of LT synthesis

Leukotriene synthesis is restricted to cells of myeloid lineage as this is where 5-LO and FLAP are expressed. Different cells also express LTC4S or LTA4H affecting which LT is synthesised. LTC4S is expressed in eosinophils and mast cells whereas neutrophils produce LTB4 as their major LT. There are also cells that produce both (18). LTA4H is expressed ubiquitously in most tissues. In addition, the interaction between different cell types with cooperation of their enzymes can allow platelets (19) and vascular endothelial cells (20) or erythrocytes (21) to produce LTC4 or LTB4, respectively, from LTA4 transferred from neutrophils (transcellular metabolism). Subcellular localisation of the enzymes/proteins is important. 5-LO, FLAP and LTC4S all co-localise to the nuclear envelope in a calcium-dependent process upon cell activation, bringing 5-LO into the proximity of its substrate and LTC4S to allow its preferential action (9, 22). 5-LO is concentrated in the cytosol of resting neutrophils, but is located in the nucleus of alveolar macrophages and mast cells (23). This localisation is possible because of three nuclear localisation signals (24, 25). The mechanisms involved here are still not fully understood, but there is evidence for cells with nuclear 5-LO to produce more LTB4 than the same cells with cytoplasmic 5-LO (26). Phosphorylation by protein kinase (mitogen-activated protein) A on Ser523 can decrease 5-LO activity, whereas phosphorylation on Ser271 by MAP kinase (mitogen-activated protein) activated protein 2 and on Ser663 by extracellular signal-regulated kinase can increase it (27–29).

Leukotrienes are lipid mediators derived from arachidonic acid by a sequence of reactions catalised by the enzymes of the 5-LO pathway. The pathway is activated by many different stimuli including biological messengers such as cytokines and immune complexes and environmental stimuli like antigens and microbes. There are multiple levels of regulation including at the genetic and biological level.

LTs and allergic inflammation

Cysteinyl leukotrienes have been shown to be involved in most of the key features of asthma including airway smooth muscle constriction (30, 31), vasodilation and increased microvascular permeability of vascular smooth muscle (32), leading to plasma exudation hence tissue oedema, mucus secretion (33) and inhibition of human respiratory cilia activity (34), airway remodelling (35), bronchial hyperresponsiveness (BHR) (17) and increased survival of inflammatory cells (36, 37). CysLT synthesis has been directly linked to asthma pathogenesis as there is extensive evidence that CysLT production is increased in asthma subjects, particularly during an exacerbation or upon challenge (1).

CysLTs are also key mediators in allergic rhinitis (AR) (38). In addition to their inflammatory effects similar to those seen in asthma, they modulate nasal allergic inflammation and clinical symptoms (especially sneezing and rhinorrhea) via activation of CYSLTR1 in nasal mucosa, in particular in vascular endothelial cells and interstitial cells (eosinophils, mast cells, macrophages and neutrophils) (39). In addition, CysLTs have been shown to participate in the process of allergic sensitisation by enhancing dendritic cell-stimulated antigen presentation (40). There is also evidence of enhanced CysLT production in the pathogenesis of atopic dermatitis (AD). Enhanced spontaneous and stimulated releasability of LTC4 from basophils and eosinophils isolated from AD patients is increased compared with healthy controls and increased production of LTs has been reported in the skin of atopic patients after allergen-specific challenge (41).

LTB4 has been proposed to play a role in a variety of acute and chronic inflammatory diseases including AD (42), asthma (1) and chronic obstructive pulmonary disease (COPD) (43). LTB4 is one of the most powerful chemotactic agents known to date (44) and its role in not only the recruitment but also activation and survival of neutrophils to the site of tissue injury (45) suggests that the mediator may amplify proinflammatory circuits in vivo. Indeed, LTB4 is overproduced in the airways of both asthma and COPD patients with levels correlating to the severity of asthma observed (46) and this LTB4 overproduction and subsequent interaction with its receptors causes the migration of inflammatory cells (mast cells, lymphocytes and eosinophils) into the airways (44).

It is important to note that in addition to a role in allergic inflammation, prominent expression of LT in the brain, adrenals, heart and vascular endothelium suggests additional unrecognised functions of those mediators, as shown by their involvement in brain injury (47), chronic cardiovascular diseases (48), cardiac anaphylaxis and ischemic myocardial infarction (MI) (48). Interestingly, a contribution to host defence against infection by augmenting the ability of macrophages and neutrophils to phagocytose microorganisms has also been suggested.

There is extensive evidence supporting a key role for LTs in allergen-induced inflammatory diseases such as asthma and AR which has been further confirmed by the clinical efficacy of strategies targeting LT production and activity (Table 1).

Table 1.   Leukotriene modifier therapy in human disease
Modifier(s)TargetUsed in clinical practice forPotential therapeutic effects on (in clinical trials)
  1. 5-LO, 5-lipoxygenase; AD, atopic dermatitis; AR, allergic rhinitis; COPD, chronic obstructive pulmonary disease; CysLT, cysteinyl leukotriene; FLAP, 5-lipoxygenase activating protein; LTB4, leukotriene B4; LTB4R, leukotriene B4 receptor; LTRA, leukotriene receptor antagonists; MI, myocardial infarction.

FLAP inhibitors: DG-031, BAYx1005FLAP MI, COPD, chronic bronchitis
5-LO inhibitor: Zileuton5-LOAsthmaCOPD
LTB4R antagonists: LY293111LTB4 receptorsAsthma, COPDPsoriasis, COPD, atherosclerosis, cancer
LTRAs: Montelukast, Pranlukast, ZafirlukastCysLT receptorsAsthma, ARCOPD, interstitial lung disease, chronic urticaria, AD, allergic fungal disease, nasal polyposis, paranasal sinus disease, migraine, respiratory syncytial virus postbronchiolitis, systemic mastocytosis, cystic fibrosis, pancreatitis, vulvovaginal candidiasis, cancer, atherosclerosis, eosinophils cystitis, otitis media, capsular contracture, eosinophilic gastrointestinal disorders

LT modifier therapy

Asthma

Three selective CYSLTR1 antagonists (LT receptor antagonists, LTRAs), have been used in clinic practice in asthma therapy for nearly 10 years (Table 1): Pranlukast (Onon®; Ono Pharmaceutical Co, Ltd, Osaka City, Tokyo, Japan), Montelukast (Singulair®; Merck & Co, Inc, Whitehouse Station, New Jersey, USA) and Zafirlukast (Accolate®; AstraZeneca Pharmaceuticals LP, Wilmington, Delaware, USA). Although there are differences in potency and pharmacokinetics (49), LTRAs have similar effects. As they are administrated orally, LTRAs are of particular value in improving patient compliance. They are also well tolerated as there have been only few class-specific side effects identified (headaches and gastrointestinal problems), rarely severe enough to cause discontinuation of the therapy (1). LTRAs are efficient in inhibiting bronchoconstrictor challenge, in particular to allergen, exercise, irritants and aspirin as well as inhaled LTD4 (50). They reduce the early and late response to allergens by approximately 50% (51). Studies comparing the effects of LTRAs with placebo in patients with mild-to-moderate symptomatic asthma have consistently shown the active drug to be more effective than placebo in improving several outcome measures including lung function tests, quality of life, night-time awakenings and asthma symptoms (52). Despite significant benefit compared with placebo, LTRAs are less effective than low doses of inhaled corticosteroids in chronic asthma (53). In the USA, the LT synthesis inhibitor (LTSI) Zileuton which targets 5-LO has been introduced to clinical practice (54). Although the effects are short lived, Zileuton blocks 70% of CysLT production. However, this drug has been associated with severe liver toxicity, limiting its clinical usefulness (55).

Allergic rhinitis

Montelukast has also been approved in the USA for treatment of seasonal and perennial AR (56). The anti-leukotriene significantly reduces day time nasal symptoms (congestion, rhinorrhea, pruritus and sneezing), day time eye symptoms, night time symptoms (difficulty sleeping, night-time awakenings and congestion on awakening) and peripheral eosinophilia (56). In addition, Zafirlukast has generated interest as it significantly improved nasal airway resistance by 68% in patients with AR (57) and reduced nasal congestion in patients allergic to cat allergens (58).

Because of the postulated role of LTs in the pathogenesis of other diseases, anti-leukotrienes are now being evaluated for other disorders (59) such as AD (41) and atherosclerosis (60) (Table 1). Although LTRAs and Zileuton are the only anti-leukotrienes used in clinical practice to date, others are in development (61, 62). Japanese studies have experimented with blocking the LTB4–LTB4R interaction with the LTB4 receptor antagonist LY293111 to prevent the chemotaxis of neutrophils into the asthmatic lung (63). This provides the potential for therapy in the future. DeCode Genetics has licensed the FLAP inhibitor DG-031 (veliflapon) created by Bayer and observed that the inhibitor reduced the incidence of MI by reducing the production of LTB4 (64).

In a randomised, double-blind, placebo controlled, 12-week trial on 895 patients with mild-to-moderate asthma, Malmstrom et al. observed that, like other anti-asthmatic drugs, the patient response to montelukast was heterogeneous (53). The reasons for this selective response pattern are not known, but may be due to many factors such as patient compliance, disease severity, co-morbidity, misdiagnosis, drug interactions and patient environment (especially for allergic phenotypes). Another factor may affect not only patient response but also the disease: the patient’s genetic profile. For this reason, the genes encoding LT synthesising enzymes, adaptor proteins and receptors are being studied for pharmacogenetic influences on LTRA and LTSI responses.

Targeting of LTs has shown clinically significant improvements in asthma and more recently allergic rhinitis particularly as add-on therapies. Interestingly, the patient response is very variable and may in part be due to the patient’s genetic profile. A better understanding of the molecular genetics of the key components of the 5-LO pathway is thus needed.

Molecular genetics of the 5-LO pathway

LT synthesis

There are many proteins involved in the formation of LTs (see Fig. 1 and Table 2). The gene encoding 5-LO (ALOX5) spans 82 kb on 10q11.2 (65). The 85 kDa protein encoded is composed of a 673 amino acid polypeptide (66). The transcription initiation site is located 65 bp upstream of the ATG and there are no TATA or CAAT sequences in the promoter (67). The gene is regulated at the transcriptional level (68) with basal mRNA levels remaining constant, but which rise rapidly during cell activation (69, 70). Another level of regulation is the calcium-dependent translocation of 5-LO to the membrane from the cytosol to arrange the protein closer to its substrate (which is derived from the membrane) (8).

Table 2.   Molecular characteristics of the genes encoding for the 5-LO pathway proteins and receptors
GenesChromosomal locationGene length (kb)Exon (no.)Protein (no. of residues)
  1. ALOX5, 5-lipoxygenase; ALOX5AP, 5-lipoxygenase activating protein; CYSLTR 1–2, cysteinyl leukotriene receptor 1–2; LTA4H, LTA4 hydrolase; LTB4R 1–2, leukotriene B4 receptor 1–2; LTC4S, LTC4 synthase; MRP1, multi-drug resistant protein 1; PPARA, peroxisome proliferator activatived receptor alpha.

ALOX510q11.28214673
ALOX5AP13q12315162
LTC4S5q352.55150
LTA4H12q223519610
CYSLTR1Xq13-q2154.93337
CYSLTR213q142.81346
LTB4R114q11.2-q1253352
LTB4R214q11.2-q122.81389
PPARA22q12-q13.183.79468
MRP116p13.16.5301531

FLAP is critical for the function of 5-LO. The 162 amino acid and 18 kDa protein is encoded by the 31 kb gene ALOX5AP. ALOX5AP is composed of five exons and four much larger introns (71) and is located on chromosome 13q12 (72). Analysis of cDNA from mammalian species has shown the protein to be highly conserved (73). Protein analysis has indicated regions more conserved between mammals (such as residues 74–123 which form the loop between transmembrane domains 2 and 3). Characterisation of the 5′ regulatory regions has identified a TATA box and glucocorticoid response elements as well as AP-1 (activator protein 1) and AP-2 transcription factor recognition sequences (71).

Both 5-LO and FLAP act upon arachidonic acid, which is released from membrane phospholipids by cytosolic phospholipase A2 (PLA2). This 85 kDa protein composed of 749 amino acids is encoded by a gene on 1q25 (PLA2G4) (74). While the genomic structure is not fully understood, there are transcription factor binding sites for NF-kappa B, AP-1, polyomavirus enhancer A- binding protein-3 (PEA-3) and NF-IL6. Crystal structure analysis identified Ser228 as involved in catalytic activity (75).

The two enzymes responsible for either CysLT or LTB4 formation are LTC4S and LTA4H respectively. LTC4S is an 18 kDa (150 amino acid) protein (76) encoded by LTC4S, a 2.5 kb gene with five exons which resides on 5q35 (77). The protein contains two consensus protein kinase C phosphorylation sites along with an N-glycosylation site and three putative transmembrane spanning regions (78). LTC4S belongs to the same family as FLAP as it shows 31% identity and a similar hydropathy profile. Promoter analysis has identified three predicted transcription start sites at −96, −69 and −66 relative to the ATG. Multiple transcription factor recognition sequences have also been identified (Sp1, AP-1, AP-2 and GATA-1) (77). LTA4H is a 35 kb gene residing on chromosome 12q22 and contains 19 exons (79). It encodes a 69 kDa enzyme from a 1833 bp open reading frame (ORF) of 610 amino acids. Promoter analysis identified a transcription start site at −151 relative to the ATG (79) and multiple transcription factor binding sites. The gene can be alternatively spliced by skipping an 83 bp exon in the 3′ coding region which produces a smaller 59 kDa protein with a different C-terminal (80).

LTB4 receptors

There are two known receptors for LTB4 named LTB4R1 and LTB4R2 (or BLT1 and BLT2). LTB4R1 is a high-affinity LTB4-specific receptor (Kd of 0.15 nm) expressed mainly on leucocytes. Despite being the high-affinity receptor, 12(R)-HETE and 20-hydroxyl LTB4 can (with much less potency than LTB4) agonize the receptor (81). LTB4R2 is less-specific, able to bind other eicosanoids and is expressed more ubiquitously (82). The structure of these LTB4R genes is yet to be fully resolved. LTB4R1 is a small gene on chromosome 14q11.2-q12. Currently, three exons have been reported, but analysis in epithelial cells has identified four different 5′ untranslated exons producing transcripts 5–7.5 kb in length (81, 83). Cloning of the cDNA expressed in leucocytes identified a 1056 bp ORF coding for a 43 kDa polypeptide of 352 amino acids. More recently, alternatively spliced variants have been identified involving a deletion of the 100 N-terminal amino acids or deletion of 39 internal amino acids (residues 43–81) (84). Analysis of the promoter region of LTB4R1 highlighted the presence of another ORF 3 kb upstream coding for a 358 amino acid polypeptide which has 45% homology to LTB4R1 (85). This gene, LTB4R2 is expressed as a major transcript of 2.5 kb (86). LTB4R2 is closely related to its mouse homolog with 93% identity. The close location of these genes indicates that they are likely to be the result of gene duplication (87). LTB4R1 and LTB4R2 show high homology in transmembrane domains 3 and 7 suggesting that this is the site for LTB4 binding (85). Agonist binding leads to the activation of a heterotrimeric guanine-nucleotide binding protein (or G-protein) made of an α-, β- and γ-subunit, which can transduce the signal from the plasma membrane to the cytoplasm. Binding to the LTB4Rs lead to the activation of phospholipase C leading to inositol-1,4,5-trisphosphate and calcium production and inhibition of adenylyl cyclase (81).

Transcription of LTB4R1 is regulated in a tissue-specific manner in macrophages and eosinophils (cells important for its chemotactic and proliferative function). The promoter contains recognition sequences for the ubiquitous transcription factor Sp-1 at -50 important for basal transcription but not tissue-specific regulation (88).

LTB4 has also been found to be a ligand for the nuclear hormone receptor peroxisome proliferator-activated receptor alpha (PPARα). The PPARA gene spans 83.7 kb and resides on chromosome 22q12-q13.1. It is expressed in tissues with a high fatty acid concentration such as the muscle, heart, kidney and liver. Activation of this receptor leads to the expression of genes containing PPAR responsive elements in their promoters (the consensus sequence AGGTCAXAGGTCA) that controls the catabolism of eicosanoids (LTs and prostaniods) (89). This produces a system of negative feedback (90, 91).

CysLT receptors

Two different CysLT-specific GPCRs have been identified based on their pharmacological profiles, CYSLTR1 and CYSLTR2. Affinity studies for CYSLTR1 agonists in transfected cells or in isolated human bronchial preparations have identified the rank order of potency: LTD4 > LTC4 > LTE4 (EC50 value: approximately 10−9, 10−8 and 10−7 m, respectively), but these results are not consistent between studies and receptor properties and function defined by functional studies differ between species and cell type (92). CYSLTR1 is antagonised by Montelukast, Pranlukast, Zafirlukast, BAY u9773 and MK-571. LTE4 has a 10-fold lower affinity for CYSLTR2 than LTC4 and LTD4 and is antagonised by the CYSLTR1/2 antagonist BAY u9773 (93, 94).

The CYSLTR1 and CYSLTR2 genes have been cloned (93, 95). CYSLTR1 is 337 amino acids, 38.5 kDa and encoded by a gene on Xq13-21. There is 32% amino acid identity to the purinoreceptor P2Y1 and 28% identity to the LTB4R1 receptor (95, 96). Cloning and sequencing of the cDNA indicated an intronless coding sequence. The genomic organisation of the human CYSLTR1 gene is only partially characterised and remains unclear. Transcripts are approximately 1.5 kb long comprising three exons with the ORF in exon 3 (92). Receptor mRNA is highly expressed in the spleen and peripheral blood leucocytes, less strongly in the lung, bronchus, small intestine, colon, skeletal muscle, pancreas and placenta (95, 96). Immunohistochemistry, Northern and reverse transcriptase-polymerase chain reaction analyses located the receptor in lung smooth muscle cells [including pulmonary artery smooth muscle (97)] and interstitial macrophages, with little or no expression in epithelial cells (95, 98).

CYSLTR2 resides on chromosome 13q14 and has been characterised by three different groups (93, 99, 100). The gene is 2548 bp long encoding a 346 amino acids polypeptide. Expression occurs in monocytes, B-lymphocytes, airway smooth muscle and eosinophils through the identification of mRNA (95, 98). The human CYSLTR2 protein is weakly expressed in lung smooth muscle cells and the strongest expression of the receptor in the lung was seen in interstitial macrophages. Moderate expression could be detected in the immune system, in particular the spleen, lymph nodes and peripheral blood leucocytes (especially monocytes and eosinophils).

The CYSLTR1 and CYSLTR2 genes show little nucleotide sequence homology and there is only 37% amino acid identity at the protein level. A third CYSLTR, CYSLTR3, not antagonised by any current antagonists has been suggested (101). Based on relatedness to the CYSLTRs and P2Y families, this could be the dual-receptor GPR17 (102).

As the genes encoding the key components of the 5-LO pathway are being elucidated, an increasing number of studies have investigated the genetic and pharmacogenetic role of polymorphisms in these genes in allergy.

LT genetics, asthma and allergy susceptibility

Asthma and allergy genetic association studies have been performed on the 5-LO pathway genes and the first two CysLT receptors, but interest is now growing towards other enzymes and LTB4 receptors. These data are summarised in Tables 3 and 4.

Table 3.   Allergic disease genetic association studies for 5-LO pathway genes
GenePolymorphismLocationPopulationAssociationReference
  1. 3′UTR, 3′ untranslated region; 5′UTR, 5′ untranslated region; A, asthma; AD, atopic dermatitis; AIA, aspirin intolerant asthma; AICU, aspirin intolerant chronic urticaria; AL, allergic; ALOX5, 5-lipoxygenase; ALOX5AP, 5-lipoxygenase activating protein; AR, allergic rhinitis; ATA, aspirin tolerant asthma; C, controls (population); C, coding region (location); I, intronic; LTA4H, leukotriene A4 hydrolase; LTC4S, leukotriene C4 synthase; NAL, nonallergic; P, promoter; SA, severe asthma; SNP, single nucleotide polymorphism.

ALOX5AP18A–21A repeatsPJapanese (71A, 71C)Yes, asthmaKoshino et al. (111)
−336 G>A,
23A–19A repeats
P
P
UK Caucasian (341 families, n = 1503)No, asthma and atopySayers et al. (107)
23A–19A repeatsPKorean (107AIA, 109 ATA, 114 C)No, AIAKim et al. (106)
SG13S25 G>A
SG13S114 T>A
SG13S89 G>A
SG13S32 C>A
rs380327 C>A
SG13S41 A>G
SG13S35 G>A
5′UTR
I
I
I
I
I
3′UTR
UK Caucasian (341 families, n = 1503)Yes, SG13S114, SG13S89, SG13S41 and haplotypes with asthma and atopy related phenotypesHolloway et al. (113)
218 A>G3′UTRKorean (93 AIA, 181 ATA, 123 C)No, AIAChoi et al. (109)
ALOX5Sp1 repeatPJapanese (55 AIA, 63 ATA, 53 C)No, AIAKawagishi et al. (105)
UK Caucasian (341 families, n = 1503)No, asthma and atopySayers et al. (107)
Korean (107 AIA, 109 ATA, 114 C)No, asthma
Yes, severity of AHR in AIA
Kim et al. (106)
Turkey (624 A)Yes, severity and exercise induced asthmaKalayci et al. (108)
−1708 G>A
21 C>T
P
C
Korean (93 AIA, 181 ATA, 123 C)Yes, AIA for both SNPsChoi et al. (109)
760 G>ACJapanese (180 A, 150 C)Yes, asthmaBai et al. (110)
LTC4S−444 A>C
(rs730012)
PPolish (47 AIA, 64 ATA, 42 C)Yes, AIASanak et al. (118)
Polish (76 AIA, 110 ATA, 75 C)Yes, AIASanak et al. (117)
Japanese (60 AIA, 100 ATA, 110 C)Yes, AIA vs ATAKawagishi et al. (105)
USA Caucasian (61 AIA, 33 ATA, 137 C)No, AIAVan Sambeek et al. (125)
Polish (26 AIA, 33 ATA)No, AIAMastalerz et al. (122)
Korean (159 AIA, 116 AICU)No, AIAKim et al. (149)
Korean (110 AIA, 125 ATA, 125 C)No, AIAChoi et al. (120)
Spanish (123 A, 43 AIA, 76 ATA, 103 C)No, asthma, AIAIsidoro-Garcia et al. (121)
Korean (93 AIA, 181 ATA, 123 C)No, AIAChoi et al. (109)
Australian Caucasian (604 A, 67 AIA, 496 ATA, 458 C)Yes, asthmaKedda et al. (115)
Japanese (349 A, 171 C)No, asthmaAsano et al. (119)
Spanish Caucasian (130 A, 78 C)No, asthmaSanz et al. (123)
UK Caucasian (341 families, n = 1503)No, asthma and atopySayers et al. (124)
UK Caucasian (23 SA, 31 C)Yes, severe asthmaSampson et al. (116)
Turkish (85 AR, 95 C)Yes, AREskandari et al. (126)
−1072 G>APUK Caucasian (341 families, n = 1503)No, asthma and atopySayers et al. (124)
Korean (110 AIA, 125 ATA, 125 C)No, AIAChoi et al. (120)
IVS1-10IAustralian Caucasian (604 A, 67 AIA, 496 ATA, 458 C)No, asthmaKedda et al. (115)
10 G>ACJapenese (141 AL, 110 NAL)Yes, allergic disease (asthma, AD, AR)Yoshikawa et al. (127)
LTA4Hrs2540475 C>T
rs2660845 A>G
rs2540482 A>G
rs17677715 T>C
rs1978331 T>C
5′UTR
5′UTR
5′UTR
I
I
UK Caucasian (341 families, n = 1503)Yes, rs1978331 and haplotypes with asthma and atopyHolloway et al. (113)
Table 4.   Allergic disease genetic association studies for CysLT receptor genes
GenePolymorphismLocationPopulationAssociationReference
  1. 3′FR, 3′ flanking region; 3′UTR, 3′ untranslated region; 5′UTR, 5′ untranslated region; A, Asthma; AD, atopic dermatitis; AIA, aspirin intolerant asthma; AICU, aspirin intolerant chronic urticaria; AL, allergic; AR, allergic rhinitis; ATA, aspirin tolerant asthma; C, controls (population); C, coding region (location); CYSLTR 1–2, cysteinyl leukotriene receptor 1–2; I, intronic; P, promoter; SNP, single nucleotide polymorphism.

  2. Unless stated otherwise (*), phenotypes were associated with the minor allele.

  3. CYSLTR1 promoter SNPs are located from the exon 4 3′end as described by Duroudier et al. (133).

CYSLTR1−945 C>T (rs321029)
−786 A>C (rs2637204)
−647A>G (rs2806489)
P
P
P
Korean
(105 AIA, 110 ATA, 125 C)
Yes, TCG haplotype with AIA (males), total IgE (females) and atopy (females)
No, BHR and FEV1
Kim et al. (131)
Japanese
(137 families with A, n = 466, 48 families with AR, n = 188)
No, asthma and ARZhang et al. (130)
−945 C>T (rs321029)PKorean
(159 AIA, 116 AICU)
Yes, AIA vs AICUKim et al. (149)
927 T>C (rs320995)CKorean
(93 AIA, 181 ATA, 123 C)
No, AIA, ATAChoi et al. (109)
Spanish Caucasian
(130 A, 78 C)
Yes, alone and in combination with LTC4S SNP −444 C allele, asthma (males)Sanz et al. (123)
Spanish Caucasian
[87 A (41 with AD), 79 C]
Yes, asthma and asthma + AD vs A + no AD (males)Arriba-Mendez et al. (128)
UK Caucasian (341 families)No, asthma, atopy and related phenotypes
Yes, atopy severity (females)*
Hao et al. (129)
899 G>A (G300S)CTristan de Cunha
(52 AL vs 60 C, 54 A vs 58 C)
Yes, atopy
Yes, asthma (females)
Thompson et al. (132)
CYSLTR2−1220 A>CPJapanese
(137 families, n = 466)
Yes, atopic asthmaFukai et al. (134)
M202V
S237L
C
C
Tristan de Cunha
(52 AL vs 60 C)
Yes, atopy (M202V)Thompson et al. (137)
601A>GCCaucasian
(359 and 384 families)
Yes, asthma (both cohorts)*Pillai et al. (136)
−819 T>G
2078 C>T
2534 A>G
2545 + 297 A>G
5′UTR
3′UTR
3′UTR
3′FR
Korean (490A, 152C)Yes, −819 T>G, 2078 C>T, 2534 A>G and haplotypes, AIA vs ATA
No, asthma
Park et al. (135)

ALOX5 and ALOX5AP

Work has focused on investigating ALOX5 and ALOX5AP as 5-LO is the key enzyme in LT synthesis which needs to interact with FLAP prior to LT synthesis. The ALOX5 5′UTR has been well characterised and by resequencing asthma patient DNA samples mutations in the region 176–146 bp upstream of the ATG (in a GC-rich region) with deletion of one, or two or addition of one zinc finger (Sp1/Egr-1) binding sites have been identified which results in reduced Sp1/Egr-1 transcription factor binding and gene transcription (103, 104). Mutations of the Sp1 repeat in the promoter are associated with airway hyper-responsiveness, but not asthma susceptibility (105, 106). Using 341 asthma enriched families, no association between the ALOX5 Sp1 polymorphism and asthma or asthma-related phenotypes was identified (107). However, the Sp1 polymorphism has been suggested as a predictor of disease severity (108). Analysis of a Korean population (n = 397) with an ALOX5 haplotype, ht1-(G-C-G-A), covering single nucleotide polymorphisms (SNPs) −1708 G>A and coding region exon 1 21 C>T, exon 2 270 G>A and exon 13 1728 A>G has shown association with aspirin intolerant asthma (AIA) [p = 0.01, odds ratio (OR) = 5.0 and 95% confidence interval (CI) of 1.54–17.9] (109). Recently, a 760 G>A (E254K) polymorphism has been described in exon 6 which results in a residue change from negative to positive in the C-terminal catalytic domain. This SNP was only seen in bronchial asthma patients (6% frequency) and not in a cohort of nonallergic subjects indicating that it may be associated with bronchial asthma (110).

Two polymorphisms have been identified in the ALOX5AP promoter, a −336 (G>A) and a poly(A) repeat at position −169 to −146 yielding 19A and 23A alleles (107). No associations were observed with asthma in a Caucasian cohort of 341 asthma families or basal transcription levels in a promoter reporter analysis (107). However, association was observed with asthma in a Japanese study with a greater frequency of the 21A allele in asthmatics compared to controls, which was statistically significant (p = 0.035) (111). The significance of this finding is questionable for the small size of the study (71 subjects and 71 controls). The ALOX5AP gene has now been completely sequenced and multiple SNPs have been identified (112). In asthma, intronic SNPs (SG13S114, SG13S89 and SG13S41) have been associated with asthma and atopy phenotypes indicating that markers in the gene are predictors for disease susceptibility (113). There has been no extensive investigation into 3′UTR SNPs; however, analysis of a 218 A>G SNP in the Korean population failed to find associations with asthma (109).

LTA4H and LTC4S

Mutation screening of the LTA4H gene identified several SNPs spanning LTA4H (114). Association with asthma and atopy phenotypes was identified for SNP rs1978331 (intron 11) and a haplotype encompassing rs1978331-rs17677715-rs2540482-rs2660845-rs240475 (113). LTC4S encodes for the main enzyme specific to CysLT synthesis and has been extensively studied, especially a polymorphism located −444 upstream of the ATG translation start site (−444 A>C, rs730012). Approximately half of the published genetic studies found the C allele associated with asthma, asthma severity or aspirin intolerance (105, 115–118). The other studies, including two also investigating another promoter polymorphism (−1072 G>A) did not detect an association with those same phenotypes (109, 119–125) (see Table 3). Interestingly, in Caucasian populations, Kedda et al. observed an association of the C allele (−444) with asthma (p = 0.03, 1062 individuals), but not asthma severity or aspirin intolerance (115). Sampson et al. and Isidoro-García et al. observed an increased frequency of this allele in severe asthma (116, 121), although this did not reach significance for the latter. Sanak et al. also detected an association of the same allele with aspirin sensitivity in two separate studies (153 and 261 individuals) (117, 118), whereas Mastalerz et al. and Van Sambeek et al. did not (59 and 201 individuals respectively) (122, 125). In Asian populations, aspirin sensitive asthma patients showed a higher frequency of the C allele in one study (270 individuals, p = 0.042), but this could not be reproduced elsewhere (109, 120). Discordances in the results are seen among the same ethnic populations (Asian and Caucasian). This raises issues regarding study design, particularly related to the relatively small size of the populations studied.

In a Turkish population of 180 individuals, heterozygotes for the −444 T>C SNP were at higher risk of developing AR (OR = 2.18; 95% CI:1.173–4.053; p = 0.014) (126). In addition, the minor allele of the recently identified coding polymorphism 10 G>A (Glu4Lys) was associated with allergic diseases (p = 0.046) among 251 Japanese individuals (127).

CysLT receptors

There are many discrepancies among studies investigating CYSLTR1 genetics. This is confounded by the fact that the gene is located on the X chromosome. In a Caucasian Spanish population, Sanz et al. reported the minor allele of the synonymous coding SNP 927 T>C (rs320995) to be associated with asthma (130 asthma vs 78 controls) (123), but Arriba-Mendez et al. observed this association only in male subjects (87 asthma vs 79 controls, p < 0.008) (128), whereas two other groups could not detect any association in a Caucasian British and a Korean population (109, 129). Although this was not reproduced in a Japanese study (137 families) (130), Kim et al. found an association of the three SNPs in the CYSLTR1 promoter with aspirin intolerant asthma among male Koreans (340 individuals, p < 0.03) (131). The authors also observed an association of these same alleles with atopy among female aspirin intolerant asthmatic individuals (p = 0.03) (131). Interestingly, we have preliminary evidence that a promoter SNP is associated with atopy incidence in a large Caucasian population (unpublished data, Duroudier et al.) In addition, in a family-based association study (341 Caucasian families), the coding SNP 927 T>C which is in high linkage disequilibrium with the promoter SNP rs2806489 was associated with atopy severity (p = 0.015), which was more significant in female asthma subjects (p = 0.0087) (129). In a high prevalence asthmatic population with a founder effect, Thompson et al. observed the minor allele of the coding polymorphism 899 G>A (Gly300Ser) to be associated with both asthma (p = 0.005) and atopy (OR = 4; 95% CI:1.2–13.3) in female subjects (132). Other studies have detected an association of the promoter or coding SNPs with allergic phenotypes such as high total IgE (p = 0.003) (131) and AD (OR = 4.97; 95% CI:1.29–19.13) (128), which could not be reproduced elsewhere (129, 130). Overall, despite these discrepancies, most studies agree that there is an association between CYSLTR1 genetic variation and allergic phenotypes and that this association is gender specific. Furthermore, the functionality of the associated promoter SNPs (tagged in the same linkage disequilibrium block) found that the minor allele (−647, rs2806489) caused a two-fold decrease in reporter expression compared to the A allele when examined using promoter-reporter analyses in human monocytes (133).

Multiple SNPs within CYSLTR2 have been found to be in association with atopic asthma or aspirin sensitivity within asthma subjects (134–137). Interestingly, a coding region variant (601 A>G) results in a decreased potency for LTD4 and was shown to be under transmitted to asthma subjects using 384 asthma families (136). The Met201Val variant was also associated with atopy (137).

LTB4 receptors and PPARA

Polymorphisms have been reported in the coding regions of LTB4R1 (138) including G>T (Ala79Ser) and T>C (Phe346Leu) and in LTB4R2 G>A (Gly196Asp), but we have not observed these in the Caucasian population (unpublished data, A.S. Tulah and I. Sayers). The role of LTB4R1 and LTB4R2 SNPs in asthma and allergy susceptibility remains to be resolved at this time. The PPARα is a nuclear hormone receptor with a cysteine-rich zinc finger DNA-binding region. Cloning and sequencing of the gene identified two non-synonymous SNPs in the DNA-binding domain (139). Leu162Val which is normally conserved between species, makes the protein unresponsive to low ligand concentrations compared to the wild-type, whilst the mutant allele has also been associated with higher plasma total and low-density lipoprotein (LDL)-apolipoprotein B and LDL-cholesterol in a French–Canadian population (140). The Arg131Gln polymorphism which is present at a lower frequency slightly raises constitutive activity of the receptor in response to the ligand. A G>A synonymous SNP in exon 7 was also identified (140). To date, association with allergy phenotypes has not been explored.

There is accumulating evidence that polymorphism spanning genes involved in leukotriene production and activity influence allergic disease phenotypes. As inflammatory diseases are complex, genetic and pharmacogenetic effects are likely to be caused by several, perhaps numerous genes, each with a small overall contribution and relative risk. Each of these genes is likely to be necessary but by themselves not sufficient to have a significant effect (141). Therefore, in addition to single SNP analyses, there is accumulating evidence to suggest that gene haplotypes over the 5-LO pathway may be functionally more relevant for activity.

Haplotype and multi-SNP analyses

Helgadottir and colleagues conducted a genome-wide linkage scan to search for genes associated with MI in an Icelandic cohort (112). ALOX5AP was shown to have the most significant association localised to a four SNP haplotype covering 33 kb designated HapA (SG13S25, SG13S114, SG13S89 and SG13S32). In a similar manner, LTA4H was investigated and HapK (defined by SNPs rs1978331, rs17677715, rs2540482, rs2660845 and rs2540475) showed association with MI (114). Interestingly, these haplotypes predicted LTB4 production in stimulated PBMCs (peripheral blood mononuclear cells) (112) suggesting that polymorphisms within ALOX5AP and LTA4H are genetic determinants of LTB4 production and may have relevance to airway disease such as asthma. Indeed, haplotype analysis with SNPs spanning ALOX5AP and LTA4H identified association or suggestive association with asthma-related phenotypes such as BHR, forced expiratory volume in 1s (FEV1), to total IgE and atopy (113).

Three groups have investigated combinations of SNPs from several genes of the 5-LO pathway. Associations of the CYSLTR2−819 G, 2078 T, 2534 G alleles with aspirin sensitivity were greater when combined with the LTC4S−444 C allele (135). Similarly, the combination of TA (927 CYSLTR1/−444 LTC4S) was significantly less common in male patients with asthma than in controls (OR = 0.37; p = 0.039) (123). Finally, GC carriers (ALOX5 SG13S41G and LTA4H rs1978331C) had an increased risk of developing asthma (OR = 2.17, 95% CI:1.41–3.32) (113).

LT pharmacogenetics

To date, there have been twelve studies specifically addressing the contribution of 5-LO pathway and receptor gene polymorphisms to clinical outcomes following LT modifier therapy (Table 5). Using a retrospective study of 221 asthmatics, the pharmacogenetic significance of the ALOX5 Sp1/Egr1 polymorphism was assessed; those possessing only mutant ALOX5 alleles were relatively resistant to treatment with the LTSI ABT-761 (68). Mean FEV1 improved by approximately 18.8 ± 3.6% (n = 64) for wild-type homozygotes and 23.3 ± 6% (n = 40) for heterozygotes compared with −1.2 ± 2.9% (n = 10) in homozygote mutant individuals (68). Similarly, Lima et al. showed a significantly higher FEV1 response to the LTRA montelukast for the ALOX5 rs2115819 SNP GG homozygotes compared to the AA and AG genotypes (p = 0.017) (142). Another study found no polymorphism specific bronchodilator responses between homozygous wild-types and heterozygotes for the ALOX5 Sp1 polymorphism in 52 asthma subjects (143). Kim et al. found no significant association between LTRA requirement and polymorphisms in the ALOX5 gene and Klotsman et al. identified an 18–25% improvement in peak expiratory flow for mutant alleles of ALOX5 SNP rs4987105 and rs4986832 (p = 0.01 for both), whereas wild-type patients only showed a 10% improvement (144, 145). These data illustrated that genes regulating the expression of LTs could influence the efficacy of therapeutics that target their synthesis and/or activity and further studies are required to determine the role of ALOX5 polymorphisms on these responses.

Table 5.   Leukotriene pharmacogenetic studies
GenePolymorphismLocationPopulationAssociationReference
  1. 3′UTR, 3′ untranslated region; A, Asthma; AA, allergic asthma; AIA, aspirin intolerant asthma; ALOX5, 5-lipoxygenase; ALOX5AP, 5-lipoxygenase activating protein; AMP, adensoine monophosphate; ATA, aspirin tolerant asthma; BHR, bronchial hyperresponsiveness; C, controls (population); C, coding region (location); CYSLTR 1–2, cysteinyl leukotriene receptor 1–2; EIA, exercise induced asthma; FE(NO), fraction of exhaled nitric oxide; FEV1, forced expiratory volume in 1 s; FVC, forced vital capacity; I, intronic; LTA4H, leukotriene A4 hydrolase; LTC4S, leukotriene C4 synthase; MRP1, multi-drug resistant protein 1; ns, not significant; P, promoter; PEF, peak expiratory force; SA, severe asthma;SNP, single nucleotide polymorphism; (-), not provided. Unless stated otherwise (*), phenotypes were associated with the minor allele. †CYSLTR1 promoter SNPs are located from the exon 4 3′end as described by Duroudier et al. (133).

ALOX5APrs3803278 C>T
rs12721458 C>G
rs3803277 A>C
I
I
I
USA (79% Caucasian, 174 A)No, FEV1, morning PEF (montelukast)Klotsman et al. (145)
23A–19A repeats
218 A>G
P
3′UTR
Korean (89 AIA)No, FEV1, PEF (montelukast)Kim et al. (144)
ALOX5Sp1 repeatPUSA Caucasian (221 A)Yes, FEV1 (ABT-761)Drazen et al. (68)
USA Caucasian (61 A)Yes, asthma exacerbations (montelukast)Lima et al. (142)
USA (79% Caucasian, 174 A)Yes, PEF (montelukast)Klotsman et al. (145)
UK Caucasian (52 AA)No BHR (montelukast, zafirlukast)Fowler et al. (143)
Korean (89 AIA)No, FEV1, PEF (Montelukast)Kim et al. (144)
Intron 2 A>G (rs2115819)IUSA Caucasian (61 A)Yes, FEV1 (montelukast)Lima et al. (142)
rs4987105 T>C
rs4986832 G>A
-
-
USA (79% Caucasian, 174 A)Yes, PEF (montelukast) for both SNPsKlotsman et al. (145)
rs2229136 A>G
rs228064 A>G
-
-
No, FEV1, PEF (montelukast)
−1708 G>A
21 C>T
270 G>A (rs2228064)
1728 A>G (rs2229136)
P
C
C
C
Korean (89 AIA)No, FEV1, PEF (montelukast)Kim et al. (144)
LTC4S−444 A>C
(rs730012)
PUK Caucasian (23 SA, 31 C)Yes, FEV1, FVC, PEF (zafirlukast, p = ns)Sampson et al. (116)
Japanese (50 A)Yes, FEV1 (pranlukast)Asano et al. (119)
Polish (26 AIA, 33 ATA)Yes, daytime symptoms, morning PEF (montelukast, p = ns)Mastalerz et al. (122)
USA (12 A)Yes, % change in FE(NO) (montelukast, p = ns)Whelan et al. (146)
USA Caucasian (61 A)Yes, asthma exacerbations (montelukast)Lima et al. (142)
USA Caucasian (150 A)No, FEV1, BHR (AMP or methacholine) (zafirlukast)Currie et al. (147)
USA (79% Caucasian, 174 A)No, FEV1, PEF (montelukast)Klotsman et al. (145)
Korean (100 EIA)No, FEV1 (montelukast)Lee et al. (148)
Korean (89 AIA)No, FEV1, PEF (montelukast)Kim et al. (144)
−1072 G>APKorean (89 AIA)No, FEV1, PEF (montelukast)Kim et al. (144)
LTA4Hrs2660845 A>G5′UTRUSA Caucasian (61 A)Yes, asthma exacerbations (montelukast)*Lima et al. (142)
CYSLTR1−945 C>T (rs321029)
927 T>C (rs320995)
PKorean (89 AIA)Yes, FEV1, PEF (montelukast)Kim et al. (144)
927 T>C (rs320995)CKorean (100 EIA)No, FEV1 (montelukast)Lee et al. (148)
CYSLTR2rs912278 T>C
rs912277 T>C
C
C
USA (79% Caucasian, 174 A)Yes, PEF (montelukast) for both SNPsKlotsman et al. (145)
MRP1Intron 1 C/T rs119774IUSA Caucasian (61 A)Yes, FEV1 (montelukast)Lima et al. (142)

For a LTA4H rs2660845 G>A SNP, possession of the minor G allele gave a greater likelihood of experiencing an asthma exacerbation, which was further increased in the homozygote state (OR for GA and GG >4, = <0.01) while taking LTRA (142).

Five of the nine published pharmacogenetic studies on the LTC4S−444 A>C polymorphism showed an improved response to LTRAs from patients carrying the C allele (116, 119, 122, 142, 146), although results were not always statistically significant (116, 122, 146) (perhaps partly due to the small number of individuals in the populations studied). The other four published studies could not reproduce these findings (144, 145, 147, 148). As these results were obtained in different populations and using different end points, this suggests that, although not sufficient to explain the variation in patient response to LTRAs, this polymorphism is a strong candidate and is potentially exerting its influence with other gene polymorphisms.

Interestingly, Lima et al. published for the first time a pharmacogenetic study of MRP1, one of the LTC4 transmembrane transporters to the extracellular space. The presence of the T allele (intronic SNP, rs119774) significantly improved the response of 61 asthmatics to Montelukast (FEV1) (142). This SNP has also shown an association with BHR (Methacholine) in asthma subjects (p = 0.003) using 341 asthma families (113).

To our knowledge, only three pharmacogenetic studies investigating CYSLTRs have been published. In a Korean population of 100 children with exercise-induced asthma, the synonymous coding CYSLTR1 polymorphism T>C (rs320995) did not affect the efficacy of montelukast to improve FEV1 (148), but the minor alleles of both the promoter and coding SNPs [C>T (rs321029) and T>C (rs320995)] were associated with requirement for montelukast in another Korean population of 89 aspirin intolerant asthmatics (p < 0.02) (144). Carriers of the variants of two polymorphisms in the CYSLTR1 gene (rs91227 and rs912278) responded better to montelukast (145).

Overall, these pharmacogenetic studies provide preliminary evidence that polymorphisms spanning genes from the 5-LO pathway may influence clinical outcomes of anti-leukotriene therapies in allergic disease.

Future implications

In the past decade, it has become clear that LTs are key inflammatory mediators and new roles in the pathogenesis of a number of diseases are being discovered. Leukotriene modifiers are already used in clinical practice for asthma and AR and are in trial for other diseases including AD. Despite the efficacy of these drugs, the patient response remains highly variable and the mechanisms of action of LTs remain to be elucidated. As the genes of the key components of the 5-LO pathway are being characterised, one could hypothesise that genetic variation in those genes could affect disease risk or severity and/or patient response to anti-leukotrienes, i.e. not only predispose to allergy but also predispose to a more LT-dependent allergy.

Discrepancies between studies are likely to be due to the experimental design of the studies, which were performed in populations from different ethnic origins, using different criteria to characterise the same phenotypes (e.g. atopy defined as elevated specific IgE or positive skin prick test) and looked at SNPs in isolation rather than in haplotypes. This reflects the importance of ethnicity in such studies as population stratification often significantly affects the effect of a polymorphism, whether genetic or pharmacogenetic, constituting a crucial element of more personalised health care. The population sizes are also of significance as most of the studies described used small numbers of individuals, creating the potential for type I and type II errors. Despite these discrepancies, overall, polymorphisms in the ALOX5, LTC4S and CYSLTR1 genes seem likely to have a real genetic and/or pharmacogenetic effect with relevance to allergy. However, individual polymorphisms may be predicted to have a small influence in complex multifunctional diseases. Combinations of polymorphisms across the entire pathway are therefore more likely to be responsible for disease susceptibility/severity and variability of the patient response to anti-leukotrienes.

Another interesting element arising out of these studies is the gender-specificity seen with CYSLTR1 polymorphisms located on chromosome X. This is of major importance regarding not only the administration of anti-leukotriene therapies but also the signalling of CysLTs. The origin of this gender segregation could have several causes such as X-inactivation or a sex-specific regulator (oestrogen).

In conclusion, polymorphisms in the genes encoding enzymes and receptors of the 5-LO pathway are likely to have a genetic and/or pharmacogenetic impact in allergic disease. However, the relevance of these associations needs to be asserted with larger more robustly designed studies.

Ancillary