Polyunsaturated lysophosphatidic acid as a potential asthma biomarker

Abstract

Lysophosphatidic acid (LPA), a lipid mediator in biological fluids and tissues, is generated mainly by autotaxin that hydrolyzes lysophosphatidylcholine to LPA and choline. Total LPA levels are increased in bronchoalveolar lavage fluid from asthmatic lung, and are strongly induced following subsegmental bronchoprovocation with allergen in subjects with allergic asthma. Polyunsaturated molecular species of LPA (C22:5 and C22:6) are selectively synthesized in the airways of asthma subjects following allergen challenge and in mouse models of allergic airway inflammation, having been identified and quantified by LC/MS/MS lipidomics. This review discusses current knowledge of LPA production in asthmatic lung and the potential utility of polyunsaturated LPA molecular species as novel biomarkers in bronchoalveolar lavage fluid and exhaled breath condensate of asthma subjects.

Asthma is a chronic and complex airway disease characterized by airway inflammation, airway hyper-responsiveness and variable reversible airway obstruction [1,2]. Despite recent advances in understanding the molecular and pathophysiological basis of asthma, there is no cure for this disease; however availability of inhaled corticosteroids, short and long-acting β-agonists, and to some extent leukotriene C4 antagonists, has proven to be effective in managing the disorder [3,4]. Progress in developing new therapies against asthma has been slow due to heterogeneity of the disease and the existence of multiple phenotypes resulting from complex gene-environment interactions [5]. A better understanding of asthma phenotypes is essential for improved diagnosis and management, and biomarkers should provide predictive information regarding diagnosis, prognosis, progression, phenotypes, disease mechanisms, risks and treatment responses. Biomarker is a term commonly used to objectively measure and evaluate the severity of biological or pathological processes or responses to therapeutic treatments. Several biomarkers have been identified and evaluated to assess the airway inflammation associated with asthma and its pathogenesis [6,7]. These include physiological measurements of air flow such as forced expiratory volume (FEV₁), numbers of eosinophils in blood and sputum, levels of total and antigen-specific immunoglobulin E (IgE), proteomics, genomics and lipidomics in blood and sputum, and fractional exhaled nitric oxide (FeNO), 8-isoprostane and oxidant markers in exhaled breath condensate (EBC). However, many of these biomarkers are not specific for asthma, as they have been identified in other respiratory pathologies such as chronic obstructive pulmonary disease, pulmonary fibrosis and pulmonary hypertension. Therefore, there is a clear need to identify and characterize novel biomarkers for asthma and asthma phenotypes for differential diagnosis and to guide effective therapies. The application of ‘Omics’ that includes genomics, proteomics, lipidomics and metabolomics to biological fluids and exhaled breath condensates will be necessary to improve the management and treatment of asthma. Lipidomics has identified the induction of polyunsaturated lysophosphatidic acid (LPA) of chain lengths C22:5 and C22:6 in bronchoalveolar lavage (BAL) fluids after subsegmental bronchoprovocation with allergen (SBP-AG) in subjects with allergic asthma [8]. The accumulation of these unusual polyunsaturated LPA molecular species in BAL fluid and not in the plasma following SBP-AG suggests that these bioactive molecular species of LPA may be useful biomarkers in certain asthma phenotypes for tracking disease activity and severity.

The present review describes the mechanisms for generation LPA including polyunsaturated molecular species during SBP-AG by autotaxin (ATX)/lysoPLD, the alveolar cell types involved in ATX-mediated LPA production in the lung from the lysophosphatidylcholine (LPC) phospholipid precursor, and polyunsaturated LPA as a potential biomarker for allergic asthma.

LPA, LPA receptors & LPA metabolism

LPA is a naturally occurring bioactive lipid and one of the simplest glycerophospholipids, with long chain fatty acids linked to either sn-1 or sn-2 position on the glycerol backbone [9,10]. In addition to acyl group(s), LPA may also have long chain alkyl- or alk-1-enyl moieties on the sn-1 position of the glycerol backbone. The predominant molecular species of LPA present in biological fluids and tissues contain saturated (C16:0 and C18:0) and monounsaturated (C18:1) long chain fatty acids; however, LPA molecular species containing polyunsaturated fatty acids (C18:2, C20:4, C22:5 and C22:6) have been reported in plasma, serum and other biological fluids, suggesting different biosynthetic pathways for LPA biosynthesis [11]. In mammalian cells, de novo biosynthesis of LPA is mediated by glycerophosphate acyltransferase and acylglycerol kinase, enzymes that are predominantly localized in endoplasmic reticulum and mitochondria, respectively [8,11,12]. Additionally, intracellular LPA can be generated from phosphatidic acid (PA) generated by phospholipase D action on phosphatidylcholine or phosphatidylethanolamine with subsequent conversion of PA to LPA by PA specific PLA₁ or PLA₂ (Figure 1) [8,11]. In contrast to pM levels of intracellular LPA, plasma LPA levels range from nM to μM [13,14], and extracellular LPA generation requires multistep enzymatic pathways involving either secretory sPLA₂ or phosphatidylserine (PS) specific PLA₁ and ATX (lysoPLD) [8,11,15]. ATX cleaves the choline, ethanolamine or serine moiety of LPC, lysophosphatidylethanolamine or lysophosphatidylserine, respectively, to generate LPA (Figure 1) [11]. ATX, can also hydrolyze sphingosinephosphorylcholine to produce sphingosine-1-phosphate [16]. Cellular ATX is mostly inactive as compared with secreted ATX [11], and ATX expression is upregulated during asthmatic airway inflammation [8,17,18] and bleomycin induced lung fibrosis [19–21].

Figure 1. . Pathways of lysophosphatidic acid (LPA) production in mammalian cells.

(A) Phospholipids including phosphatidylcholine (PC), phosphatidylethnolamine (PE) and phosphatidylserine (PS) are hydrolyzed by phospholipase A (PLA) 1 or 2 to 1-acyl- or 2-acyl-lysophospholipids, respectively, which are subsequently converted to 1-acyl- or 2-acyl-lysophosphatidic acid (LPA) by autotaxin/lysophospholipase D (LysoPLD). Alternatively, PC, PE or PS is hydrolyzed by phospholipase D (PLD) 1 or 2 to generate phosphatidic acid (PA) that is converted to 1-acyl- or 2-acyl-LPA by membrane bound PA selective PLA₁ or PLA₂, respectively. (B) Diacylglycerol (DAG) in cells is converted to PA by DAG kinase (DAGK), which is then converted to LPA (1-acyl- or 2-acyl-) by PA specific PLA₁ or PLA₂, respectively. (C) Monoacylglycerol in cells is converted to LPA by the action of acylglycerol kinase (AGK) enzyme.

The biological effector functions of LPA are attributed to at least six G-protein coupled LPA receptors (LPA_1–6) with overlapping specificities and varying tissue distribution [22]. LPA receptors are expressed by lung epithelial and endothelial cells as well as infiltrating inflammatory cells including eosinophils, macrophages, neutrophils, T cells, mast cells and dendritic cells (DCs) [23–27]. Eosinophils express LPA₁ and LPA₃, but not LPA₂ [ackerman sj, natarajan v, unpublished data] and LPA_1–3 are expressed on both mature and immature DCs [28,29]. LPA_1&2 interact with Gα_i, Gα_q and Gα_12/13; LPA₃ interacts with Gα_i and Gα_q but not Gα_12/13; LPA₄ appears to couple with all the G-proteins; LPA₅ is coupled to Gα_s, Gα_12/13 and Gα_q, and LPA₆ is coupled to Gα_12/13 [12]. The orphan GPR87 and P2Y10 receptors have similarities to LPA receptors, while peroxisome proliferation-activated receptor γ has been identified as an intracellular receptor for LPA [30].

Activities of LPA on airway epithelial cells, smooth muscle cells, T lymphocytes & dendritic cells

LPA has a potent action on most of the lung cells thought to be involved in airway inflammation in asthma including airway epithelial cells, smooth muscle cells, T-lymphocytes and DCs. Human airway epithelial cells express LPA₁, LPA₂ and LPA₃ [23], whereas qRT-PCR assessments of LPA receptor mRNA in mouse tracheal epithelium show expression of LPA₂ > LPA₄ > LPA₁ ≥ LPA₃ [31,32]. LPA induces murine tracheal epithelial cells to express TSLP and CCL20 via CARMA3-mediated NF-κB activation [33]. In contrast, in human bronchial epithelial cells, LPA has potential anti-inflammatory activity through induction of IL-13 decoy receptor (IL-13Rα2) and prostaglandin E2 (PGE₂) expression [23]. IL-13Rα2 has higher affinity than IL-13Rα1 for IL-13, and blocks its signaling and activities. LPA-induced IL-13Rα2 expression/secretion by human bronchial epithelium occurs through the PLD- and JNK-mediated pathways, and LPA attenuates IL-13-induced STAT6 phosphorylation and cytokine release, suggesting an anti-inflammatory role for LPA via attenuation of IL-13-induced airway remodeling activities. Finally, LPA induces cyclooxygenase-2 (COX-2) and PGE₂ release by human bronchial epithelial cells, and unlike other tissues, COX-2 and PGE₂ have anti-inflammatory roles in the lung [34,35]. Airway smooth muscle (ASM) cells express mRNA encoding LPA₁-LPA₃ [36]. LPA induces ASM cell contraction and increases contractile responses to methacholine in isolated tracheal rings from rabbits and cats, and decreases relaxation to the β-adrenergic agonist adenylate cyclase activator, forskolin [37]. LPA also induces actin reorganization through Gαi-2 and Gαq proteins and stimulates cAMP accumulation, PI hydrolysis, and activation of Rho in ASMs, and induces ASM cell proliferation in a concentration-dependent manner [38]. LPA activates transcription factors including NF-κB, AP-1, CREB, NFAT and SRE complex in ASM cells, factors known to regulate inflammatory cytokine expression, suggesting LPA may induce cytokine production and release by ASM [39]. For T-lymphocytes, Th1 and Th2 cells both express cell-surface LPA₁, LPA₂ and LPA₃ based on flow cytometry, with Th2 cells expressing higher LPA₁ levels compared with Th1 [17,40]; however, the flow cytometry data have to be interpreted with caution as the LPA receptor antibodies were used in permeabilized cells. LPA has been shown to increase intracellular Ca²⁺ and chemotaxis of Th2 cells in a pertussis toxin-sensitive and PI3K-independent manner [40], and LPA and LPA₂ regulates T cell motility in vitro and in vivo [41] with no major changes in LPA receptor mRNA expression in T helper subsets. As well, LPA augments IL-13 gene expression and secretion in submaximally activated T cells, and enhances transcriptional activity of the IL-13 promoter in Th2 cells [42]. Thus, these studies suggest a role for LPA in both Th2 chemotaxis and cytokine release. In terms of DC activation and function, both murine bone marrow-derived immature DCs and LPS-matured DCs express LPA₁, LPA₄ and LPA₅, while immature DCs express higher levels of LPA₃ [17,43]. Both immature and mature human DCs express similar levels of LPA₁, LPA₂ and LPA₃. The effects of LPA on immature versus mature DCs differ, with LPA inducing intracellular Ca²⁺, actin polymerization and chemotaxis only on immature DCs [44]. Of relevance to the current review (see below), only polyunsaturated LPA species induced chemotaxis of immature DCs, and mature mouse DCs did not respond to LPA. The effect of polyunsaturated LPA on DC migration was dependent on expression of LPA₃, since immature DCs generated from LPA₃ ^-/- null mice showed a 50% reduction, and LPA₃ antagonist a 70% reduction, in migration. In contrast, for LPS-matured myeloid DCs, LPA enhanced IL-4 secretion, and enhanced IL-10 and inhibited IL-12 and TNF-α secretion from mature DCs [45], and LPA inhibited LPS-dependent DC activation in part via LPA₂ deficient DCs, which were hyperactive for in vivo adoptive transfer [46].

LPA in airway inflammation

LPA exerts multiple effects on airway and lung inflammation in vivo and in vitro suggesting its potential role in the pathobiology of asthma. LPA inhalation increased eosinophils and neutrophils in BAL fluids, and also enhanced the airway response induced by iv. administration of acetylcholine in guinea pigs [26]. Intratracheal administration of LPA into mouse lung enhanced MIP-2 levels and neutrophil infiltration in the alveolar space at 6 h, which returned to near normal basal levels at 12 h, suggesting initial rapid recruitment in the lung in response to LPA [47]. In contrast to asthma models, LPA administration protected mice against LPS-induced acute lung inflammation and injury, indicating its action on epithelial cells and not smooth muscle cells [48,49]. In vitro, LPA exhibits multiple effects on airway cells such as increased smooth muscle cell contractility, mesenchymal cell proliferation, epithelial cell barrier enhancement, stimulation of pro- and anti-inflammatory cytokine expression by human bronchial epithelial cells and T cell homing, suggesting multiple and differential roles of LPA in various lung cell types [8,17,47].

LPA in the pathogenesis of allergic airway inflammation in asthma

Georas and Natarajan were the first to report that LPA is detectable in the BAL fluid of allergic subjects at baseline, and was significantly increased 18 h after SBP-AG. LPA, detectable in the BAL fluid from control lung segments, was significantly induced 18 h after allergen challenge [50]. Total LPA levels were poorly correlated with the numbers or percentages of eosinophils, and only weakly with neutrophils and lymphocytes, suggesting LPA is not a dominant chemoattractant for eosinophils or other inflammatory cells in SBP-AG-induced late allergic responses in the airways. Of note, LPA levels from control sites were inversely correlated with total BAL protein, suggesting LPA enhances epithelial barrier function at baseline; LPA was similarly found to increase human bronchial epithelial cell barrier activity in vitro. Thus, LPA is present in human airways at baseline and its expression increases significantly during allergic airway inflammation. This finding was reproduced in both human asthma SBP-AG subjects from two different asthma cohorts subjected to SBP-AG, and mouse allergic asthma models [8]. In both models, total LPA levels were substantially increased in the airspaces as measured in BAL fluid in response to allergen challenge (Figure 2). One of the functions of LPA in the normal lung may be maintenance of epithelial barrier integrity, while its induction, for example in allergic airway inflammation in asthma, is more likely to be proinflammatory as elaborated in detail above. Another study reported increased LPA levels in the airways in idiopathic pulmonary fibrosis (IPF) in both patients and the mouse bleomycin IPF model [24]. LPA levels were increased in BAL fluid following bleomycin-induced lung injury, and mice lacking one of the LPA receptors, LPA₁, were markedly protected from fibrosis and mortality; the LPA₁ deficiency reduced fibroblast recruitment and vascular leak, both of which contribute to fibrosis. Similarly, patients with IPF had increased LPA levels in their BAL fluid, and inhibition of LPA₁ ex vivo markedly reduced fibroblast chemotactic responses to the BAL fluid.

Figure 2. . Allergic airway inflammation induces increased levels of total and polyunsaturated LPA C22:5 in BAL fluid of human asthma subjects and a mouse allergic asthma model.

(A) Total and polyunsaturated LPA C22:5 induced by SBP-AG in the BAL fluid of three mild allergic asthma subjects. BAL was performed before (0 h) and 48 h after allergen instillation (house dust mite, ragweed or Aspergillus sp. based on patient skin-test sensitivity). Control lung segments were saline-challenged. Total LPA and saturated and polyunsaturated LPA species were quantitated by LC/MS/MS. Insert shows the LC/MS/MS profile of LPA molecular species from the allergen-challenged segment after 48 h. The LPA C22:5 represents the major species induced by allergen challenge. (B) Triple (DRA) allergen-induced total and polyunsaturated LPA C22:5 in the BAL fluid of ATX transgenic over-expressing (ATX-TG) mice. BAL was performed on Day 15 of the DRA allergen sensitization/challenge protocol in ATX-TG (right) and wild-type FVB mice (left) (n = 4/group, mean ± SEM). Control (CTR) mice were DRA-sensitized/saline-challenged. The insert shows the lipidomics profile of LPA species in the BAL fluid from a representative ATX-TG over-expressing, allergen-challenged mouse; LPA C22:5 was the major molecular species (>50%) induced by allergen challenge.

*p < 0.05 compared with saline-challenged control mice.

Polyunsaturated molecular species of LPA in allergic airway inflammation in asthma

SBP-AG in patients with intermittent allergic asthma resulted in a remarkable increase in BAL fluid levels of LPA, with LPAs containing polyunsaturated fatty acids (C22:5 and C22:6) showing the greatest increases, up to 22-fold. The enrichment of polyunsaturated species of LPA is particularly notable, since saturated and monounsaturated LPAs (C16:0, C18:0 and C18:1) were only minimally induced. In BAL fluid from allergen challenged asthmatics, polyunsaturated 22:5 LPA molecular species constituted >50% of the total LPA released into the airspace with much lesser amounts of polyunsaturated 20:4 LPA, unsaturated 16:1 and saturated 18:0 LPA (Figure 2A). In contrast to increased 22:5 LPA in asthmatic BAL fluid, saturated 18:0 LPA was elevated in the plasma of asthmatics [51] suggesting that LPA in BAL fluid is not likely simply due to plasma leak. We therefore concurrently analyzed LPA molecular species in plasma following SBP-AG and clearly showed that plasma LPA molecular species were largely unchanged after airway allergen challenge [8], strongly indicating that LPA species in the lung are not simply derived from plasma extravasation. As well, this implies that LPA production in the lung is compartmentalized from that generated at other tissue sites such as the peripheral circulation. The increased accumulation of the signature polyunsaturated LPAs (C22:5 and C22:6) in SBP-AG challenged allergic asthma subjects was verified by us in a murine model of allergic asthma (Figure 2B). Exposure of mice to a triple allergen mixture containing house dust mite, ragweed and Aspergillus sp. (DRA) increased the level of LPA 22:5 in BAL fluid compared with control mice [8]. Another murine model of allergic airway inflammation induced by Schistosoma mansoni egg antigen sensitization and challenge also showed increased polyunsaturated LPA molecular species (C22:5 and C22:6) compared with control mice [31]. Currently, no data are available comparing signaling of the saturated versus monounsaturated versus polyunsaturated LPAs, and it is possible that the degree of LPA poly-unsaturation may affect its affinity and specificity for the various LPA receptors.

Autotaxin expression in BAL fluid of asthma subjects during allergic airway Inflammation

ATX is the key enzyme that regulates extracellular generation of LPA from LPC [11]. ATX is present in plasma where it is responsible for generating plasma levels of LPA [14]. ATX is minimally expressed in the BAL fluid of normal control subjects and correlates with the very low levels of LPA detectable in the airways. In contrast, in allergic asthma subjects, following performance of the SBP-AG procedure, ATX expression was significantly increased in BAL fluids from allergen challenged lung segments compared with BAL fluids from contralateral control and preallergen challenge segments, suggesting a causal link between allergic airway inflammation, induction of ATX expression and increased LPA levels [8,50]. A role for ATX in allergic asthmatic airway inflammation was further established, as ATX-overexpressing transgenic mice exhibited a more robust response with increased eosinophilic airway inflammation and Th2 cytokine responses, whereas blocking ATX activity with an inhibitor, GWJ-23 and knockdown of LPA₂ receptor in mice attenuated allergic airway inflammation and Th2 cytokine responses [8], confirming a role for the ATX/LPA/LPA receptor pathway in the pathogenesis of asthma. However, in a recent study, pharmacological activation of LPA₂ with a small molecule LPA₂ agonist (2-[4-(1,3-Dioxo-1H,3H-benzoisoquinolin-2-yl) butylsulfamoyl]benzoic acid (DBIBB) attenuated Th2-driven allergic airway inflammation in a mouse model of asthma [52]. The discrepancy between genetic deletion of LPA₂ and LPA₂ agonist in affecting allergen-driven airway inflammation is unclear, and the effect of the LPA₂ agonist on the expression of other LPA receptors and LPA metabolism in vivo needs to be established. The mechanism(s) and cells involved in the secretion and transcriptional regulation of ATX in the asthmatic lung after SBP-AG challenge are unclear; however, infiltrating macrophages and eosinophils appear to be the principal cell types that secrete ATX into the alveolar space [8]. LPC is the preferred substrate for ATX in generating LPA [11], and notably, BAL fluids of SBP-AG-challenged asthma subjects exhibited enhanced levels of polyunsaturated LPC (C22:5 and C22:6) as determined by lipidomics mass spectrometry [natarajan v, berdyshev e and ackerman sj, unpublished data], further indicating a different origin of the polyunsaturated LPAs identified in BAL fluids compared with plasma in these subjects.

Potential role of phospholipase A1 (PLA₁) in the generation of LPA C22:5 & C22:6 molecular species in allergic airway inflammation in asthma

The fatty acid composition of PC in biological fluids is distinct at the sn-1 and sn-2 position of the glycerol backbone. The sn-1 position is predominantly occupied by C16:0, C18:0 and C18:1 fatty acids while the sn-2 position consists primarily of C18:2, C20:4, C22:5 and C22:6 fatty acids [53]. Generation of LPC, the substrate for ATX, from PC is mediated by either PLA₁ or PLA₂ resulting in two different types of LPC with saturated/monounsaturated fatty acids at sn-1 or polyunsaturated fatty acids at sn-2 position of the glycerol backbone (Figure 3). Hence, accumulation of the polyunsaturated C22:5 and C22:6 LPA molecular species in BAL fluid of SBP-AG challenged asthmatics suggests a potential role for the PLA₁ pathway. A preliminary analysis of BAL fluids by Western blotting and ELISA confirmed higher levels of phosphatidylserine-specific PLA₁ (PSPLA₁) expression in BAL fluid from SBP-AG challenged allergic asthma subjects compared with BAL fluids from the same subjects prior to allergen challenge and nonallergen challenged control lung segments [natarajan v, ackerman sj et al., unpublished data]; however, the PLD→PC to PA and PA to LPA by PA specific PLA₁ cannot be ruled out (Figure 1).

Figure 3. . Role of Phospholipase A1 (PLA₁) in the generation of polyunsaturated LPA species.

Generation of polyunsaturated LPC substrate (from PC) for cleavage by ATX/lysoPLD that contains the C22:5 (or C-22:6) fatty acid in the sn-2 position is mediated by either PLA₁ or PLA₂ enzymes to produce two different LPC substrates containing saturated/monounsaturated fatty acids at sn-1 or polyunsaturated fatty acids at sn-2 position of the glycerol backbone. Allergic inflammation-induced accumulation of polyunsaturated C22:5 and C22:6 LPA molecular species in BAL fluids of SBP-AG-challenged asthmatics and allergen sensitized/challenged mice (see Figure 2) suggests a probable role for the PLA₁ pathway.

Detection of LPA molecular species in exhaled breath condensate

As noted above, we have demonstrated increased levels of polyunsaturated LPAs (C22:5, C22:6) and total LPA in mild allergic asthma subjects following SBP-AG [8]. We have similarly analyzed the profile of LPA molecular species in EBC from five of these allergic asthma subjects undergoing SBP-AG; asthma diagnosis was confirmed by spirometry and bronchodilator reversibility or methacholine challenge. EBCs and BAL fluid were collected before and 48 h after the SBP-AG protocol and analyzed by mass spectrometry [54]. Three of the five subjects provided adequate EBC sample for lipidomics analysis. Of the three subjects, two demonstrated strong airway inflammatory responses to allergen challenge while one had an attenuated response. Lipidomics analysis by LC/MS/MS in the subjects with strong allergic responses showed a selective increase in the polyunsaturated LPA C22:5 and LPA C22:6 in their BAL fluid. However, the LPA composition in the EBCs did not correspond to the BAL fluid, with EBC LPA molecular species being predominantly saturated LPA C16:0 and C18:0. These findings demonstrate that LPA C22:5 is selectively induced in the airway of allergic asthma subjects, and may translate into a novel biomarker and target for asthma treatment. However, the increases in LPA C22:5 in BAL fluid were unfortunately not reflected in EBCs, suggesting poor and potentially differential partitioning of polyunsaturated versus saturated LPA molecular species into the micelle/vapor phase during respiration. Alternatively, since polyunsaturated fatty acids in LPA are prone to lipid peroxidation, EBCs may need to be monitored for oxidized LPAs and fatty acids derived from the LPA 22:5 and 22:6 molecular species. A recent study of EBCs obtained from a small cohort of 11 patients with IPF and equivalent number of normal control subjects reported a small but significant increase in polyunsaturated docosatetraenoyl LPA (C22:4), but not total LPA in the IPF patients [55]. However, the levels of LPA C22:4 detected were extremely low and it is unclear whether they will have any utility as a biomarker for IPF in terms of disease severity or progression.

Polyunsaturated LPAs as potential biomarkers of allergic airway inflammation & disease severity in asthma

Three separate but small asthma cohorts have been analyzed to date, all of which demonstrated potent induction of polyunsaturated LPAs in alveolar airspaces in response to airway allergen challenge using SBP-AG and BAL after 48 h [8,49]. In all three cohorts, LPA production was variable from subject to subject with some minimal or nonresponders (e.g., see Figure 4), suggesting there may be subsets of allergic asthmatics whose airway inflammation is more or less dependent on the ATX/LPA axis. Notably, for patient safety reasons, all of these SBP-AG studies have been conducted in subjects with mild intermittent allergic asthma. To the best of our knowledge, there have not been any studies to date addressing potential correlations between elevated total LPA or the polyunsaturated LPA C22:5 or C22:6 molecular species that comprise the majority of the LPA response, with asthma disease severity, comparing LPA levels in BAL fluids from subjects with mild, moderate and severe asthma to normal healthy controls or subjects with other pulmonary diseases. As such, the sensitivity and specificity of lipidomics assessment of polyunsaturated LPAs as biomarkers for asthma diagnosis, phenotyping, disease severity or status during treatment warrants further investigation.

Figure 4. . LPA molecular species in the BAL fluid of subjects with mild allergic asthma before and after airway allergen challenge.

(A) Profiles of saturated and polyunsaturated LPA molecular species in the BAL fluid of patients with mild allergic asthma subjected to the SBP-AG protocol at the University of Wisconsin–Madison (Mean ± SEM, n = 21). Significant increases in polyunsaturated LPA C22:5, and to a lesser degree LPA C22:6, were measured by LC/MS/MS in BAL fluids from the allergen-challenged lung subsegments (*p < 0.05, paired t test). (B) Levels of total and polyunsaturated LPA C22:5 and C22:6 in the BAL fluid of the individual subjects in (A) induced by the SBP-AG protocol. Modified and reprinted with permission from [8] © American Thoracic Society (2015). The American Journal of Respiratory and Critical Care Medicine is an official journal of the American Thoracic Society.

Conclusion

Novel polyunsaturated LPAs have now been firmly identified as potential biomarkers of airway inflammation in patients with allergic asthma. Whether levels of these polyunsaturated LPA species will correlate with disease severity in asthma and/or have utility for phenotyping asthma or following patient responses to treatment remains to be addressed. Emerging evidence suggests that the composition of the LPA fatty acyl side chains may affect their affinity for the various LPA_1–6 receptors. Whether this translates directly into novel biological effects for the polyunsaturated LPA biomarkers identified in allergic late phase reactions in asthma subjects remains to be determined. Support for this possibility comes from an earlier report that unsaturated LPAs preferentially induced chemotaxis of immature murine DCs [43]. Future studies of cell-specific receptor-mediated functions of polyunsaturated LPAs in the asthma diathesis need to address this possibility. A small number of human and mouse model studies now establish a role for the ATX/LPA axis in the pathogenesis of allergic airway inflammation in asthma (Figure 5). Of note, in the three human asthma cohorts analyzed to date, LPA production in response to airway allergen challenge was highly variable, suggesting there could be subsets of asthma subjects whose airway inflammation is more or less dependent on the ATX/LPA axis. Many unanswered questions remain, most importantly the specific cell sources for the enzymes and substrates involved in the LPA metabolic pathway, and the mechanism(s) by which LPA enhances allergic airway inflammation (Figure 5). Future studies need to address whether LPA promotes leukocyte (e.g., eosinophil, macrophage) recruitment to the airways directly through LPA-R signaling or indirectly (e.g., by inducing chemokine expression by airway epithelium). Whether LPA acts directly on ASM to enhance airway contractility and/or hyper-reactivity in asthma also warrants additional investigation. As well, if repeated allergen exposures induce repeat cycles of airway LPA production, then LPA-driven injury/repair responses might contribute to long-term airway remodeling in chronic asthma. Finally, studies are needed to reconcile the pro- versus anti-inflammatory activities described for LPA, with their potential for playing a more regulatory role in the development of allergic airway inflammation and asthma pathogenesis in general.

Figure 5. . Schema depicting the generation and potential roles of polyunsaturated LPA in asthma pathogenesis.

The cell source(s) of the initial phospholipid substrate containing the polyunsaturated C22:5 (and C22:6) fatty acids has not been determined. Enzymatic activity of one of the PLA₁ enzymes, possibly phosphatidylserine-specific PLA1 (PSPLA1) expressed by airway epithelium, generates the LPC 22:5 (and 22:6) substrate for autotaxin (ATX)/lysophospholipase D-mediated generation of the predominant polyunsaturated LPA 22:5 and 22:6 molecular species shown to be induced during allergen-mediated airway inflammation in patients with asthma and in mouse asthma models. ATX is expressed and secreted into the airspaces mainly by alveolar macrophages and airway epithelial cells in response to activation by IL-4, contributed by either Th2 lymphocytes, activated type 2 innate lymphoid cells (ILC2s) and/or eosinophils, which express considerable amounts of IL-4. Polyunsaturated LPA 22:5 (and 22:6) interact and signal through the LPA2 and possibly other LPA receptors on inflammatory cells, airway epithelium and other resident lung cells to promote inflammatory cell recruitment and activation, leading to airway inflammation, hyperreactivity (AHR) and remodeling (goblet cell metaplasia, subepithelial fibrosis), cardinal pulmonary tissue responses associated with asthma pathogenesis.

Future perspective

Identification of asthma biomarkers, which could be used for diagnosis, severity of the disease, progression and responsiveness to treatment regimens is a major challenge in pediatric and adult asthma. Recent advances in biotechnology and ‘Omics’ will allow basic, translational and clinical researchers to identify new and novel asthma biomarkers. An important issue will be the validation and translation of the present findings of polyunsaturated molecular species of LPA, their LPA receptors and enzymes of LPA metabolism in allergic airway inflammation and asthma into patient phenotyping, personalized care and treatment.

Executive summary.

LPA, LPA receptors & LPA metabolism in airway inflammation & asthma

• Biological effector functions of lysophosphatidic acid (LPA) can be attributed to at least six LPA receptors (LPA_1–6) that have overlapping specificities and varying tissue distributions.

• The specific roles of these G-protein coupled receptors in asthma pathogenesis in both human subjects and mouse models remain to be fully delineated.

• No data are available comparing LPA receptor binding and signaling for saturated versus monounsaturated versus polyunsaturated LPAs; it is possible that the degree of LPA poly-unsaturation may affect its affinity and specificity for its various receptors on airway epithelial, smooth muscle and inflammatory cells in asthma.

• The metabolic pathways and specific enzymes involved in generating polyunsaturated LPA molecular species in the asthmatic lung are unresolved and an important area for future research to identify novel targets for asthma therapeutics.

Activities of LPA on airway epithelial cells, smooth muscle cells, T lymphocytes & dendritic cells

• LPA has potent activities in vitro on most lung cells thought to be involved in airway inflammation in asthma including airway epithelial cells, smooth muscle cells, T-lymphocytes and dendritic cells.

• Whether and how the activities of LPA on resident lung, inflammatory and immune cells translate in vivo in terms of contributions to asthma pathogenesis and pathophysiology is an important subject for continued investigation.

• Expression, specificity, signaling pathways and roles of the various LPA receptors on resident and airway inflammatory cells remain to be fully defined, both at baseline and during allergic airway inflammation.

Polyunsaturated molecular species of LPA as biomarkers of allergic airway inflammation in asthma

• Airway allergen challenge in patients with allergic asthma results in a remarkable increase in alveolar airway levels of total LPA, comprised principally of two LPAs containing polyunsaturated fatty acids (C22:5 and C22:6).

• Enrichment of these polyunsaturated LPA molecular species is particularly notable, since saturated and monounsaturated LPAs (C16:0, C18:0 and C18:1) are only minimally induced, and they are not induced systemically, in other words, in plasma.

• Inducible increases in polyunsaturated LPAs at sites of allergic airway inflammation in asthma subjects cannot be explained by plasma leak during the allergic response.

• Preliminary study suggests that Increased polyunsaturated LPAs in the airways are not reflected in exhaled breath condensates, with poor correlation between capture and/or detection in exhaled breath condensates compared with BAL, an issue requiring further investigation in a larger cohort of asthma subjects of varying disease phenotype and severity.

• Accumulation of these unusual polyunsaturated LPAs in the alveolar airspace following allergen challenge suggests they could be useful biomarkers in certain asthma phenotypes for tracking disease activity and severity, an area ripe for future investigation.

Future perspective

• Identification of novel lipid mediators or other noninvasive biomarkers useful for determining disease phenotype, status and severity, and for following patient responses to treatment, continues to be a major challenge in the asthma field.

• Advances in biotechnology and ‘Omics’ should enable basic, translational and clinical researchers to identify and validate the utility of novel biomarkers such as polyunsaturated LPAs in asthma for these purposes.

• Validation studies, and translation of the present findings of polyunsaturated species of LPA, their LPA receptors, and the contribution of enzymes involved in LPA metabolism to the asthma diathesis should further advance disease phenotyping, and the development of personalized patient care and novel therapies.