A novel, orally active LPA₁ receptor antagonist inhibits lung fibrosis in the mouse bleomycin model

Abstract

Background and purpose: The aim of this study was to assess the potential of an antagonist selective for the lysophosphatidic acid receptor, LPA₁, in treating lung fibrosis We evaluated the in vitro and in vivo pharmacological properties of the high affinity, selective, oral LPA₁-antagonist (4′-{4-[(R)-1-(2-chloro-phenyl)-ethoxycarbonylamino]-3-methyl-isoxazol-5-yl}-biphenyl-4-yl)-acetic acid (AM966).

Experimental approach: The potency and selectivity of AM966 for LPA₁ receptors was determined in vitro by calcium flux and cell chemotaxis assays using recombinant and native cell cultures. The in vivo efficacy of AM966 to reduce tissue injury, vascular leakage, inflammation and fibrosis was assessed at several time points in the mouse bleomycin model.

Key results: AM966 was a potent antagonist of LPA₁ receptors, with selectivity for this receptor over the other LPA receptors. In vitro, AM966 inhibited LPA-stimulated intracellular calcium release (IC₅₀= 17 nM) from Chinese hamster ovary cells stably expressing human LPA₁ receptors and inhibited LPA-induced chemotaxis (IC₅₀= 181 nM) of human IMR-90 lung fibroblasts expressing LPA₁ receptors. AM966 demonstrated a good pharmacokinetic profile following oral dosing in mice. In the mouse, AM966 reduced lung injury, vascular leakage, inflammation and fibrosis at multiple time points following intratracheal bleomycin instillation. AM966 also decreased lactate dehydrogenase activity and tissue inhibitor of metalloproteinase-1, transforming growth factor β1, hyaluronan and matrix metalloproteinase-7, in bronchoalveolar lavage fluid.

Conclusions and implications: These findings demonstrate that AM966 is a potent, selective, orally bioavailable LPA₁ receptor antagonist that may be beneficial in treating lung injury and fibrosis, as well as other diseases that are characterized by pathological inflammation, oedema and fibrosis.

Abbreviations

BALF

bronchoalveolar lavage fluid

BID

twice a day

CHO

Chinese hamster ovary

CTGF

connective tissue growth factor

DMEM

Dulbecco's modified Eagle's medium

DMSO

dimethyl sulfoxide

ECM

extracellular matrix

EMEM

Eagle's minimum essential medium

FBS

foetal bovine serum

H&E

haematoxylin and eosin

HBSS

Hank's buffered salt solution

HEK

human embryonic kidney

IPF

idiopathic pulmonary fibrosis

LDH

lactate dehydrogenase

LPA

lysophosphatidic acid

MMP-7

matrix metalloproteinase-7

PBS

phosphate buffered saline

PDGF

platelet derived growth factor

TGFβ-1

transforming growth factor β-1

TIMP-1

tissue inhibitor of metalloproteinase-1

UUO

unilateral ureteral obstruction

Introduction

Idiopathic pulmonary fibrosis (IPF) is one of the most common forms of interstitial lung disease and is characterized by alveolar damage and exaggerated scar tissue production that results in fibrous obliteration of lung parenchyma and subsequent lung dysfunction (Wilson and Wynn, 2009). Currently, 5 million people worldwide are affected by IPF with over 200 000 patients in the United States and the median survival time for IPF patients is approximately 3–5 years from the time of diagnosis (Raghu et al., 2006). Despite its prevalence, there are currently no therapies available to halt or reverse the progression of IPF and there are no Food and Drug Administration-approved courses of treatment. Thus, there is a large, unmet medical need for new therapeutic strategies to treat IPF and other diseases that involve pathological tissue fibrosis.

IPF is believed to result from an abnormal wound-healing response following alveolar injury. In this setting, the primary effector cell, the myofibroblast, produces excess scar tissue which leads to dramatic changes in lung architecture and deleterious effects on lung function (Selman and Pardo, 2002). Myofibroblasts are phenotypically differentiated mesenchymal cells that contain a contractile apparatus of α-smooth muscle actin-containing microfilaments and exhibit increased collagen production capacity (Swaney et al., 2005). Although myofibroblasts were initially believed to derive solely from resident lung fibroblasts (Gabbiani et al., 1971), recent studies have suggested that myofibroblasts can also originate from bone marrow-derived fibroblast precursor cells called ‘fibrocytes’ (Andersson-Sjoland et al., 2008) and from mesenchymal transformation of alveolar epithelial cells (Willis et al., 2005; Tanjore et al., 2009). Regardless of their cellular origin, myofibroblasts are the main effector cell in IPF and other fibrotic diseases (Scotton and Chambers, 2007). The formation and the subsequent fibrogenic capacity of myofibroblasts are regulated by mechanical stimuli, as well as by the biochemical milieu (Tomasek et al., 2002). In response to tissue injury, elevated local and/or circulating concentrations of growth factors, such as platelet-derived growth factor (PDGF) and cytokines [transforming growth factor β (TGFβ) and interleukins] promote myofibroblast transformation of precursor cell populations with subsequent increases in tissue fibrosis (Wynn, 2008).

Lysophosphatidic acid (LPA) is a bioactive lysophospholipid that regulates numerous aspects of cellular function and has been recognized as a novel mediator of wound healing and tissue fibrosis (Watterson et al., 2007). LPA signals through a family of at least five G protein-coupled receptors, designated LPA_1–5 (nomenclature follows Alexander et al., 2009; Anliker and Chun, 2004). Several studies have recently linked LPA₁ receptors to the pathogenesis of lung and kidney fibrosis and identified these receptors as a potential clinical target for IPF (Ley and Zarbock, 2008; Tager et al., 2008). Tager et al. showed that LPA levels were increased in the bronchoalveolar lavage fluid (BALF) of IPF patients. In addition, LPA₁ receptors were the most highly expressed LPA receptors in lung fibroblasts isolated from IPF patients and BALF-stimulated chemotaxis by these cells was inhibited by an LPA_1,3 receptor-selective antagonist. Consistent with these findings, LPA₁ receptor null mice were protected from bleomycin-induced lung fibrosis via a reduction in vascular leakage and fibroblast recruitment (Tager et al., 2008). In a renal fibrosis model, tubulointerstitial fibrosis was reduced in LPA₁ receptor knockout mice following unilateral ureteral obstruction (UUO) (Pradere et al., 2007). In this same study, an LPA_1,3 receptor antagonist recapitulated the decrease in renal fibrosis and inhibited mRNA expression of TGFβ and connective tissue growth factor (CTGF) (Pradere et al., 2007), suggesting that these fibrogenic factors act downstream of LPA₁ and/or LPA₃ receptors.

In this study, we characterized the in vitro and in vivo pharmacological properties of an oral, high affinity, selective LPA₁ receptor antagonist (AM966) and assessed the efficacy of AM966 in a mouse model of lung fibrosis induced by bleomycin. Furthermore, we examined the ability of AM966 to modulate cytokine and protease concentrations in the lung in order to elucidate the potential mechanism by which LPA₁ receptor antagonism inhibited lung inflammation and fibrosis. Here we report that AM966 exhibited good oral bioavailability and pharmacokinetics in rats and mice. AM966 reduced tissue injury, vascular leakage, inflammation and fibrosis at several time points in the bleomcyin lung fibrosis model and decreased concentrations of several pro-fibrotic and pro-inflammatory cytokines in BALF. These data establish the LPA₁ receptor antagonist as a potential anti-fibrotic therapy and pave the way for the clinical development of ‘first in class’ molecules that treat IPF via blockade of LPA₁ receptors.

Methods

Cells/recombinant expression

A2058 human melanoma cells were obtained from ATCC and cultured in Dulbecco's modified Eagle's medium (DMEM) containing 10% FBS. IMR-90 human foetal lung fibroblasts were obtained from ATCC and cultured in Eagle's minimum essential medium (EMEM) containing 10% foetal bovine serum (FBS). B103 rat neuroblastoma cells were obtained from Jerold Chun at The Scripps Research Institute and cultured in DMEM containing 10% FBS. Human and mouse LPA₁ and human LPA₃ receptors were stably expressed in Chinese hamster ovary (CHO) cells and cultured in F12 media containing 10% FBS and 1 mg/mL hygromycin B. Mouse LPA₃ receptors were stably expressed in human embryonic kidney (HEK) cells and cultured in DMEM containing 10% FBS and 200 µg/mL hygromycin B. Human and mouse LPA₂ and human and mouse LPA₅ receptors were transiently expressed in rat neuroblastoma B103 cells using Lipofectamine™ 2000 (Invitrogen, Carlsbad, CA, USA) following the manufacturers' instruction. Human LPA₄ receptors were expressed stably in rat neuroblastoma B103. Briefly, on the day before the assay, 30 000–35 000 cells per well were seeded together with 0.2 µL lipofectamine 2000 and 0.2 µg expression vector in 96-well Poly-D-Lysine coated black-wall clear-bottom plates (BD BioCoat, Franklin Lakes, NJ, USA) in DMEM + 10% FBS. Following an overnight culture, cells were washed once with phosphate buffered saline (PBS) then cultured in serum-free media for 4 h prior to dye loading.

Calcium flux

Stably expressing cells were plated in 96-well Poly-D-Lysine coated black-wall clear-bottom plates (BD BioCoat) at a density of 20 000–40 000 cells per well and cultured overnight in complete media. The following day, cells were washed once with PBS then cultured in 75 µL serum-free media either overnight (for the stably expressing cells) or 4 h (for the transient tranfectants) prior to dye loading. On the day of the assay, cells were loaded for 1 h at 37°C with 100 µL FLIPR Calcium 4 dye (Molecular Devices, Sunnyvale, CA, USA) prepared in Hank's buffered salt solution (HBSS) supplemented with 20 mM HEPES (pH 7.4), 2 mM probenecid and 0.3% fatty-acid-free human serum albumin. 25 µL test compounds (prepared in 1% DMSO) were added to each well and incubated at room temperature for 30 min. After 15 s of baseline measurement, 50 µL of 50 nM LPA [prepared in HBSS supplemented with 20 mM HEPES (pH 7.4), and 0.3% fatty-acid-free human serum albumin] was added. Intracellular calcium mobilization was measured using the FLEXstation III (Molecular Devices).

Cell chemotaxis

Neuroprobe ChemoTx^® System plates (8 µm pore size, 5.7 mm diameter sites; Gaithersburg, MD, USA) were coated on both sides with 20 µL of 0.001% fibronectin and allowed to dry. The bottom wells were loaded with 100 nM LPA or vehicle in DMEM containing 0.1% fatty acid free bovine serum albumin (BSA). A2058 human melanoma cells and IMR-90 human lung fibroblasts were serum starved for 24 h, harvested and resuspended in DMEM containing 0.1% fatty acid free BSA. CHO cells over-expressing mouse LPA₁ receptors were serum starved for 24 h, harvested and resuspended in F12K media containing 0.2% fatty acid free BSA. The cells were incubated with inhibitor or vehicle for 15 min at 37°C before applying 25 000 (A2058 and CHO) or 50 000 (IMR-90) cells to the upper portion of the ChemoTx^® plate. The plates were incubated for 3 h (A2058), 4 h (CHO) or 16–18 h (IMR-90) at 37°C. Cells were removed from the upper portion of the filter by rinsing with PBS and scraping. The filter was allowed to dry before staining with the HEMA 3 Staining System (Fisher Scientific, Pittsburg, PA, USA). The absorbance of the filter was then read at 590 nm and calculation of the cell number was made from the standard curve.

Pharmacokinetics

The oral exposure of AM966 was determined in fasted mice. Animals received AM966 (10 mg·kg⁻¹) in vehicle (water) by oral gavage and were then killed by CO₂ inhalation at 1, 2, 4, 8 and 24 h post dose (n= 2 animals per time point for each test compound). Blood (approximately 300 µL) was collected via cardiac puncture into EDTA-containing tubes and centrifuged at 1450×g for 10 min. The plasma was removed and analysed for AM966 content by liquid chromatography-mass spectrometry (LCMS). Briefly, known amounts of AM966 were added to thawed mouse plasma to yield a concentration range from 0.8 to 4000 ng·mL⁻¹. Mouse plasma samples were precipitated using acetonitrile (1:4, v:v) containing the internal standard buspirone. A 10 µL aliquot of the analyte mixture was injected using a Leap PAL autosampler (Leap Technologies, Carrboro, NC, USA). Analyses were performed using an Agilent Zorbax SB-C8 column (Agilent Technologies, Santa Clara, CA, USA) (2.1 × 50 mm; 5 µm) linked to a Shimadzu LC-10AD VP with SCL-10A VP system controller (Shimadzu Scientific Instruments, Colombia, MD, USA). Tandem mass spectrometric detection was carried out on a PE Sciex API3200 (Applied Biosystems, Foster City, CA, USA) in the positive ion mode (ESI) by multiple reaction monitoring. The calibration curves were constructed by plotting the peak–area ratio of analysed peaks against known concentrations. The lower limit of quantitation was 0.8 ng·mL⁻¹. The data were subjected to linear regression analysis with 1/x² weighting.

Bleomycin model

Female C57BL/6 mice (Harlan, San Francisco, CA, USA; 20–25 g) were housed at four per cage, given free access to food and water and allowed to acclimatize for at least 7 days prior to test initiation. After the habituation phase, mice were lightly anaesthetized with isoflurane (5% in 100% O₂) and treated with bleomycin sulfate (Blenoxane, Henry Schein, Melville, NY, USA; 1.5–5 units·kg⁻¹) via intratracheal instillation. Mice were returned to their cages and monitored daily for the duration of the experiment. Vehicle (water), AM966, and Pirfenidone were delivered twice daily (BID) and dexamethasone was delivered once daily (QD) over the course of each study. All compounds were delivered by oral gavage. At 3, 7, 14 or 28 days after bleomycin instillation, animals were killed with inhaled isoflurane. Following death, mice were intubated with a 20 gauge angiocatheter attached to a 1 mL syringe. Lungs were lavaged with saline (2 × 0.75 mL) to obtain BALF and cells and then the lungs were removed and fixed in 10% neutral buffered formalin for subsequent histopathological analysis.

BALF analysis

BALF was centrifuged for 10 min at 800×g to pellet the cells and the BALF supernatant was removed and frozen at −80°C for subsequent protein analysis using the DC protein assay kit (Biorad, Hercules, CA, USA) and soluble collagen analysis. BALF was analysed for concentrations of inflammatory, pro-fibrotic and tissue injury biomarkers including tissue inhibitor of metalloproteinase-1 (TIMP-1), TGFβ1, hyaluronan and matrix metalloproteinase (MMP)-7, and lactate dehydrogenase (LDH) activity, using commercially available kits. The cell pellet was re-suspended in PBS/0.1% BSA. Total cell counts were obtained using a Hemavet haematology system (Drew Scientific, Wayne, PA, USA) and differential cells counts were determined using Shandon cytospin (Thermo Scientific, Waltham, MA, USA). All counts were expressed as cell·mL⁻¹.

Histopathology

Lung tissue was stained using H&E and Masson's Trichrome and the severity of lung fibrosis was determined by semiquantitative histopathological scoring (Ashcroft et al., 1988) using light microscopy (100× magnification). To do this, lung fibrosis was scored on a scale of 0–8 by examining 45 randomly chosen fields (5 lobes / mouse × 3 sections / lobe × 3 fields / section) for each mouse. As defined by Ashcroft et al. (1988) the criteria for grading lung fibrosis were as follows: grade 0 = normal lung; grade 1 = minimal fibrous thickening of alveolar or bronchiolar walls; grade 3 = moderate thickening of walls without obvious damage to lung architecture; grade 5 = increased fibrosis with definite damage to lung structure and formation of fibrous bands or small fibrous masses; grade 7 = severe distortion of structure and large fibrous areas; and grade 8 = total fibrous obliteration of the fields. If there was any difficulty in deciding between two odd-numbered categories, the field was given the intermediate even-numbered score.

Collagen quantification in BALF

Soluble collagen concentrations in BALF were assessed using a modification of previously described methods (Jankov et al., 2002). Briefly, BALF samples (20–50 µL in 1 mL of 0.5 M acetic acid) and collagen type I standards (0–500 µg·mL⁻¹) in 1 mL acetic acid (0.5 M) were added to an equal volume of Sirius Red (direct red 80; 120 µg·mL⁻¹ in 0.5 M acetic acid). Samples were incubated at room temperature for 30 min with gentle agitation, centrifuged at 10 000×g for 10 min and the absorbance of 200 µL of sample supernatant was read at 540 nm and plotted against standards with a known concentration of collagen type I.

Statistical analyses

Values are expressed as mean ± standard error of the mean. Groups were compared using one- and two-way analysis of variance with Bonferroni and Dunnett's, post hoc, multiple comparison tests. Kaplan-Meier curves were compared using a log-rank (Mantel–Cox) test. Statistical significance was set at P < 0.05. Statistical comparisons and graphical representations were performed using GraphPad Prism 5.0 (GraphPad Software, San Diego, CA, USA).

Materials

LPA (18:1 LPA 1-oleoyl-2-hydroxy-sn-glycero-3-phosphate, sodium salt) was purchased from Avanti Polar Lipids (Alabaster, AL, USA). AM966 and pirfenidone (5-methyl-1-phenylpyridin-2-one) were synthesized at Amira Pharmaceuticals (San Diego, CA, USA). Pirfenidone synthesized at Amira was of identical chemical purity to material obtained commercially from 3B Pharmachem (Wuhan, Libertyville, IL, USA) International Inc., as determined by NMR and LCMS. Bleomycin sulphate (Blenoxane) was purchased from Henry Schein (Melville, NY). Total TGFβ₁ and TIMP-1 kits were obtained from R&D Systems (Minneapolis, MN, USA). Hyaluronan kits were obtained from Echelon Biosciences (Salt Lake City, UT, USA). MMP-7 kits were obtained from USCN Life Science Inc. (Wuhan, China). LDH activity assay kits, dexamethasone, direct red 80 and rat tail collagen were purchased from Sigma Aldrich (St. Louis, MO, USA).

Results

AM966 is a selective LPA₁ receptor antagonist

AM966 (Figure 1A) was synthesized and its potency for LPA₁ receptors and selectivity for the different LPA receptors was determined by Ca flux assays using cells over-expressing human and mouse LPA_1–5 receptors. AM966 inhibited LPA-stimulated intracellular calcium release from CHO cells stably expressing human and mouse LPA₁ receptors (Figure 1B; Table 1). As shown in Table I, AM966 was 100-fold and 10-fold more selective in human and mouse cells, respectively, for LPA₁ relative to the other LPA receptors. AM966 was also evaluated for inhibition of LPA-induced chemotaxis in A2058 human melanoma cells and IMR-90 human lung fibroblasts endogenously expressing LPA₁ receptors and in CHO cells stably expressing mouse LPA₁ receptors. As shown in Figure 1C, AM966 inhibited LPA₁-mediated chemotaxis of human A2058 melanoma cells (IC₅₀= 138 ± 43 nM), IMR-90 human lung fibroblasts (IC₅₀= 182 ± 86 nM) and CHO mLPA₁ cells (IC₅₀= 469 ± 54 nM).

Figure 1

Pharmacological characterization of AM966 as an LPA₁ receptor antagonist. (A) Structure of AM966. (B) Inhibition of LPA-stimulated intracellular calcium release from CHO cells stably expressing the mouse (mLPA1) and human LPA₁ (hLPA1) receptors. CHO cells were pre-treated with increasing concentrations of AM966 for 30 min and then stimulated with LPA (10–30 nM) and calcium release was measured. (C) AM966-mediated inhibition of cell chemotaxis of human A2058 melanoma cells (IC₅₀= 138 nM), IMR-90 human lung fibroblasts (IC₅₀= 181 nM) and CHO mLPA₁ cells (IC₅₀= 469 nM). LPA, lysophosphatidic acid; CHO, Chinese hamster ovary.

Table 1. Inhibition of LPA-stimulated intracellular calcium release from cells expressing the human or mouse LPA_1–5 receptors

	AM966 IC50 (µM)
	LPA1	LPA2	LPA3	LPA4	LPA5
Human	0.017 ± 0.002 (n= 8)	1.7 (n= 1)	1.6 ± 0.1 (n= 12)	7.7 (n= 1)	8.6 (n= 2)
Mouse	0.019 ± 0.002 (n= 6)	∼25 (n= 1)	0.17 ± 0.03 (n= 8)	ND	∼23 (n= 1)

• Data represent the mean ± standard error of the mean.
• LPA, lysophosphatidic acid; ND, not determined.

Time- and dose-dependent effects of bleomycin on lung markers of injury and fibrosis

In order to characterize the time- and dose-dependent effects of bleomycin on biochemical markers of lung injury and fibrosis, we measured soluble collagen, LDH activity and TIMP-1, TGFβ1, hyaluronan and MMP-7 concentrations in BALF at several points (3, 7 and 14 days) following intratracheal instillation of bleomycin (1.5, 3, and 5 units·kg⁻¹). BALF collagen concentrations were increased at all times and in response to all doses of bleomycin (Figure 2A). Analysis of LDH activity revealed very little change at the 3 day point, except for the 5 units·kg⁻¹ bleomycin dose (Figure 2B). However, by days 7 and 14, LDH activity was increased at all bleomycin doses and showed an approximately 15-fold elevation in activity by day 14. The bleomycin-induced changes in soluble collagen prompted us to examine BALF concentrations of several well-known pro-fibrotic mediators. Among these factors, the matrix metalloproteinase inhibitor, TIMP-1, was elevated in the early window at days 3 and 7 following bleomycin injury (Figure 2C) but decreased dramatically by day 14. In contrast, TGFβ1, hyaluronan and MMP-7 were only slightly elevated on day 3 but dramatically increased between days 7 and 14 (Figure 2D–F). Notably, 25% mortality was observed at the high (5.0 units·kg⁻¹) bleomycin dose, whereas all mice survived at the lower (1.5 and 3.0 units·kg⁻¹) doses in this study. These data demonstrate that several biochemical mediators of tissue inflammation and fibrosis are elevated in this model and these factors correlate with increased lung injury and fibrosis following bleomycin administration.

Figure 2

Time course of inflammatory, fibrotic and cell death markers in the mouse bleomycin model. Effects of bleomycin on (A) soluble collagen (B) LDH activity (cell death marker) and (C) TIMP-1 (D) TGFβ1 (E) hyaluronan and (F) MMP-7 concentrations in BALF at several time points (3, 7 and 14 days) following intratracheal instillation of bleomycin (1.5, 3, and 5 units·kg⁻¹). Data represent the mean ± standard error of the mean of n≥ 4 mice per group. * denotes a significant increase (P < 0.05) compared to vehicle at each time point. LDH, lactate dehydrogenase; TGFβ-1, transforming growth factor β-1; TIMP-1, tissue inhibitor of metalloproteinase-1; MMP-7, matrix metalloproteinase-7; BALF, bronchoalveolar lavage fluid; Veh, vehicle.

AM966 demonstrates a good pharmacokinetic profile in vivo

AM966 was given orally to non-fasted mice at 10 mg·kg⁻¹ and plasma concentrations were measured over a 24 h period (Figure 3). AM966 peaked within 1 h to a concentration of 9 µM and decreased to 10 nM over a 24 h period. The dashed line in Figure 3 represents the IC₅₀ value of 469 nM for AM966-mediated inhibition of chemotaxis by CHO cells expressing recombinant mouse LPA₁ (see Figure 1C). Based on these data, a single 10 mg·kg⁻¹ oral dose of AM966 was not sufficient to maintain plasma drug concentrations above the IC₅₀ for chemotaxis for 24 h. Therefore, in all bleomycin studies we used twice a day (BID) dosing to attain full drug coverage over a 24 h period.

Figure 3

Time-concentration profile for AM966. Fasted mice (n= 2) were dosed with AM966 and blood was sampled at 1, 2, 4, 8 and 24 h after oral dosing (10 mg·kg⁻¹). The dashed line in Figure 3 represents the IC₅₀ value for AM966-mediated inhibition of cellular chemotaxis by Chinese hamster ovary cells expressing recombinant mouse LPA₁ receptors (See Figure 1D).

The LPA₁ receptor antagonist, AM966, reduces vascular leakage, fibrosis and inflammation in the acute (3–7 day) window after bleomycin lung injury

Increases in vascular permeability and inflammation are observed as early as day 3 after bleomycin lung injury (Izbicki et al., 2002; Tager et al., 2008) and by day 7, tissue fibrosis is detectable (Cuzzocrea et al., 2007). As early as day 3 (Figure 4), AM966 decreased (P < 0.05) BALF protein concentrations in a dose-dependent manner, resulting in 43% reduction at the 30 mg·kg⁻¹ dose (Figure 4A). This reduction is comparable with that observed in the LPA₁ receptor knockout mouse (Tager et al., 2008), demonstrating full antagonism of LPA₁ receptors by AM966 at 30 mg·kg⁻¹. A similar reduction in BALF cells was observed in response to AM966, albeit this decrease was not statistically significant (Figure 4B). AM966 also inhibited the bleomycin-induced lung damage as demonstrated by the decrease in LDH activity at 10 and 30 mg·kg⁻¹ (Figure 4C). Based on these findings at day 3, a 7 day study was conducted in which the dose range was expanded to include a 1 mg·kg⁻¹ AM966 dose (Figure 5). Consistent with the 3 day results, AM966 significantly reduced BALF protein at the 10 and 30 mg·kg⁻¹ doses (Figure 5A). In addition, AM966 reduced BALF collagen and total TGFβ concentrations (Figure 5B,C) in the lung. In order to assess the effects of AM966 on tissue fibrosis at day 7, mouse lungs were stained with Trichrome (Figure 5D). Compared to vehicle controls (Figure 5D, panel a.), increased tissue remodelling and fibrosis were observed in bleomycin-instilled mice (Figure 5D, panel b.). In contrast, lungs from mice treated with AM966 (30 mg·kg⁻¹) exhibited a decrease in tissue fibrosis and maintenance of normal lung architecture (Figure 5D, panel c.). Furthermore, the decrease in lung leakage and fibrosis at day 7 was accompanied by a lowering of total BALF cells, with this reduction comprised primarily of macrophages and neutrophils (Figure 5E). Thus, LPA₁ receptor antagonism decreases vascular permeability, inflammation and fibrosis in the early (3–7 day) window after bleomycin-induced lung injury.

Figure 4

AM966 reduces vascular leakage, inflammation and lung injury and inflammation in a 3 day bleomycin model. Mice were given intratracheal bleomycin sulfate (BLM; 1.5 units·kg⁻¹) or saline vehicle (Veh.), followed by oral (gavage) administration of AM966 (10 and 30 mg·kg⁻¹, BID) for a period of 3 days. BALF was isolated and analysed for changes in (A) total protein, (B) inflammatory cells, and (C) LDH activity. Data represent the mean ± standard error of the mean of n= 5–7 mice per group. # denotes a significant increase (P < 0.05) compared to vehicle control. * denotes a significant decrease (P < 0.05) compared to BLM. BID, twice a day; BALF, bronchoalveolar lavage fluid; LDH, lactate dehydrogenase.

Figure 5

AM966 reduces vascular leakage and fibrosis in a 7 day bleomycin model. Mice were given intratracheal bleomycin sulfate (BLM; 3.0 units·kg⁻¹) or saline vehicle (Veh.), followed by oral (gavage) administration of AM966 (1, 10 and 30 mg·kg⁻¹, BID) for a period of 7 days. BALF was isolated and analysed for changes in (A) total protein, (B) collagen, (C) total (TGFβ₁). (D) Representative histopathological images (100× magnification; trichrome staining) are shown of the lungs of mice treated with (a) vehicle, (b) BLM or (c) BLM + AM966 (30 mg·kg⁻¹, BID). Blue staining denotes regions of fibrosis. (E) Total cellularity was analysed using a Hemavet system and differential cell counts for macrophages, neutrophils and lymphocytes were measured by cytospin in BALF. Data represent the mean ± standard error of the mean of n= 8 mice per group. # denotes a significant increase (P < 0.05) compared to vehicle control. * denotes a significant decrease (P < 0.05) compared to BLM. BID, twice a day; BALF, bronchoalveolar lavage fluid.

AM966 inhibits lung fibrosis, maintains mouse body weight and decreases lung inflammation 14 days after bleomycin lung injury

To assess the therapeutic potential of LPA₁ antagonism in treating lung fibrosis at later time points, we examined the anti-fibrotic activity of AM966 14 days after bleomycin (1.5 unit·kg⁻¹) lung injury. Day 14 after a single intratracheal instillation of bleomycin is reported to be the most appropriate time point for assessing lung fibrosis in this model (Izbicki et al., 2002). Histological examination of mouse lungs following bleomycin instillation revealed significant damage to the lung tissue, including thickening of the alveolar septae, alveolar inflammation and fibrous obliteration of the peribronchiolar and parenchymal regions (Figure 6A, panel b.), compared to the vehicle group (Figure 6A, panel a.). No reduction in lung fibrosis was observed in response to either the dexamethasone (Figure 6A, panel c) or low dose AM966 (10 mg·kg⁻¹; Figure 6A, panel d.). However, AM966 at 30 and 60 mg·kg⁻¹ dramatically reduced lung tissue remodelling and fibrosis (Figure 6A, panels e–f.) so that lung architecture in these groups was similar to that of the vehicle group (Figure 6A, panel a.). The reduction in fibrosis was confirmed by quantitative analysis of BALF collagen concentrations (Figure 6B) and histopathological scoring of lung fibrosis (Figure 6C), according to Ashcroft et al. (1988). Consistent with the reduction in lung pathology, the bleomycin-induced body weight loss was inhibited in the 30 and 60 mg·kg⁻¹ treatment groups (Figure 6D) demonstrating an enhanced overall state of health.

Figure 6

AM966 inhibits lung fibrosis and maintains body weight 14 days after bleomycin-induced lung injury. Mice were given intratracheal bleomycin sulfate (BLM; 1.5 units·kg⁻¹) or saline vehicle (Veh.), followed by oral (gavage) administration of dexamethasone (1 mg·kg⁻¹, QD) or AM966 (10, 30 and 60 mg·kg⁻¹, BID) for a period of 14 days. (A, a-f) Representative histopathological images (100× magnification; Trichrome staining) are shown of the lungs of mice treated with (a) vehicle, (b) BLM, (c) BLM + dexamethasone (1 mg·kg⁻¹, QD), (d) BLM + AM966 (10 mg·kg⁻¹, BID), (e) BLM + AM966 (30 mg·kg⁻¹, BID) or (f) BLM + AM966 (60 mg·kg⁻¹, BID). (B) Soluble collagen in BALF. (C) Histopathological scoring of lung fibrosis. (D) Percentage change in mouse body weight (wt.). (E) Total cellularity and differential cell counts for macrophages, neutrophils and lymphocytes. Data represent the mean ± standard error of the mean of n= 8 mice per group. # denotes a significant increase (P < 0.05) compared to vehicle control. * denotes a significant decrease (P < 0.05) compared to BLM. QD, once a day; BID, twice a day; BALF, bronchoalveolar lavage fluid.

Pulmonary inflammation is a precipitating or complicating component of many fibrotic lung diseases and very few clinical treatment strategies address both the inflammatory and fibrotic components of IPF. For this reason, we examined the role of LPA₁ receptors in inflammatory cell infiltration in the lungs after bleomycin injury (Figure 6E). Similar to previous reports, dexamethasone inhibited lung inflammation, as indicated by a reduction in the total number of inflammatory cells, as well as specific decreases in macrophages, neutrophils and lymphocytes in the BALF. The fact that dexamethasone had an anti-inflammatory effect but did not affect lung fibrosis suggests that, at least in our hands, decreasing inflammation is not sufficient to decrease fibrosis. AM966 at 10 mg·kg⁻¹ had no effect on total cells, macrophages, neutrophils or lymphoctyes. In contrast, at the 30 and 60 mg·kg⁻¹ doses, AM966 reduced total and differential cell counts to a greater extent than dexamethasone. These data demonstrate that LPA₁ receptor antagonism limits the infiltration of inflammatory cells in the lung following tissue injury.

AM966 reduces vascular leakage, tissue injury and pro-fibrotic cytokine production in the 14 day bleomycin study

Mice lacking LPA₁ receptors exhibit decreased pulmonary vascular leakage following bleomycin injury, which is believed to contribute to a reduction in lung damage (Tager et al., 2008). In order to understand the mechanism by which LPA₁ receptor blockade inhibits lung injury and fibrosis (see 4-6), we examined the effects of AM966 on BALF protein, LDH activity and BALF concentrations of various inflammatory and pro-fibrotic mediators. In the 14 day bleomycin model, AM966 attenuated vascular leakage, as shown by a significant (P < 0.05) reduction in BALF protein (Figure 7A). In addition, AM966-treated animals showed decreased LDH activity in BALF (Figure 7B), indicative of a decrease in lung tissue injury. AM966 significantly reduced BALF concentrations of TIMP-1, TGFβ1, hyaluronan and MMP-7 (Figure 7C–F), factors that are linked to lung inflammation and fibrosis (Bjermer et al., 1989; Broekelmann et al., 1991; Swiderski et al., 1998; Rosas et al., 2008). Dexamethasone (1 mg·kg⁻¹) had no effect on BALF protein or LDH activity and actually increased BALF concentrations of TIMP-1, TGFβ1 and hyaluronan (Figure 7A–E). However, a reduction in MMP-7 by dexamethasone was observed (Figure 7F). When combined with the 3 and 7 day bleomycin data (Figure 4), these findings provide mechanistic insight into the ability of AM966 to attenuate lung tissue damage, inflammation and fibrosis at multiple time points after lung insult.

Figure 7

AM966 reduces vascular leakage, tissue injury and pro-fibrotic cytokine production. Mice were treated in the same manner as described for Figure 6. (A) Total protein (B) LDH activity (cell death marker) and concentrations of the pro-fibrotic factors (C) TIMP-1 (D) TGFβ1 (E) hyaluronan and (F) MMP-7 were analysed using commercially available kits. Data represent the mean ± standard error of the mean of n= 8 mice per group. # denotes a significant increase (P < 0.05) compared to vehicle control. Φ denotes a significant (P < 0.05) increase compared to BLM. * denotes a significant decrease (P < 0.05) compared to BLM. LDH, lactate dehydrogenase; TIMP-1, tissue inhibitor of metalloproteinase-1; TGFβ-1, transforming growth factor β-1; MMP-7, matrix metalloproteinase-7; BLM, bleomycin sulfate.

AM966 demonstrates greater efficacy compared to pirfenidone in the 14 day bleomycin model

Pirfenidone is an anti-fibrotic agent that has recently been approved for IPF in Japan and is under regulatory review in the United States. Therefore, we compared the efficacy of AM966 (30 mg·kg⁻¹, BID) to that of pirfenidone (20, 100 and 400 mg·kg⁻¹, BID) 14 days after bleomycin (3.0 units·kg⁻¹) challenge. As shown in Figure 8, lung fibrosis (Figure 8A, panel c.) was once again reduced by AM966, compared to the bleomycin-only group (Figure 8A, panel b.). Remarkably, no therapeutic benefit was observed in pirfenidone-treated mice (Figure 8A, panels d–f.). Consistent with these findings, both BALF collagen concentrations (Figure 8B) and tissue fibrosis scores (Figure 8C) were significantly reduced by AM966, whereas pirfenidone had no effect when compared to mice treated with bleomycin alone. Although the mouse body weight loss in the AM966 group was similar to that observed at this increased dose of bleomycin (3.0 unit·kg⁻¹; see Figure 6D for comparison), a decrease in body weight was seen in response to the 100 mg·kg⁻¹ and 400 mg·kg⁻¹ pirfenidone doses (Figure 8D). This same trend was observed for BALF protein, LDH activity and TIMP-1, which were all decreased by AM966, whereas pirfenidone showed little or no therapeutic benefit (Figure 8E–G). Thus, AM966 demonstrates greater efficacy than pirfenidone at reducing lungtissue damage and fibrosis in a 14 day mouse bleomycin model.

Figure 8

AM966 demonstrates greater efficacy than pirfenidone after i.t. instillation of bleomycin. Mice were given intratracheal bleomycin sulfate (BLM; 3.0 units·kg⁻¹) or saline vehicle (Veh.) followed by oral administration of AM966 (30 mg·kg⁻¹, BID) or pirfenidone (20, 100 and 400 mg·kg⁻¹, BID) for a period of 14 days. Representative (A, panels a–f.) Trichrome-stained images (100× magnification) of lungs from mice treated with (a) vehicle, (b) BLM, (c) BLM + AM966 (30 mg·kg⁻¹, QD), (d) BLM + pirfenidone (20 mg·kg⁻¹, BID), (e) BLM + pirfenidone (100 mg·kg⁻¹, BID) and (f) BLM + pirfenidone (400 mg·kg⁻¹, BID). (B) soluble collagen concentrations in BALF (C) Histopathological scoring of lung fibrosis (D) Percentage change in mouse body weight (E) total protein in BALF (F) LDH activity and (G) TIMP-1 concentration are also shown. Data represent the mean ± standard error of the mean of n= 4–6 mice per group. # denotes a significant increase (P < 0.05) compared to vehicle control. * denotes a significant decrease (P < 0.05) compared to BLM. QD, once a day; BID, twice a day; BALF, bronchoalveolar lavage fluid; LDH, lactate dehydrogenase; TIMP-1, tissue inhibitor of metalloproteinase-1.

AM966 decreases mortality and fibrosis at late time points after bleomycin injury

To assess the beneficial effects of LPA₁ receptor antagonism on survival, mice received an intratracheal instillation of 5 units·kg⁻¹ of bleomycin and were treated with AM966 (30 mg·kg⁻¹, BID) for a period of 21 days (Figure 9A). In response to bleomycin alone, 60% of the mice were dead by day 12 with majority of the mortalities occurring between days 7 and 9. In the AM966-treated group, only 2 out of 10 mice died and the first death did not occur until day 12, demonstrating an increase in both the magnitude and duration of the survival effect. In a separate, longer term study, AM966 decreased BALF collagen concentrations as late as day 28 after bleomycin (3.0 unit·kg⁻¹) injury (Figure 9B).

Figure 9

Effects of AM966 on mouse survival and lung fibrosis 21–28 days after BLM administration. (A) AM966 (30 mg·kg⁻¹) significantly (P= 0.04) increased mouse survival over a 21 day period following i.t. instillation of BLM (5.0 units·kg⁻¹). (B) AM966 (30 mg·kg⁻¹) significantly (P < 0.05) decreased BALF collagen 28 days after i.t. instillation of BLM (3.0 units·kg⁻¹). # denotes a significant increase (P < 0.05) compared to vehicle control. * denotes a significant decrease (P < 0.05) compared to BLM. BLM, bleomycin sulfate; BALF, bronchoalveolar lavage fluid.

Discussion

IPF is a devastating and grievous condition for which the only beneficial treatment modality is lung transplantation. Although a number of potential therapeutic targets are being investigated, no approved course of treatment currently exists that halts the progression of IPF and prolongs survival (Scotton and Chambers, 2007). As a result, there is a large unmet medical need for the identification of novel therapeutic strategies for treating IPF. Recently, some compelling findings by Tager et al. (2008) shed new light on this therapeutic area and identified LPA as a novel lysophospholipid mediator of pulmonary vascular leakage and fibroblast migration associated with IPF. Specifically, analysis of LPA receptor expression levels in IPF patients and preclinical studies using a mouse LPA₁ receptor knockout mouse identified the signalling through LPA₁ receptors as a potential pathogenic pathway in IPF. In addition to the fibrogenic role of LPA₁ receptors in the lung, Pradere et al. (2007) demonstrated a pathological role for these receptors in renal fibrosis. Using a UUO model, this group showed a reduction in tubulointerstitial fibrosis in LPA₁ knockout mice. In addition, administration of an LPA_1,3 selective antagonist (Ki16425) reduced kidney fibrosis and the expression of TGFβ and CTGF after UUO in wild-type mice. Thus, these and other studies have highlighted the role of LPA signalling in fibrotic disease and have identified the LPA₁ receptor as a therapeutic target for treating lung and kidney fibrosis. For this reason, we developed a novel LPA₁ receptor selective antagonist (AM966) and used the mouse bleomycin model to evaluate the potential of AM966 as a treatment for IPF.

Prior to initiating AM966 efficacy studies, we analysed the time- and dose-dependent effects of bleomycin on several BALF biomarkers that have been correlated with the severity of IPF or are known to be regulated by or exhibit crosstalk with LPA. LDH is an indicator of tissue damage and TIMP-1, TGFβ1, hyaluronan and MMP-7 are mediators of lung inflammation and fibrosis. These biomarkers are increased in both animal models of lung injury and in association with human lung fibrosis (Bjermer et al., 1989; Broekelmann et al., 1991; Drent et al., 1996; Swiderski et al., 1998; Rosas et al., 2008; Ou et al., 2009). Thus, by characterizing the time-dependent effects of bleomycin on these various markers we provided a framework for our interpretation of the mechanism(s) by which AM966 reduces lung inflammation and fibrosis. Based on the BALF and tissue-specific endpoints, AM966 appears to not only reduce the lung tissue damage, vascular leakage and inflammation that occur in the acute (3–7 day) phase after bleomycin injury, but it also decreases cytokine production and attenuates tissue remodelling and fibrosis in the chronic (14–28 day) phase.

It must be noted that, although AM966 is 100-fold more selective for human LPA₁ receptor over the other LPA receptors based on calcium flux data, the selectivity for mouse LPA₁ receptor is only 10-fold greater than that for mouse LPA₃ receptors. Because the plasma concentrations are above the IC₅₀ for mouse LPA₃ receptors during the first 8 h following a dose of 10 mg·kg⁻¹ (see Figure 3 and Table 1), it is likely that some of the effects observed in this report are mediated by LPA₃ receptors. This is particularly true at the high (30 and 60 mg·kg⁻¹) AM966 doses, for which peak plasma concentrations are predicted to be above the IC₅₀ for mouse LPA₁ and LPA₃ receptors for at least the first 8 h after dosing. Although the anti-fibrotic effects of AM966 are consistent with the findings of Tager et al. using the LPA₁ knockout mice, AM966 treatment produced a much greater reduction in inflammatory cells all time points tested. Interestingly, LPA_1–3 receptors have all been implicated in the regulation of lung inflammation (Saatian et al., 2006). Therefore, the anti-inflammatory effects of AM966 seen in these studies, both in terms of cellular infiltration and cytokine production in the lung, may be mediated in part via a reduction in LPA₃ signalling. Recent studies have also shown that LPA₂ receptors are mediators of lung fibrosis in both mice and humans (Xu et al., 2009) and regulate myofibroblast transformation and TGFβ production by human mesenchymal stem cells in culture (Jeon et al., 2008). The work by Xu et al. expanded upon the work of Tager et al. and suggested that LPA₂ receptors may regulate TGFβ activation and matrix synthesis, whereas LPA₁ receptors promote inflammation and vascular leakage associated with IPF. However, because the IC₅₀ for mouse LPA₂ receptors for AM966 is approximately 25 µM (Table 1), the effects described in our studies are unlikely to result from LPA₂ receptor antagonism. In terms of treating human disease, AM966 is at least 100-fold more selective for human LPA₁ receptors relative to any of the other LPA receptors. Furthermore, to address the issue of receptor selectivity, we are currently developing LPA receptor antagonists with greater selectivity for mouse LPA₁ receptors.

Of the LPA receptor antagonists that have been developed thus far, the LPA_{1, 3} antagonist Ki16425 has shown some therapeutic potential. Ki16426 has been used in a number of in vitro and in vivo studies (Ohta et al., 2003; Boucharaba et al., 2006) and has been shown to reduce renal fibrosis in a mouse UUO model (Pradere et al., 2007). Although the in vitro pharmacokinetic profiles of Ki16425 and AM966 are similar, the in vivo pharmacokinetic characteristics of AM966 are better in that AM966 is orally bioavailable, whereas Ki16425 is not and requires a subcutaneous route of administration.

Mechanistically, AM966 appears to exert its beneficial effects via multiple modes of action. Tissue injury initiates a precise series of molecular and cellular events, including (i) homeostasis; (ii) inflammation; (iii) cellular proliferation, migration and differentiation; (iv) matrix and tissue remodelling; and finally, (v) wound contracture and scar formation. It is widely recognized that many of these events are controlled by the coordinated release of biochemical factors within and around the site of injury (Wynn, 2008). In terms of LPA₁ receptors, it is believed that, following initial tissue injury, LPA is released into the lung interstitium via plasma exudation from the surrounding vascular compartment (Ley and Zarbock, 2008). These increased local LPA concentrations then act on LPA₁ receptors to promote vascular leakage, potentiate epithelial damage and recruit fibroblasts to the region of injury. As shown by our 3 day bleomycin model data, AM966 decreased protein content and LDH activity in the lung, indicating a potential for AM966 to reduce the LPA-mediated effects on vascular leakage and epithelial cell death that occur in IPF and other interstitial lung diseases. As the initial injury progresses, it is hypothesized that pathologically elevated concentrations of LPA may continue to activate LPA₁ receptors on pulmonary cells, thereby potentiating tissue inflammation and stimulating excessive ECM production. By day 7, when both inflammatory cell infiltration and fibrosis are observed in the bleomycin model (Cuzzocrea et al., 2007), AM966 reduced both of these parameters, indicating a continued ability of AM966 to antagonize LPA₁-mediated inflammation and scar tissue production. By day 14, when inflammation is resolving and fibrosis is at its peak, AM966 continues to decrease collagen production as well as inflammation. In addition, AM966 reduces BALF concentrations of TGFβ as well as other cytokines and proteases that are associated with the progression of lung fibrosis (Oga et al., 2009). Although bleomycin injury-induced fibrosis in mice is different from that observed in IPF, the findings of Tager et al. (2008) suggest that LPA₁ receptors are involved in both scenarios. Furthermore, these findings continue to support a very complex inflammatory and fibrotic role for LPA₁ receptors after lung injury and demonstrate that antagonism of LPA₁ receptors is sufficient to modify these pathological endpoints. Finally at days 21–28, although fibrosis is beginning to resolve, a significant reduction in collagen by AM966 was still observed. Furthermore, the combined effects of AM966 throughout this entire wound healing process, from initial injury to wound resolution, manifested itself in increased survival of AM966-treated mice compared to bleomycin controls, over the 21 survival day study.

The histopathological findings in IPF patients are highly varied and retrospective analyses of necropsy samples from IPF patients have revealed clinical features such as diffuse alveolar damage, usual interstitial pneumonia, metaplastic squamous epithelium and inflammatory cell infiltration (Tiitto et al., 2006). Via its pleiotropic effects on tissue damage, inflammation, vascular leakage and fibrosis, LPA may regulate numerous aspects of IPF pathogenesis through LPA₁ receptors. As a result, we believe that AM966 may target multiple aspects of the disease, thereby conferring greater therapeutic efficacy in comparison to approaches targeted on a single mechanism. Plasma exudation into the alveolar space and pulmonary interstitium is believed to be a major source of LPA in the IPF lung. Thus, the AM966-mediated reduction in vascular leakage may diminish the plasma-derived increases in LPA, which help to potentiate IPF disease progression. The ability of AM966 to reduce numerous BALF cytokines provides further mechanistic insight into the pathological role of LPA₁ receptors in lung inflammation and fibrosis. In addition, it may aid in identifying BALF biomarkers that can be used to extrapolate AM966 pre-clinical efficacy to the clinical setting.

One of the most interesting findings of this report was our comparison of AM966 and pirfenidone in the 14 day bleomycin model. Pirfenidone is a broad spectrum anti-fibrotic agent that is currently approved in Japan for the treatment of IPF and is in Phase III clinical trials in the United States for the same indication. Despite the limited data regarding its mode of action, pirfenidone has demonstrated anti-fibrotic effects in cell culture and in a number of pre-clinical lung, liver and kidney fibrosis models (Kakugawa et al., 2004; Oku et al., 2008; Lin et al., 2009; RamachandraRao et al., 2009; Zhao et al., 2009). The anti-fibrotic effects of pirfenidone are attributed to its ability to regulate TGFβ signalling, as well as that of other cytokines, and to scavenge free radicals (Misra and Rabideau, 2000; Oku et al., 2008). Using doses within the published range for pirfenidone in our mouse bleomycin studies, we conducted a direct head-to-head comparison between AM966 and pirfenidone. AM966 demonstrated greater efficacy than pirfenidone at all endpoints tested, whereas pirfenidone showed no therapeutic activity compared to bleomycin alone. This finding was unexpected based on the efficacy of pirfenidone in previously published mouse bleomycin studies (Kakugawa et al., 2004; Oku et al., 2008). In contrast to the data in Figure 6D that demonstrate a maintenance of mouse body weight at the 30 and 60 mg·kg⁻¹ doses, AM966 did not maintain body weight in the studies involving pirfenidone (Figure 8E). This lack of a body weight effect was most likely due to the increased dose of bleomycin (3.0 units·kg⁻¹ vs. 1.5 units·kg⁻¹) that was employed in the pirfenidone studies (Figure 8E), relative to those described in Figure 6D. It must also be noted that in previous studies where pirfenidone was shown to reduce lung fibrosis, bleomycin was administered via i.v. injection over a 5 day period at the beginning of the study (Kakugawa et al., 2004; Oku et al., 2008). This systemic route of administration is sufficient when trying to mimic a chemotherapeutic dosing regime, but requires a long period of time (6–8 weeks) for the development of lung fibrosis (Walters and Kleeberger, 2008). In contrast, our studies employed a single intratracheal instillation of bleomycin on the first day of the study. Intratracheal instillation of bleomycin elicits a peak fibrotic response within 2–3 weeks and produces a more variable distribution of fibrotic foci. Thus, the discrepancy in pirfenidone efficacy between our study and those of other investigators may be due to differences in the route of bleomycin administration and the resulting progression of lung injury and fibrosis.

In conclusion, our findings confirm a role for LPA₁ receptors in inflammation, production of extracellular matrix and tissue remodelling following lung injury. In addition, we demonstrate that antagonizing LPA₁ receptors in the lung with a small molecule has effects consistent with those observed in the LPA₁ receptor knockout mouse. Therefore, the ability of AM966 to modify these pathological endpoints further supports LPA₁ receptors as a novel clinical target in treating IPF and paves the way for the clinical development of an oral, ‘first in class’ LPA₁ receptor antagonist to treat IPF and other lung diseases characterized by pathological inflammation, oedema and fibrosis.

Conflict of interest

All authors of this manuscript are paid employees of Amira Pharmaceuticals.