4,5Dihydro1H-imidazol-2-yl)-[4-(4-isopropoxy-benzyl)-phenyl]-amine (RO1138452) Is a Selective, Pseudo-Irreversible Orthosteric Antagonist at the Prostacyclin (IP)Receptor Expressed

4,5-Dihydro-1H-imidazol-2-yl)-[4-(4-isopropoxy-benzyl)-phenyl]-amine (RO1138452) Is a Selective, Pseudo-Irreversible OrthostericAntagonist at the Prostacyclin (IP)-Receptor Expressed by HumanAirway Epithelial Cells: IP-Receptor-Mediated Inhibition of CXCL9and CXCL10 Release□S

Linda M. Ayer, Sylvia M. Wilson, Suzanne L. Traves, David Proud, and Mark A. GiembyczDepartments of Pharmacology and Therapeutics (L.M.A., S.M.W., M.A.G.) and Physiology and Biophysics (S.L.T., D.P.),Airways Inflammation Group, Institute of Infection, Immunity, and Inflammation, Faculty of Medicine, University of Calgary,Calgary, Alberta, Canada

Received July 27, 2007; accepted October 24, 2007

ABSTRACTThe extent to which the prostacyclin (IP) receptor regulates therelease of two proinflammatory chemokines from human airwayepithelial cells was investigated using the novel and competi-tive IP-receptor antagonist 4,5-dihydro-1H-imidazol-2-yl)-[4-(4-isopropoxy-benzyl)-phenyl]-amine (RO1138452). In BEAS-2Bhuman airway epithelial cells, taprostene, a selective IP-recep-tor agonist, suppressed interferon-�-induced CXCL9 andCXCL10 release in a concentration-dependent manner. Theseeffects were mimicked by 8-bromo-cAMP, and they were abol-ished in cells infected with an adenovirus vector encoding ahighly selective inhibitor of cAMP-dependent protein kinase(PKA). RO1138452 blocked the inhibitory effect of taprosteneon chemokine output in a manner inconsistent with surmount-able competitive antagonism. Comparable results were ob-tained using primary cultures of human airway epithelial cells.The basis of the antagonism imposed by RO1138452 wasstudied further using BEAS-2B cells stably transfected with a

cAMP-response element (CRE) luciferase reporter. On this out-put, RO1138452 also behaved insurmountably. Mechanisti-cally, this could not be attributed to covalent receptor inactiva-tion, allosterism, or a state of hemiequilibrium. Other studiesestablished that the degree by which RO1138452 antagonizedtaprostene-induced CRE-dependent transcription was not re-versed over a 20-h “washout” period. This pharmacologicalprofile is consistent with the behavior of a pseudo-irreversibleantagonist where dissociation from its cognate receptor is soslow that re-equilibration is not achieved at the time the re-sponse is measured. Collectively, these data provide compel-ling evidence that human airway epithelial cells express inhib-itory IP-receptors linked to the activation of PKA. Moreover,contrary to existing literature, RO1138452 behaved pseudo-irreversibly, emphasizing the need, in drug discovery, to screenpotential new medicines in the target tissue(s) of interest.

Prostacyclin (PGI2) is a labile eicosanoid produced by thewall of arterial blood vessels (Bunting et al., 1976), and it isderived from arachidonic acid following the sequential actionof cyclooxygenase and PGI2 synthase (Vane and Botting,1995). Like other prostanoids formed by the cyclo-oxygen-ation of arachidonic acid, the biological actions of PGI2 aremediated through one or more G protein-coupled receptors

M.A.G. is an Alberta Heritage Foundation for Medical Research SeniorScholar and is funded by the Canadian Institutes of Health Research (CIHR).D.P. is the recipient of a Canada Research Chair in Inflammatory AirwayDiseases and is supported by the CIHR. S.L.T. acknowledges Nycomed Canadafor postdoctoral fellowship support.

Article, publication date, and citation information can be found athttp://jpet.aspetjournals.org.

doi:10.1124/jpet.107.129312.□S The online version of this article (available at http://jpet.aspetjournals.org)

contains supplemental material.

ABBREVIATIONS: PGI2, prostacyclin; GPCR, G protein-coupled receptor; IP, prostacyclin; RO1138452, 4,5-dihydro-1H-imidazol-2-yl)-[4-(4-iso-propoxybenzyl)-phenyl]-amine; COPD, chronic obstructive pulmonary disease; CXCL9, monokine induced by interferon-�; CXCL10, interferon-�inducible protein of 10 kDa; CRE, cAMP-response element; KSFM, keratinocyte serum-free medium; HAEC, human primary airway epithelial cell; IFN,interferon; ELISA, enzyme-linked immunosorbent assay; MOI, multiplicity of infection; PKA, cAMP-dependent protein kinase; PKI, cAMP-dependentprotein kinase inhibitor; CMV, cytomegalovirus; BWA 868C, 3-benzyl-5-(6-carboxyhexyl)-1-(2-cyclohexyl-2-hydroxyethylamino)hydantoin; L-161,982,[4�-[3-butyl-5-oxo-1-(2-trifluoromethyl-phenyl)-1,5-dihydro-[1,2,4]triazol-4-ylmethyl]-biphenyl-2-sulfonic acid (3-methyl-thiophene-2-carbonyl)-amide];L-798,106, 5-bromo-2-methoxy-N-[3-(2-naphthalen-2-ylmethyl phenyl)acryloyl]-benzene sulfonamide; Br, bromo; iloprost, (5E)-5-[(3aS,4S,5R,6aS)-5-hydroxy-4-[(E,3S)-3-hydroxy-4-methyl-oct-1-en-6-ynyl]-3,3a,4,5,6,6a-hexahydro-1H-pentalen-2-ylidene]pentanoic acid; taprostene, 3-[(Z)-[(1S,3R,4S,5R)-4-[(E,3S)-3-cyclohexyl-3-hydroxy-prop-1-enyl]-3-hydroxy-8-oxabicyclo[3.3.0]oct-7-ylidene]methyl]benzoic acid; DP, PGD2 receptor;EP, PGE2 receptor; TP, thromboxane receptor; AFP-07, 18,19-didehydro-7,7-difluro-16S,20-dimethyl-PGI2; TEI-9063, 17,20-dimethylisocarbacyclin.

0022-3565/08/3242-815–826$20.00THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS Vol. 324, No. 2Copyright © 2008 by The American Society for Pharmacology and Experimental Therapeutics 129312/3292562JPET 324:815–826, 2008 Printed in U.S.A.

815

http://jpet.aspetjournals.org/content/suppl/2007/10/29/jpet.107.129312.DC1.htmlSupplemental material to this article can be found at:

at ASPE

T Journals on July 1, 2016

jpet.aspetjournals.orgD

ownloaded from

(GPCRs). Studies performed since the early 1980s have pro-vided pharmacological evidence for five main classes ofGPCRs for the naturally occurring prostanoid agonists, andthese classes have been designated DP, EP, FPGF� receptor,IP, and TP (Narumiya et al., 1999). This taxonomy wasformulated from rank orders of agonist potency obtained invarious pharmacological preparations where the first letterdenotes the agonist most selective at that receptor, with thisbeing at least 1 order of magnitude more potent than theother natural ligands. Molecular techniques have now con-firmed this pharmacological classification with the cloningand expression of the five prostanoid receptors in a number ofspecies, including the human (Narumiya et al., 1999).

Currently, the explicit classification of responses mediatedby IP-receptors is hindered by a paucity of suitable pharma-cological tools. Indeed, PGI2 and many PGI2 analogs andmimetics, including iloprost, carbacyclin, AFP-07, and TEI-9063 (see Wise and Jones for details), are not sufficientlyselective for their biological actions to be diagnostic of IP-receptor agonism. Even cicaprost, which is often the agonistof choice in studies examining IP-receptor pharmacology,must be used with caution because it does not effectivelydiscriminate IP-, EP4-, and, to a lesser extent, EP3-receptor-mediated responses (Abramovitz et al., 2000; Wise andJones, 2000). To overcome these problems, we have used herea synthetic PGI2 analog, taprostene (Schneider et al., 1993).This ligand was initially described as a full agonist (Jones etal., 1997). However, subsequent studies established that ithas relatively lower efficacy compared with other IP-agonistsand that it behaves as a partial agonist in tissues whereIP-receptor density is limiting and/or receptor-effector cou-pling efficiency is low (Wise and Jones, 2000; Jones andChan, 2001; Chan and Jones, 2004; Chow et al., 2004). Nev-ertheless, taprostene is selective for the IP-receptor subtypeacross several species, including the human (Chan andJones, 2004; Chow et al., 2004) (Supplemental Fig. 1), and itwas the ligand of choice for most experiments describedherein. Moreover, several IP-receptor antagonists based onthe 2-(phenylamino)imidazoline template were reported re-cently (Clark et al., 2004; Bley et al., 2006). Some of thesecompounds, including RO1138452 (Fig. 1), are highly selec-tive for the IP-receptor subtype, and they act competitively,with affinities in the low nanomolar range (Jones et al.,2006). Thus, these pharmacological tools permit the (patho)physiological role of IP-receptors to be evaluated unambigu-ously that, hitherto, has not previously been possible.

Accumulating evidence suggests that agonism of IP-receptorsin the lung may exert anti-inflammatory activity, antiviralactivity, or both (Jaffar et al., 2002; Takahashi et al., 2002;Hashimoto et al., 2004; Zhou et al., 2007a,b). For example,

exposure of sensitized IP-receptor-deficient mice to allergenis associated with enhanced pulmonary inflammation com-pared with wild-type animals (Takahashi et al., 2002). Sim-ilarly, agonism of IP-receptors has been shown to stimulatethe production of interleukin-10 from murine CD4� T-helper2 cells, and, in vivo, to attenuate allergen-induced, T-helper 2cell-mediated inflammation (Jaffar et al., 2002). Given theemerging anti-inflammatory activity of PGI2 in the lung, twoprinciple objectives formed the basis of the research de-scribed herein. First, we wanted to extend the experimentsperformed with mice (described above) to a human in vitrosystem that may have relevance to the inflammation thatcharacterizes chronic obstructive pulmonary disease (COPD).To this end, we evaluated the role of IP-receptors in regulat-ing the release from human airway epithelial cells (bothprimary and the BEAS-2B cell line) of two CXC chemokines:monokine induced by interferon-� (CXCL9) and interferon-�-inducible protein of 10 kDa (CXCL10). These CXC chemo-kines are potent chemoattractants for CD8� (Tc1) T lympho-cytes (Farber, 1997), and they are elevated in the sputum andairways of subjects with COPD (Saetta et al., 2002; Hardakeret al., 2004; Donnelly and Barnes, 2006). Indeed, the abnor-mal secretion of such CXC chemokines from airway epitheliamay account for the high number of pulmonary CD8� T cellsin subjects with COPD compared with normal healthy indi-viduals (O’Shaughnessy et al., 1997; Saetta et al., 1999).Given that CD8� T cells elaborate perforins and granzyme B(Garcia-Sanz et al., 1988; Chrysofakis et al., 2004), theirinappropriate activation may result in secretion of CXCL9and CXCL10 and promote emphysematous changes in thelung by inducing apoptosis of alveolar epithelial cells (Don-nelly and Barnes, 2006).

The second aim of this study was to characterize in detailthe antagonist properties of RO1138452 and to evaluate theutility of this novel ligand in human IP-receptor classifica-tion. For this purpose, BEAS-2B cells stably harboring acAMP response element (CRE) reporter construct were usedbecause they provide a robust method of interrogating ago-nist-antagonist interactions.

Materials and MethodsCell Culture

BEAS-2B cells were cultured for 3 days under a 5% CO2, airatmosphere at 37°C in 6- or 24-well plastic plates containing kera-tinocyte serum-free medium (KSFM) supplemented with 5 ng/mlepidermal growth factor, 50 �g/ml bovine pituitary extract, 100mg/ml penicillin, and 100 U/ml streptomycin. The cells were culturedfor a further 3 days in fresh KSFM with supplements and thengrowth-arrested for 24 h in supplement-free KSFM. At this time,cultures were tightly confluent, and they were processed for bio-chemical and functional measurements as described below.

Human primary airway epithelial cells (HAECs) were obtained byproteinase digestion of nontransplanted normal human lung (Inter-national Institute for the Advancement of Medicine, Edison, NJ), asdescribed previously (Churchill et al., 1989). Cells were seeded in96-well plates (Costar, Cambridge, MA) containing serum-free, bron-chial epithelial cell growth medium (BioWhittaker, Walkersville,MD). Cells were cultured under a 5% CO2, air atmosphere at 37°Cuntil confluent (after �14 days of culture), growth arrested for 18 hin serum-free, bronchial epithelial cell basal medium, which is de-void of all supplements, and then processed for biochemical andfunctional measurements as described below. Ethics approval for the

Fig. 1. Chemical structure of RO1138452 (Clark et al., 2004; Bley et al.,2006).

816 Ayer et al.

at ASPE

T Journals on July 1, 2016

jpet.aspetjournals.orgD

ownloaded from

use of human tissues has been granted by the Conjoint HealthResearch Ethics Board of the University of Calgary.

Measurement of CXCL9 and CXCL10

Growth-arrested epithelial cells were treated with IP-receptoragonists/antagonists as indicated in the text and figure legends.IFN� was then added, and the cells incubated at 37°C under a 5%CO2 atmosphere for 6 to 48 h. The amount of CXCL9 and CXCL10released into the culture supernatant was quantified by sandwichELISA (Human DuoSet Development System; R&D Systems, Min-neapolis, MN) according to the manufacturer’s instructions.

Infection of Cells with Ad5.CMV.PKI�

In some experiments, subconfluent (�70%), BEAS-2B cells wereinfected (MOI � 20) with an E1�/E3� replication-deficient adeno-virus vector (Ad5.CMV.PKI�) containing a 251-base pair DNA frag-ment encoding the complete amino acid sequence of the �-isoform ofcAMP-dependent protein kinase (PKA) inhibitor (PKI) downstreamof the constitutively active CMV immediate early promoter (Meja etal., 2004). After 48 h, cells were processed for the assessment of CXCchemokine release and activation of a CRE reporter construct asdescribed below. To control for biological effects of the virus per se,the vector Ad5.CMV.Null, expressing no transgene, was used inparallel. Using this experimental protocol, we have reported previ-ously that �90% of BEAS-2B cells are infected with Ad5.CMV.PKI�,resulting in the expression of a completely functional transgene withno adverse effect on cell viability (Meja et al., 2004).

Generation of a CRE-Dependent Luciferase ReporterConstruct

The plasmid pADneo2-C6-BGL contains six CREs in tandem up-stream of a minimal �-globin promoter driving a luciferase gene.BEAS-2B cells at �50% confluence in T-75 flasks were transfectedwith 8 �g of plasmid DNA using Tfx50 (Promega, Madison, WI).After overnight incubation, cells were passaged and cultured inT-162 flasks in the presence of 75 �g/ml G-418 (Geneticin; Invitro-gen, Carlsbad, CA). Medium was changed every 2 to 3 days until fociof stable transfectants were noted (after 2 to 3 weeks of culture).These foci were then harvested to create a heterogeneous populationof cells in which the sites of plasmid integration were randomized.

Measurement of Luciferase

CRE reporter cells were growth-arrested and treated with IP-receptor agonists/antagonists or salbutamol as indicated in the textand figure legends. Cells were incubated at 37°C under a 5% CO2

atmosphere for 1 to 9 h as indicated, and then they were harvestedin reporter lysis buffer; luciferase was then measured according tothe manufacturer’s instructions. Data are expressed as -fold induc-tion of luciferase relative to unstimulated cells. In some experiments,the IP-receptor agonist iloprost was used. Given that this compoundhas nanomolar affinity for several Gs-coupled prostanoid receptors(Abramovitz et al., 2000; Wise and Jones, 2000), the culture mediumwas supplemented with 1 �M BWA 868C, 100 nM L-798,106, and500 nM L-161,982 to block the DP1-, EP3-, and EP4-receptor sub-types, respectively. Although BEAS-2B cells also express the EP1-receptor subtype, selective agonists neither activate nor inhibit CRE-dependent transcription (unpublished observations). Accordingly, aselective EP1-receptor antagonist was not used in these experiments.

RO1138452 Washout Experiments

BEAS-2B cells were incubated for 30 min at 37°C in supplement-free KSFM in the absence and presence of 100 nM RO1138452. Cellswere washed with supplement-free KSFM, incubated in the samemedium for defined periods (see Fig. 8), and exposed to 1 �M tapro-stene. Four hours later, cells were harvested in reporter lysis buffer,and luciferase activity was measured.

Assessment of Cell Viability

The viability HAECs and BEAS-2B cells was determined colori-metrically by measuring the reduction of the tetrazolium salt 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide to formazan,by mitochondrial dehydrogenases.

Curve Fitting

Monophasic agonist concentration-effect (E/[A]) curves were fit-ted by least-squares, nonlinear iterative regression to the follow-ing form of the Hill equation (Prism 4; GraphPad Software Inc.,San Diego, CA):

E � Emin ��Emax � Emin�

1 � 10�pA 50�pA�n (1)

where E is the effect, Emin and Emax are the lower and upper asymp-tote (i.e., the basal response and maximal agonist-induced response,respectively), p[A] is the log molar concentration of agonist, p[A]50 isa location parameter equal to the log molar concentration of agonistproducing Emax/2, and n is the gradient of the E/[A] curve at thep[A]50 level.

Characterization of the IP-Receptor AntagonistRO1138452

The nature of the antagonism produced by RO1138452 at IP-receptors was evaluated by least-squares, nonlinear regression. Dataderived from the CRE-reporter studies were analyzed using a mod-ification of the Hill and Gaddum/Schild equations derived by Waudet al. (1978). Each family of E/[A] curves (i.e., the control E/[A] curveand E/[A] curves constructed in the presence of increasing concen-trations of antagonist) were fitted simultaneously to eq. 2. Thus,

E � Emin � ��Emax � Emin�

1 � �10pA �1 � �B�10�pA2�S�A

�n� (2)

where [A] and [B] are the molar concentration of agonist and antag-onist, respectively, S is the Schild slope factor, and pA2 is the affinityof the antagonist when S � 1, which is equivalent to the pKB. Todetermine whether S deviated significantly from unity, the entirefamily of E/[A] curves that made up an individual experiment wasfitted globally to eq. 2 under two conditions: one condition where Swas constrained to a constant equal to 1, and the other conditionwhere it was a shared value for all data sets. The F-test was appliedto determine statistically which equation gave the best fit, and thisequation was then used for the analysis.

Because complete agonist E/[A] curves could not always be con-structed, especially at high antagonist concentrations, the antago-nism imposed by RO1138452 of CRE-dependent transcription wasalso determined using an alternative experimental approach out-lined by Lazareno and Birdsall (1993). This was also the preferredmethod in the chemokine release studies as a detailed Schild anal-ysis was not practicable. Thus, taprostene was used at a fixed con-centration of 1 �M (�p[A]95; �10 � p[A]50) in the absence andpresence of increasing concentrations of RO1138452. For this type ofanalysis, knowledge of the precise location of the taprostene E/[A]curve is also required; accordingly, this was determined in parallelon the same batch of cells. The resulting pair of antagonist andtaprostene E/[A] curves was then fitted simultaneously to eq. 2. Asbefore, the F-test was applied to determine whether S deviatedsignificantly from unity.

Determination of the Equilibrium Dissociation Constant ofTaprostene

The affinity of taprostene (KA) was estimated by operational modelfitting (Black and Leff, 1983) using two experimental approaches:

RO1138452 and IP-Receptors on Human Airway Epithelia 817

at ASPE

T Journals on July 1, 2016

jpet.aspetjournals.orgD

ownloaded from

following “irreversible” inactivation of a fraction of the total func-tional receptor population ([Ro]) (Furchgott, 1966) and by compari-son of taprostene E/[A] curves with those of the full reference agonistiloprost (Barlow et al., 1967).

Receptor Inactivation. E/[A] curves were generated to tapro-stene in the absence and presence of concentrations of RO1138452that depressed the upper asymptote, Emax, of the control curve. Forthis type of experiment only a single E/[A] curve can be determinedper batch of cells. Accordingly, the data (i.e., all E/[A] curves con-structed in the absence and presence of either a single concentrationof antagonist or every concentration of antagonist), were fitted si-multaneously to eq. 3 (Leff et al., 1990c), which describes a theoret-ical relation between pharmacological effect and agonist concentra-tion (Black and Leff, 1983). Thus,

E �Em � � n � An

�KA � A�n � � n � An (3)

where Em is the theoretical maximal response of the tissue (i.e.,maximal effect produced by a full agonist), and � is the operationalefficacy of the agonist, which is the ratio of [Ro] to the concentrationof agonist-receptor complexes ([AR]) required to produce half-maxi-mal effect ([KE]) (Leff et al., 1990c). This analysis yielded a singleestimate of Em, n, and KA as well as an efficacy value for agonistbefore (��) and after (�) receptor inactivation. The percentage offunctionally active receptors (q) remaining after treatment of cellswith RO1138452 is given by (�/��) � 100.

Comparative Method. E/[A] curves were constructed to tapro-stene, which is a partial agonist in BEAS-2B cells, and the fullagonist iloprost. The experiment was conducted nine times and thereference full agonist and partial agonist E/[A] curves were fittedsimultaneously to eqs. 1 and 3, respectively.

Determination of Receptor Reserve

Receptor occupancy-effect curves were constructed to taprosteneusing the KA determined by receptor “inactivation” and the compar-ative method. At each concentration of agonist and, therefore, ateach level of response, fractional IP-receptor occupancy [i.e., the ratioof agonist occupied IP-receptors to the total number of availablereceptors (RA/Rt)] in untreated, “control” cells was determined ap-plying eq. 4 (Furchgott, 1966), where

RA�Rt � A��KA � A� (4)

As each E/[A] curve was constructed to taprostene in half-logincrements of concentration, the point at which Emax is achievedcannot be determined with accuracy. Thus, an estimate of this valuewas obtained by determining the fraction of agonist-occupied IP-receptors that elicited 95% of the maximal response.

Drugs and Analytical Reagents

RO1138452 (a.k.a. CAY 10441), BWA 868C, iloprost, and taprostenewere from Cayman Chemical (Ann Arbor, MI). L-161,982 andL-798,106 were donated by Merck Frosst (Montreal, QC, Canada). Allother chemicals and reagents were from Sigma-Aldrich (Oakville, ON,Canada). Prostanoid agonists and antagonists were dissolved in di-methyl sulfoxide, and they were diluted to the desired working concen-tration in the appropriate culture medium. None of the compounds ortheir vehicles used in the experiments described herein significantlyaffected cell viability using the reduction of 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide to formazan as an indicator.

General Statistics

Data points, and values in the text and figure legends, representthe mean S.E.M. of n independent determinations. Where appro-priate, data were analyzed statistically using Student’s paired t testor by one-way analysis of variance/Newman-Keuls multiple compar-ison test. The null hypothesis was rejected when p � 0.05.

ResultsChemokine Release

IFN�-Induced CXCL9 and CXCL10 Release fromBEAS-2B Cells. The amount of CXCL9 and CXCL10 releasedspontaneously from BEAS-2B cells after they had been culturedfor 24 h in KSFM (without supplements) was below the detec-tion limit of the assays. In contrast, cells exposed to IFN� (0.1ng/ml–3 �g/ml) released CXCL9 and CXCL10 into the culturesupernatant in a concentration-dependent manner, with p[A]50

values (grams per milliliter) of �7.32 0.2 and �7.53 0.14,respectively (p � 0.05; Fig. 2a). At a near maximally effectiveconcentration of IFN� (3 �g/ml), CXCL9 and CXCL10 werepresent in the culture medium at concentrations of 1.21 0.18and 1.58 0.33 ng/ml, respectively (p � 0.05; Fig. 2a). Unlessstated otherwise, IFN� was used in all further experimentswith BEAS-2B cells at a concentration of 100 ng/ml, whichequated to the p[A]60 for the release of both chemokines.

Kinetic studies established that CXCL9 and CXCL10 werereleased into the culture medium after a lag of approximately4 h. Thereafter, the amount of both chemokines increasedsteadily, with t1/2 values of 29.9 4.1 h (CXCL9) and 18.2 0.8 h (CXCL10; p � 0.05). The amount of CXCL10 peaked atapproximately 30 h after exposure to IFN�, and no furtherchange was detected up to 48 h (Fig. 2b). In contrast, the timecourse of CXCL9 release was more protracted and the max-imum response could not accurately be defined (Fig. 2b).

Effect of Taprostene on CXCL9 and CXCL10 Releasefrom BEAS-2B Cells. Taprostene (1 nM–10 �M) inhibited, ina concentration-dependent manner, the release of CXCL9 fromIFN�-treated BEAS-2B cells, with a p[A]50 (molar) and Emax of�7.31 0.07 and 32.4 1.1%, respectively (n � 18). Statisti-cally indistinguishable data were obtained for taprostene on theoutput of CXCL10 (p[A]50 � �7.40 0.05; Emax � 30.3 0.9%;n � 21; p � 0.05). These inhibitory effects were abolished incells infected with the PKI� expression vector, Ad5.CMV.PKI�(MOI � 20), but not the empty virus, Ad5.CMV.Null (Fig. 3, aand c).

Effect of 8-Br-cAMP on CXCL9 and CXCL10 Releasefrom BEAS-2B Cells. Pretreatment (30 min) of BEAS-2Bcells with 30 �M 8-Br-cAMP attenuated the ability of 100ng/ml IFN� to release CXCL9 and CXCL10 by 31.1 and55.4%, respectively (Fig. 3, b and d). This inhibitory effectwas abolished in cells infected with Ad5.CMV.PKI� (MOI �

Fig. 2. Concentration-dependence and kinetics of IFN�-induced CXCL9and CXCL10 release from BEAS-2B cells. a, growth-arrested cells weretreated with IFN� (0.1 ng/ml–3 �g/ml) for 24 h, and the amount of CXCL9and CXCL10 released into the culture supernatant was determined byELISA. b, growth-arrested cells were treated with a fixed concentration ofIFN� (100 ng/ml), and the concentration of CXCL9 and CXCL10 in theculture supernatant was determined at defined times over a period of48 h. Data points in a and b represent the mean S.E.M. of eight andfour independent determinations, respectively.

818 Ayer et al.

at ASPE

T Journals on July 1, 2016

jpet.aspetjournals.orgD

ownloaded from

20) but not Ad5.CMV.Null (Fig. 3, b and d). Neither virusvector influenced the release of CXCL9 and CXCL10 evokedby IFN� (Fig. 3, b and d).

Effect of RO1138452 on Taprostene-Induced Inhibi-tion of CXCL9 and CXCL10 Release from BEAS-2BCells. RO1138452 had no effect on IFN�-induced chemokinerelease from BEAS-2B cells at any concentration examined.In contrast, RO1138452 (10 pM–10 �M) added to cells con-currently with a fixed concentration of taprostene (1 �M)prevented, in a concentration-dependent manner, the inhibi-tion of CXCL9 and CXCL10 release, with p[A]50 (molar)values of �8.73 0.11 and �8.47 0.16 (p � 0.05), respec-tively (Fig. 4, a and b). Enumeration of the Schild slopefactor, S, by simultaneously fitting to eq. 2 each RO1138452

and taprostene E/[A] curve indicated that this parameterdeviated significantly (p � 0.05) from unity for the antago-nism of both CXCL9 (S � 0.745 0.05; n � 9) and CXCL10(S � 0.674 0.08; n � 9). Thus, RO1138452 behaved in amanner that was inconsistent with surmountable competi-tive antagonism (Neubig et al., 2003).

IFN�-Induced CXCL9 and CXCL10 Release fromHAECs: Effect of Taprostene and Antagonism byRO1138452. The amount of CXCL9 and CXCL10 releasedspontaneously from HAECs after they had been cultured for18 h in serum-free, bronchial epithelial cell basal medium wasroutinely low or below the detection limit of the ELISAs (Fig. 5,a and b). In contrast, exposure of cells to IFN� (0.1–300 ng/ml)resulted in a robust, concentration-dependent release of CXCL9and CXCL10 into the culture supernatant [p[A]50 values (gramsper milliliter) � �8.15 0.09 and �8.25 0.12, respectively;p � 0.05; Fig. 5, a and b]. At the highest concentration of IFN�(300 ng/ml) studied, which was equivalent to the p[A]95 pre-dicted by eq. 1, CXCL9 and CXCL10 were present in the culturemedium in amounts (12.1 1.7 and 12.0 0.54 ng/ml, respec-tively; p � 0.05; Fig. 5, a and b) that were �10-fold higher thanthat produced by BEAS-2B cells under similar cell culture con-ditions (see Fig. 2a). As shown in Fig. 5, a and b, concurrenttreatment of cells with IFN� and taprostene (1 �M) signifi-cantly depressed the upper asymptote, Emax, of the mean E/[A]curves that described the release of both chemokines withoutsignificantly affecting potency [p[A]50 (grams per milliliter):CXCL9 � �8.19 0.09; CXCL10 � �8.32 0.21; p � 0.05].Exposure of HAECs to 30 �M 8-Br-cAMP also attenuated theability of IFN� to release CXCL9 and CXCL10 by 58.9 and50.1%, respectively (Fig. 5, c and d).

RO1138452 had no effect on IFN�-induced chemokine re-lease from HAECs at any concentration examined. However,

Fig. 3. Effect of PKI� on the inhibi-tion by taprostene and 8-Br-cAMP ofCXCL9 and CXCL10 release fromBEAS-2B cells. Naive and Ad5-in-fected BEAS-2B cells (MOI � 20; 48 h)were pretreated (30 min) with tapro-stene (1 nM–10 �M; a and c) or 8-Br-cAMP (30 �M; b and d). IFN� (100ng/ml) was then added, and after 24 hthe amount of CXCL9 (a and b) andCXCL10 (c and d) released into theculture supernatant was quantifiedby a sandwich ELISA. Data in eachpanel represent the mean S.E.M. ofthree independent determinations. �,p � 0.05, significant inhibition of che-mokine release.

Fig. 4. Waud/Birdsall/Lazareno analysis of the antagonism byRO1138452 of taprostene-induced inhibition of CXCL9 and CXCL10 re-lease from BEAS-2B cells. Growth-arrested cells were pretreated (30min) concurrently with RO1138452 (10 pM–1 �M) in the presence of 1 �Mtaprostene, or with taprostene (1 nM–10 �M) alone. IFN� (100 ng/ml) wasthen added, and after 24 h the amount of CXCL9 (a) and CXCL10 (b)released into the culture supernatant was quantified by a sandwichELISA. Each pair of E/[A] curves was then fitted simultaneously to eq. 2from which the Schild slope factor (S) was derived. Data in each panelrepresent the mean S.E.M. of nine independent determinations.

RO1138452 and IP-Receptors on Human Airway Epithelia 819

at ASPE

T Journals on July 1, 2016

jpet.aspetjournals.orgD

ownloaded from

when added to cells concurrently with a fixed concentrationof taprostene (1 �M), RO1138452 (100 pM–1 �M) reversed, ina concentration-dependent manner, the inhibition of CXCL9and CXCL10 release, with mean p[A]50 (molar) values of�7.88 and �8.27, respectively (Fig. 5, e and f). Consistentwith the data shown in Fig. 4, these E/[A] curves were veryshallow, with Hill coefficients considerably less than unity(nH: CXCL9 � 0.313; CXCL10 � 0.425).

CRE-Dependent Transcription

Effect of Taprostene on CRE-Dependent Transcrip-tion. Treatment of BEAS-2B cells stably harboring a CREreporter with taprostene (100 nM, 1 �M, and 10 �M) inducedthe luciferase gene in a time-dependent manner (Fig. 6a).The kinetic of this effect was independent of agonist concen-tration [t1/2(on) �1.7 h], reached a maximum at the 4 to 5-htime point, and thereafter it decayed toward baseline levels

[t1/2(off) � 3.5 h]. At the 5-h time point, the induction bytaprostene of the luciferase gene was concentration-related,with a p[A]50 (molar) and maximal -fold induction, Emax, overuntreated cells of �7.08 0.03 and 3.60 0.10, respectively(n � 48). Infection of cells with Ad5.CMV.PKI� (MOI � 20)abolished taprostene-induced CRE-dependent transcription,whereas Ad5.CMV.Null was inactive (Fig. 6b).

Characterization of the IP-Receptor AntagonistRO1138452. Pretreatment (30 min) of BEAS-2B cells with100 nM RO1138452 had no effect on the basal expression ofluciferase, but it dramatically depressed (by 71.4% at 10 �M)the upper asymptote, Emax, of the mean taprostene E/[A]curve (Fig. 7a). This effect was agonist independent; indeed,a similar reduction in Emax (70% at 1 �M) was producedwhen iloprost was substituted for taprostene (Fig. 7b). Theability of RO1138452 to reduce the taprostene Emax wasconcentration-related, and it was accompanied by a gradeddextral shift of the E/[A] curves and an associated (up to �1.3log10 units at 300 nM) reduction in agonist potency (Fig. 7c;Table 1).

RO1138452 (100 nM) also dramatically reduced Emax whenadded to cells concurrently with taprostene (30.5%) or ilo-prost (43.3%) and luciferase measured at 5 h (Fig. 7, a and b).However, in these experiments the magnitude of Emax de-pression was less than when cells were pretreated with theantagonist before being exposed to taprostene or iloprost(Fig. 7, a and b). Increasing the time cells were exposed toRO1138452 in combination with taprostene or iloprost in anattempt to achieve equilibrium conditions before reporteractivity was measured was impracticable, because the “sig-nal-to-noise” ratio was too high at time points beyond 7 h dueto a rapid decline of the luciferase signal (Fig. 6a).

Figure 7d shows the effect on CRE-dependent transcrip-tion of RO1138452 (10 pM–10 �M) added to cells concur-rently with a fixed concentration of taprostene (1 �M). Usingthis alternative experimental approach, RO1138452 pre-

Fig. 5. Effect of taprostene and 8-Br-cAMP on chemokine release fromHAECs and antagonism by RO1138452.a and b, growth-arrested cells werepretreated (30 min) with 1 �M tapro-stene or vehicle, and then they wereexposed to IFN� (0.1–300 ng/ml). cand d, cells were pretreated (30 min)with 30 �M 8-Br-cAMP, and then theywere exposed to a fixed concentrationof IFN� (100 ng/ml). e and f, cells werepretreated (30 min) with taprostene(T; 1 �M) in the absence and presenceof RO1138452 (10 pM–1 �M), andthen they were exposed to 100 ng/mlIFN�. In each experiment, theamount of CXCL9 (a, c, and e) andCXCL10 (b, d, and f) released into theculture supernatant at 24 h was de-termined by ELISA. Data representthe mean S.E.M. of nine indepen-dent determinations using cells har-vested from four donors (a and b) andfour independent determinations us-ing cells harvested from four donors(c–f). �, p � 0.05, significant inhibitionof chemokine release.

Fig. 6. Kinetics and concentration dependence of taprostene-inducedactivation of a CRE-reporter construct stably expressed in BEAS-2B cells.a, growth-arrested cells were exposed to taprostene (0.1, 1, and 10 �M),and the expression of luciferase was measured every 60 min over a periodof 9 h. b, cells were infected with Ad5.CMV.PKI�, Ad5.CMV.Null, or leftuntreated (naive), and then they were exposed to taprostene (1 nM–10�M). Luciferase was measured at 5 h. Data in a and b represent themean S.E.M. of four and three independent determinations, respec-tively.

820 Ayer et al.

at ASPE

T Journals on July 1, 2016

jpet.aspetjournals.orgD

ownloaded from

vented, in a concentration-dependent manner, the activationof the reporter construct. However, Waud/Lazareno/Birdsallanalysis could not be performed, because the RO1138452E/[A] curves in the presence of taprostene were biphasic,with p[A]50 values for the high- and low-affinity componentsof �8.95 0.46 (57.6 7.1%) and �6.81 0.15, respectively.Thus, using CRE-dependent transcription as a functionaloutput, RO1138452 also behaved in a manner that was in-consistent with surmountable competitive antagonism (Neu-big et al., 2003).

An additional set of washout experiments was conducted togauge the reversibility of RO1138452. BEAS-2B cells werepretreated (30 min) with 100 nM RO1138452 or vehicle,

washed with antagonist-free medium (3 � 2.5 ml), and thenincubated in fresh medium for 0 to 20 h as indicated. Tapro-stene (1 �M) was added, and luciferase was measured at 5 h.As shown in Fig. 8, taprostene activated the reporter �3-foldat all time points when measurements were made, and thisresponse was significantly antagonized by RO1138452 before“washout” (i.e., at time 0) and at all time points thereafter.Indeed, at 20 h, taprostene-induced CRE-dependent tran-scription was suppressed by 55.5 17.8%, and this was notsignificantly different (p � 0.05) from the antagonism pro-duced at time 0 (84.2 9.4%).

Effect of RO1138452 on �2-Adrenoceptor-Mediated,CRE-Dependent Transcription. Treatment of BEAS-2BCRE reporter cells with the selective �2-adrenoceptor agonistsalbutamol (1 �M) increased transcription 8.16 0.72-fold(n � 3). RO1138452 (100 nM), added concurrently with salbu-tamol, failed to significantly antagonize this response (7.83 0.84-fold; n � 3; p � 0.05) under conditions where taprostene-induced transcription was markedly attenuated (Fig. 7a).

Estimation of the KA of Taprostene for the HumanIP-Receptor: Inactivation Method. The insurmountableantagonism produced by RO1138452 in BEAS-2B cells at 5 hwas exploited to estimate the affinity of taprostene for theIP-receptor subtype using operational curve fitting (Tables 2and 3). Each data set that makes up Fig. 7c (i.e., the controlE/[A] curve and the associated E/[A] curves obtained at eachconcentration of antagonist) was fitted globally to eq. 3, fromwhich a mean pKA of �5.89 (KA � 1.3 �M) was determined(Table 3). Analysis of the control E/[A] curves relative tothose generated at each individual concentration of antago-nist showed that the affinity of taprostene estimated in thisway was essentially invariant irrespective of the concentra-tion of RO1138452 used to depress Emax and that it differedby only 0.3 log10 units from the pKA (�6.19) estimated fromthe simultaneous analysis of E/[A] curves at all concentra-tions of antagonist (Table 2). Operational curve fitting im-plied that taprostene had reasonable efficacy in this system(mean �’ � 14.3; Leff et al., 1990c), but, nevertheless, it wasa partial agonist relative to the estimated Em, with an in-trinsic activity (�), calculated as the ratio of (Emax � 1)/(Em �1), of 0.84 (Tables 1 and 2).

Fig. 7. Effect of RO1138452 on taprostene- and iloprost-induced activa-tion of the CRE reporter construct stably expressed in BEAS-2B cells. aand b, growth-arrested cells were exposed to taprostene (1 nM–10 �M) oriloprost (100 pM–1 �M) concurrently (C) with RO1138452 (100 nM) orafter pretreatment (P; 30 min) with the antagonist. c, cells were pre-treated (30 min) with RO1138542 (10, 30, and 300 nM), and then theywere exposed to taprostene (1 nM–10 �M). d, growth-arrested cells wereexposed (30 min) concurrently to RO1138452 (10 pM–1 �M) in the pres-ence of taprostene (1 �M) or to taprostene (1 nM–10 �M) alone. In allexperiments, cells were processed at 5 h, and luciferase activity wasdetermined. Each data set of E/[A] curves in c was fitted globally to eq. 3,to give model parameter estimates (see Table 2). In d, Waud/Birdsall/Lazareno analysis could not be performed because the RO1138452 E/[A]curves in the presence of taprostene were biphasic. Data in a, b, c, and drepresent the mean S.E.M. of four, six, five, and six paired experi-ments, respectively.

TABLE 1Effect of the IP-receptor antagonist RO1138452 on the potency andability of taprostene to induce a CRE-driven luciferase gene stablyexpressed in BEAS-2B epithelial cellsCells were incubated with RO1138452 or vehicle for 30 min at the concentrationsindicated. Taprostene E/A curves were then constructed in the presence of antag-onist and fitted to eq. 1 from which estimates of agonist potency, pA50, and -foldinduction (Emax) of the luciferase gene were interpolated. Data represent the mean S.E.M. of five independent determinations.

Treatment pA50 Emax

Taprostene �7.15 0.08 2.89 0.19Taprostene � RO1138452 (10 nM) �6.63 0.19 2.49 0.16Taprostene � RO1138452 (30 nM) �6.25 0.10 1.89 0.11Taprostene � RO1138452 (300 nM) �5.87 0.16 1.62 0.10

Fig. 8. RO1138452 washout experiments. BEAS-2B cells in 24-wellplates were incubated for 30 min in medium containing 100 nMRO1138452 or vehicle, washed (3 � 2.5 ml) with antagonist-free medium,and incubated in 2.5 ml of the same medium for 0 to 20 h. Taprostene (1�M) was then added, and 5 h later cells were harvested in reporter lysisbuffer, and luciferase activity was measured. Data represent the mean S.E.M. of four independent paired determinations.

RO1138452 and IP-Receptors on Human Airway Epithelia 821

at ASPE

T Journals on July 1, 2016

jpet.aspetjournals.orgD

ownloaded from

Estimation of KA of Taprostene for the Human IP-Receptor: Comparative Method. The IP-receptor agonistiloprost (0.1 nM–1 �M) activated the CRE reporter in aconcentration-dependent manner. As shown in Fig. 9, ilo-prost was �15-fold more potent (p[A]50 � �8.25 0.13; p �0.05) than taprostene, and it was more effective. Thus, tap-rostene was a partial agonist in this system, with an intrinsicactivity [i.e., Emax � 1(taprostene)/Emax � 1(iloprost))] of 0.82,which was similar to that predicted by the inactivationmethod (� � 0.84). Simultaneously fitting each pair of tap-rostene and iloprost E/[A] curves that make up Fig. 9 to eqs.1 and 3, respectively, yielded for taprostene pKA, p�’, n, andEm estimates that were very similar to the same parametersdetermined by receptor inactivation (Table 3).

Relationship between IP-Receptor Occupancy andCRE Activation. Using the KA of taprostene estimated by theinactivation method (1.3 �M), IP-receptor occupancy expressedas a function of response was described by a curvilinear rela-tionship that deviated significantly from the line of identity(where response is directly proportional to occupancy; Fig. 10).Thus, an IP-receptor “reserve” for taprostene was present inBEAS-2B cells at submaximal responses (Fig. 10; Table 4).However, the “spare receptors” in this system declined incre-mentally with concentration and, therefore, magnitude of re-sponse, and they were almost depleted at the Emax (1.6-foldexcess at the p[A]95; Fig. 10; Table 4). The IP-receptor occupan-cy/response relationship was also curvilinear using the KA es-

timated by the comparative method (851 nM), and it was es-sentially indistinguishable from the affinity of taprosteneestimated by receptor inactivation (Fig. 10). Thus, both meth-ods indicated an IP-receptor reserve for taprostene in BEAS-2Bcells at low levels of response that became limiting toward themaximal asymptote (Fig. 10; Table 4).

DiscussionMany IP-receptor agonists have affinity at multiple pro-

stanoid receptors (see Introduction). Accordingly, responsesthey elicit cannot be considered diagnostic of IP-receptoragonism. This lack of selectivity greatly hinders the design ofexperiments and interpretation of data in which the charac-terization of IP-receptor antagonists in a particular tissue isa primary objective. To circumvent these problems, we usedin the present study the stable PGI2 analog taprostene(Schneider et al., 1993). This ligand is selective for the IP-subtype at concentrations up to 10 �M (Chan and Jones,2004; Supplemental Fig. 1); therefore, it can be used in tis-sues expressing multiple prostanoid receptors.

Anti-Inflammatory Potential of IP-Receptor Agonists.Airway epithelial cells can elaborate a plethora of proinflam-matory mediators (Saetta et al., 2002), and for this reason theyare implicated in the pathogenesis of COPD. Herein, we reportthat taprostene suppressed the output, from BEAS-2B cells, oftwo CXC chemokines that are implicated in pulmonary CD8�

T-lymphocyte recruitment. Furthermore, RO1138452 antago-nized these effects of taprostene at concentrations that selec-tively target the IP-receptor (Bley et al., 2006, Jones et al.,2006). RO1138452 also blocked the effect of taprostene onHAECs, indicating that data obtained from BEAS-2B cells onIP-receptor function may reliably be extrapolated to primaryairway epithelia. Further studies established that PKI� abol-ished the effects of taprostene and 8-Br-cAMP on chemokinerelease and CRE-dependent transcription (see below), indicat-ing that human airway epithelial cells express inhibitory IP-receptors coupled to the activation of the cAMP/PKA cascade.

RO1138452 Is a Pseudo-Irreversible Orthosteric An-tagonist. On BEAS-2B cells, RO1138452 did not antagonizethe actions of taprostene in a competitive manner (Schild slopes�1). The interaction of RO1138452 with the IP-receptor onHAECs was also complex as evinced by the very shallow Hillslopes associated with the reversal of CXCL9 and CXCL10release. Therefore, the basis for this antagonism was inves-tigated further using BEAS-2B cells stably transfected witha CRE-reporter construct. This experimental system wasselected as an output over chemokine release because itproduces highly robust and reproducible data; thus, it isideally suited for the interrogation of agonist/antagonist

TABLE 2Estimates of pKA, Emax, n, p��, p�, and q for the effect of RO1138452 on taprostene-induced activation of a CRE-driven luciferase gene stablyexpressed in BEAS-2B airway epithelial cellsCells were incubated with RO1138452 or vehicle for 30 min at the concentrations indicated in the table. Taprostene E/A curves were then constructed in the continuedpresence of antagonist, and the data were fitted to eq. 3 as described under Materials and Methods. Data represent the mean S.E.M. of five independent determinations.

Treatment pKAa Em n p�� p� q (�/�� � 100)

RO1138452 (10 nM) �6.02 0.03 3.32 0.38 1.017 0.13 1.092 0.16 �0.264 0.12 14.9RO1138452 (30 nM) �6.05 0.12 3.19 0.23 0.971 0.10 0.957 0.21 �0.123 0.05 8.3RO1138452 (300 nM) �5.59 0.17 3.21 0.27 1.185 0.31 1.504 0.27 �0.159 0.15 2.2RO1138452 (all conc.)b �6.19 3.14 0.976 0.943a pKA values determined at 10, 30, and 300 nM deviated by �0.46 log10 units (2.9-fold), but this was unrelated to antagonist concentration.b Values were derived by simultaneously fitting the mean E/A curves at all concentrations of antagonist in Fig. 7c to eq. 3.

Fig. 9. Analysis of taprostene E/[A] curve data by comparison with thereference full agonist iloprost. E/[A] curves were constructed in parallel totaprostene (1 nM–10 �M) and iloprost (100 pM–1 �M) for activation ofCRE-dependent transcription. Each resulting pair of taprostene and ilo-prost E/[A] curves was then fitted simultaneously to eqs. 1 and 3, respec-tively. See Table 3 for model parameter estimates. Data represent themean S.E.M. of nine independent paired determinations. All experi-ments with iloprost were performed in medium supplemented with 1 �MBWA 868C, 100 nM L-798,106, and 500 nM L-161,982 to block the DP1-,EP3-, and EP4-receptor subtypes, respectively.

822 Ayer et al.

at ASPE

T Journals on July 1, 2016

jpet.aspetjournals.orgD

ownloaded from

interactions. On CRE-dependent transcription, RO1138452also behaved insurmountably. Thus, taprostene and iloprostE/[A] curves were displaced dextrally, and Emax was de-pressed. In contrast, �2-adrenoceptor-mediated transcriptionwas unaffected by RO1138452, confirming that this antago-nist selectively interacts with the IP-receptor subtype (Clarket al., 2004; Bley et al., 2006). Further evidence for a devia-tion from simple competitive behavior was our inability toperform Waud/Birdsall/Lazareno analysis because RO1138452E/[A] curves in the presence of taprostene (1 �M) were biphasic(see below for discussion of this effect).

Several mechanisms can account for insurmountableantagonism, including covalent receptor inactivation,pseudo-irreversible binding, allosterism, and conditionswhere equilibrium among agonist, antagonist, and recep-tor are incomplete at the time a given response is mea-sured. RO1138452 does not contain any obvious reactivemoieties, implying that the depression of Emax is not due todepletion of functional IP-receptors. A state of hemiequi-librium is also unlikely to account for these results. If sucha mechanism was operative, RO1138452 should have dis-placed taprostene E/[A] curves to the right, similar to thebehavior of a competitive antagonist, and also reducedEmax to a new steady-state level reflecting a situationwhere agonist, antagonist, and receptor have partially re-equilibrated (Kenakin et al., 2006). Reference to Fig. 7cshows that RO1138452 did not produce this pattern ofantagonism. Indeed, RO1138452 depressed Emax in a con-centration-dependent manner that was associated with aprogressive dextral displacement of the taprostene E/[A]curve. Several pieces of evidence also argue against an allo-steric interaction (i.e., the binding of RO1138452 to a sitewithin the receptor that is topographically distinct fromwhere agonists interact). At saturating concentrations, anallosteric inhibitor often incompletely suppresses the func-tional response of interest. In this scenario, a condition isreached where increasing the concentration of antagonist

further fails to produce any additional dextral displacementof the agonist E/[A] curve or depression of Emax (Kenakin et al.,2006). Thus, receptor signaling remains partially intact, andthis accounts for the residual functional response that is mea-sured. The finding that RO1138452 abolished CRE-dependenttranscription (Fig. 7d) clearly excludes this mode of antago-nism. An allosteric antagonist can also completely block re-ceptor signaling and so abolish functional responses medi-ated by that receptor. A prediction of this form of antagonismis that the allosteric ligand, which by definition acts at a sitedistinct from the agonist binding domain, should produce thesame effect on Emax and on the location of the E/[A] curveregardless of whether it is added before, or together with,agonist. The data shown in Fig. 7, a and b, shows that thedepression of Emax and degree of dextral displacement ofiloprost and taprostene E/[A] curves were significantlygreater when cells were pretreated with RO1138452 beforeagonist exposure than when agonist and antagonist wereadded concurrently. Thus, RO1138452 did not behave in amanner consistent with an allosteric antagonist that abol-ishes receptor signaling. On the balance of available evi-dence, we conclude that RO1138452 probably behaves as anorthosteric insurmountable antagonist (i.e., directly interactswith the primary agonist binding site to preclude access foractivating ligands) at the human IP-receptor on human air-way epithelial cells. The relatively modest reduction in Emax

produced by RO1138452 when added to cells concurrentlywith agonist (Fig. 7, a and b) compared with cells subjected toan antagonist pretreatment protocol indicates that the insur-mountability is only apparent. This type of behavior usuallyarises with antagonists that dissociate very slowly from theircognate receptor such that agonist, antagonist, and receptor arenot in equilibrium at the time the response of interest is mea-sured (Vauquelin et al., 2002). Indeed, washout studies estab-lished that RO1138452 was a long-acting antagonist, whichmight be explained by a tight binding, pseudo-irreversibleinteraction with the IP-receptor. Moreover, the finding thatthe degree of antagonism imposed by RO1138452 was notreversed after a 20-h washout period is consistent with thistenet and would also account for the lack of competitivebehavior when chemokine release (measured at 24 h) wasused as a functional output. Thus, on the BEAS-2B cellIP-receptor, our data strongly suggest that RO1138452 be-haved pseudo-irreversibly. Given that equilibrium is not at-tained under these conditions, the affinity of RO1138452cannot be determined (Kenakin et al., 2006).

The mechanism for the long-lasting interaction ofRO1138452 with the IP-receptor was not investigated in thisstudy. However, the biphasic RO1138452 E/[A] curve thatdescribed the antagonism of taprostene-induced transcrip-tion (Fig. 7d) has been documented previously for the inter-action of certain antagonists with the type 1 angiotensin IIreceptor in vascular smooth muscle (Vanderheyden et al.,

TABLE 3Estimates of pKA, Em, n, and p�� determined by the inactivation and comparative methods on taprostene-induced activation of a CRE-drivenluciferase gene stably expressed in BEAS-2B airway epithelial cells

Method pKA Em n p�’

Inactivationa �5.89 0.15 3.24 0.04 1.058 0.06 1.184 0.17Comparativeb �6.07 0.10 3.76 0.17 0.913 0.05 1.041 0.13

a Data represent the mean of the parameter estimates obtained at 10, 30, and 300 nM RO1138452 shown in Table 2.b Data obtained by simultaneously fitting each pair of taprostene and iloprost E/A curves that make up Fig. 9 to eqs. 1 and 3, respectively.

Fig. 10. Relationship between IP-receptor occupancy and activation ofCRE-dependent transcription. The KA values derived by the inactivationand comparative methods were used to calculate the fractional receptoroccupancy (RA/Rt; eq. 4) for taprostene (1 nM–10 �M), and then they wereplotted against defined increments of response.

RO1138452 and IP-Receptors on Human Airway Epithelia 823

at ASPE

T Journals on July 1, 2016

jpet.aspetjournals.orgD

ownloaded from

2000). In those studies, it was proposed that this pharmaco-logical profile represented an equilibrium between “tightbinding/slowly reversible” and “relatively weak binding/rap-idly reversible” states of the antagonist/receptor complex.Indeed, mathematical modeling of such a mechanism wasshown subsequently to readily account for the behavior ofthese antagonists at the type 1 angiotensin II receptor (seeLee et al., 2007, and references therein). Therefore, it ispossible that this two-state model could also help explain theinteraction of RO1138452 with the IP-receptor.

RO1138452 Is a Competitive Surmountable Antagonistin Other Tissues. The insurmountability of RO1138452 wasunexpected given its competitive behavior in other tissues (Bleyet al., 2006; Jones et al., 2006). However, such conflicting dataare not unprecedented. For example, the 5-hydroxytrypta-mine2-receptor antagonist methysergide behaves competitivelyand noncompetitively across different tissues (even from thesame species) that are thought to express the same 5-hydroxy-tryptamine2-receptor (see Vauquelin et al., 2002, and refer-ences therein).

Several explanations could account for this anomaly. Dif-ferences in sequence between orthologous GPCRs that do notaffect the pharmacology of endogenous agonists may, never-theless, modify the interaction of synthetic ligands. Multiplereceptor subtypes expressed variably across tissues couldalso give rise to insurmountability. However, both of thesepossibilities are improbable given that RO1138452 is an ap-parently competitive antagonist in some tissues of humanorigin (Jones et al., 2006) and IP-receptor heterogeneity hasnot convincingly been demonstrated. Two other theories arerelevant. First, the cellular microenvironment in which aGPCR is expressed can influence the behavior of antagonists(Kenakin, 2003). Such differences are, thus, tissue-depen-dent, and antagonists will, therefore, have distinct “pheno-typical” profiles. Whether tissue-dependent factors can alterthe behavior of RO1138452 is unknown, but our data do notallow phenotypical considerations to be excluded. Second, thepresence of a substantial IP-receptor reserve in a tissue maymask, initially, the ability of an antagonist to act insur-mountably (Vauquelin et al., 2002). In such tissues, whichcould include many of those studied by Jones et al. (2006), afull agonist will still elicit the maximum response in thepresence of RO1138452 because it needs to occupy only asmall fraction of the receptor population. Thus, the agonistE/[A] curve is displaced to the right without depression ofEmax. Only when the concentration of RO1138452 is in-creased to a level where IP-receptor number becomes limit-ing is insurmountable behavior seen. This theoretical sce-nario contrasts with the results described herein wheretaprostene was a partial agonist on CRE-dependent tran-scription and IP-receptor number was limiting. Hence,RO1138452, even at very low concentrations, produced a

dextral displacement of the taprostene E/[A] curve and con-comitantly depressed Emax. It is noteworthy that Jones et al.(2006) have reported that the Emax of cicaprost (IP-agonist)for the relaxation of rabbit mesenteric artery and guinea pigaorta was slightly depressed (by �20%) in the presence of100 nM and 1 �M RO1138452, respectively. Although theseinvestigators attributed this effect to functional antagonism(i.e., the activation of contractile EP3-receptors by high ago-nist concentrations), the data could also be explained ifRO1138452 bound pseudo-irreversibly to the IP-receptor un-der conditions where it had depleted the receptor reserve forcicaprost.

Estimation of the KA of Taprostene and ReceptorReserve. An antagonist that behaves pseudo-irreversiblyshould behave identically to an irreversible competitive an-tagonist (Kenakin, 1984). Accordingly, RO1138452 was ex-ploited to estimate the affinity of taprostene for the humanIP-receptor and to determine the relationship between recep-tor occupancy and response. Estimating the affinity of anagonist for a GPCR using pharmacological means is theoret-ically invalid due to the operation of ternary complex mech-anisms (Leff et al., 1990a; Colquhoun, 1998). Indeed, theorypredicts that the method of receptor inactivation overesti-mates affinity by a factor proportional to agonist intrinsicefficacy. Therefore, in the present study, the KA of taprostenewas also determined by the comparative method, using ilo-prost as a reference full agonist. By definition, the analysis ofpartial agonists is considerably less prone to error, and itprovides a close approximation of the true KA if its intrinsicefficacy is low relative to the full agonist. As described by Leffand Harper (1989), simultaneous application of these twomethods provides an experimental “check” for the operationof ternary complex mechanisms because these independentestimates of affinity should not correspond.

Using this practical test, the KA of taprostene was esti-mated to be 1.3 and 0.85 �M using the receptor inactivationand comparative methods, respectively. Thus, in this exper-imental system, there was no indication that the receptorinactivation method introduced the error in affinity predictedby theory. Similar results have been reported in other tissues(Waud, 1969; Leff et al., 1990b), indicating that under certaincircumstances, pharmacological methods can provide reliableestimates of affinity despite the operation of ternary complexmechanisms (Leff et al., 1990a). Moreover, the KA of tapro-stene for the human IP-receptor reported herein is very sim-ilar to its pA2 (6.1) determined using the Schild equation forthe antagonism of IP-receptor-mediated relaxation of rabbitsaphenous vein (Jones and Chan, 2005). However, speciesdifferences in IP-receptor pharmacology may be apparentsince the affinity of taprostene is 6- to 25-fold higher for therelaxant IP-receptor expressed in pig and guinea pig saphe-

TABLE 4IP-receptor occupancy/response relationship for taprostene-induced activation of CRE-dependent transcription

Response Level (% Max) 25% 50% 75% 95%

Occupancy (%)a 2.5 0.2 7.3 0.5 20.6 1.4 59.6 3.3Receptor excess (-fold)a 10 6.8 3.6 1.6Occupancy (%)b 3.8 0.3 10.7 0.8 27.9 1.7 68.7 3.0Receptor excess (-fold)b 6.6 4.7 2.7 1.4

a Data calculated using the KA (1.3 � M) determined by the inactivation method.b Data calculated using the KA (0.85 � M) determined by the comparative method.

824 Ayer et al.

at ASPE

T Journals on July 1, 2016

jpet.aspetjournals.orgD

ownloaded from

nous veins (pA2 � 7.4 and 6.8, respectively) (Jones and Chan,2005).

The KA of taprostene estimated by both pharmacologicalmethods was then used to establish the relationship betweenIP-receptor occupancy and response. Irrespective of whetherthe KA was set to a value of 1.3 or 0.85 �M, there was areceptor reserve for taprostene at low agonist concentrationsthat declined incrementally with response and was essen-tially depleted at the Emax. These data are consistent withthe ability of RO1138452 to depress the taprostene Emax atconcentrations as low as 10 nM. Stated differently, occupancyof all available IP-receptor is required for taprostene to elicitits maximum response.

Conclusions. Herein, we report that human airwayepithelial cells express functional IP-receptors. This find-ing adds to an expanding literature that documents theubiquitous expression of IP-receptors in the lung andwhere, in vivo, selective agonists may have anti-inflamma-tory activity, antiviral activity, or both. Although tapro-stene incompletely suppressed chemokine output, greaterinhibition can, theoretically, be achieved. Indeed, tapro-stene was a partial agonist, indicating that the IP-receptorsystem in airway epithelia can be activated further withligands of higher efficacy. Furthermore, because cAMPinhibits the release of many mediators implicated in air-way inflammation, selective agonists may have wide-spread activity on those proinflammatory and immunecells that express the IP-receptor subtype.

Our investigation also suggest that RO1138452 is apseudo-irreversible orthosteric IP-receptor antagonist.The unexpected deviation from surmountable competitivebehavior confirms the need, in drug discovery, to screenpotential new medicines in the target tissue(s) of the rel-evant species as differences in receptor reserve and tissuephenotypes could have profound pharmacological and ther-apeutic implications.

Acknowledgments

We thank Merck Frosst for supplying L-161,982 and L-798,106and Dr. Robert Newton (Department of Cell Biology and Anatomy,University of Calgary, Calgary, AB, Canada) for donating the CRE-reporter constructs.

ReferencesAbramovitz M, Adam M, Boie Y, Carriere M, Denis D, Godbout C, Lamontagne S,

Rochette C, Sawyer N, Tremblay NM, et al. (2000) The utilization of recombinantprostanoid receptors to determine the affinities and selectivities of prostaglandinsand related analogs. Biochim Biophys Acta 1483:285–293.

Barlow RB, Scott NC, and Stephenson RP (1967) The affinity and efficacy of oniumsalts on the frog rectus abdominis. Br J Pharmacol Chemother 31:188–196.

Black JW and Leff P (1983) Operational models of pharmacological agonism. Proc RSoc Lond B Biol Sci 220:141–162.

Bley KR, Bhattacharya A, Daniels DV, Gever J, Jahangir A, O’Yang C, Smith S,Srinivasan D, Ford AP, and Jett MF (2006) RO1138452 and RO3244794: charac-terization of structurally distinct, potent and selective IP (prostacyclin) receptorantagonists. Br J Pharmacol 147:335–345.

Bunting S, Gryglewski R, Moncada S, and Vane JR (1976) Arterial walls generatefrom prostaglandin endoperoxides a substance (prostaglandin X) which relaxesstrips of mesenteric and coeliac arteries and inhibits platelet aggregation. Prosta-glandins 12:897–913.

Chan KM and Jones RL (2004) Partial agonism of taprostene at prostanoid IPreceptors in vascular preparations from guinea-pig, rat, and mouse. J CardiovascPharmacol 43:795–807.

Chow KB, Jones RL, and Wise H (2004) Agonists can discriminate between clonedhuman and mouse prostacyclin receptors. Prostaglandins Leukot Essent FattyAcids 70:423–429.

Chrysofakis G, Tzanakis N, Kyriakoy D, Tsoumakidou M, Tsiligianni I, Klimathia-naki M, and Siafakas NM (2004) Perforin expression and cytotoxic activity ofsputum CD8� lymphocytes in patients with COPD. Chest 125:71–76.

Churchill L, Chilton FH, Resau JH, Bascom R, Hubbard WC, and Proud D (1989)

Cyclooxygenase metabolism of endogenous arachidonic acid by cultured humantracheal epithelial cells. Am Rev Respir Dis 140:449–459.

Clark RD, Jahangir A, Severance D, Salazar R, Chang T, Chang D, Jett MF, SmithS, and Bley K (2004) Discovery and SAR development of 2-(phenylamino)imida-zolines as prostacyclin receptor antagonists. Bioorg Med Chem Lett 14:1053–1056.

Colquhoun D (1998) Binding, gating, affinity and efficacy: the interpretation ofstructure-activity relationships for agonists and of the effects of mutating recep-tors. Br J Pharmacol 125:924–947.

Donnelly LE and Barnes PJ (2006) Chemokine receptors as therapeutic targets inchronic obstructive pulmonary disease. Trends Pharmacol Sci 27:546–553.

Farber JM (1997) Mig and IP-10: CXC chemokines that target lymphocytes. J LeukocBiol 61:246–257.

Furchgott RF (1966) The use of �-haloalkylamines in the differentiation of receptorsand in the determination of dissociation constants of receptor-agonist complexes.Adv Drug Res 3:21–55.

Garcia-Sanz JA, Velotti F, MacDonald HR, Masson D, Tschopp J, and Nabholz M(1988) Appearance of granule-associated molecules during activation of cytolyticT-lymphocyte precursors by defined stimuli. Immunology 64:129–134.

Hardaker EL, Bacon AM, Carlson K, Roshak AK, Foley JJ, Schmidt DB, Buckley PT,Comegys M, Panettieri RA Jr, Sarau HM, et al. (2004) Regulation of TNF�- andIFN�-induced CXCL10 expression: participation of the airway smooth muscle inthe pulmonary inflammatory response in chronic obstructive pulmonary disease.FASEB J 18:191–193.

Hashimoto K, Graham BS, Geraci MW, FitzGerald GA, Egan K, Zhou W, Gole-niewska K, O’Neal JF, Morrow JD, Durbin RK, et al. (2004) Signaling through theprostaglandin I2 receptor IP protects against respiratory syncytial virus-inducedillness. J Virol 78:10303–10309.

Jaffar Z, Wan KS, and Roberts K (2002) A key role for prostaglandin I2 in limitinglung mucosal Th2, but not Th1, responses to inhaled allergen. J Immunol 169:5997–6004.

Jones RL and Chan KM (2001) Distinction between relaxations induced via prosta-noid EP4 and IP1 receptors in pig and rabbit blood vessels. Br J Pharmacol134:313–324.

Jones RL and Chan KM (2005) Investigation of the agonist activity of prostacyclinanalogues on prostanoid EP4 receptors using GW 627368 and taprostene: evidencefor species differences. Prostaglandins Leukot Essent Fatty Acids 72:289–299.

Jones RL, Wise H, Clark R, Whiting RL, and Bley KR (2006) Investigation of theprostacyclin (IP) receptor antagonist RO1138452 on isolated blood vessel andplatelet preparations. Br J Pharmacol 149:110–120.

Jones RL, Qian YM, Wise H, Wong HN, Lam WL, Chan HW, Yim AP, and Ho JK(1997) Relaxant actions of non-prostanoid prostacyclin mimetics on human pul-monary artery. J Cardiovasc Pharmacol 29:525–535.

Kenakin TP (1984) The classification of drugs and drug receptors in isolated tissues.Pharmacol Rev 36:165–222.

Kenakin T (2003) Predicting therapeutic value in the lead optimization phase of drugdiscovery. Nat Rev Drug Discov 2:429–438.

Kenakin TP, Jenkinson S, and Watson C (2006) Determining the potency andmolecular mechanism of action of insurmountable antagonists. J Pharmacol ExpTher 319:710–723.

Lazareno S and Birdsall NJ (1993) Estimation of competitive antagonist affinityfrom functional inhibition curves using the Gaddum, Schild and Cheng-Prusoffequations. Br J Pharmacol 109:1110–1119.

Lee MT, Pugsley MK, Vauquelin G, and van Liefde I (2007) Molecular characteri-sation of the interactions between olmesartan and telmisartan and the humanangiotensin II AT1 receptor. Br J Pharmacol 151:952–962.

Leff P, Dougall IG, Harper DH, and Dainty IA (1990a) Errors in agonist affinityestimation: do they and should they occur in isolated tissue experiments? TrendsPharmacol Sci 11:64–67.

Leff P, Harper D, Dainty IA, and Dougall IG (1990b) Pharmacological estimation ofagonist affinity: detection of errors that may be caused by the operation of receptorisomerisation or ternary complex mechanisms. Br J Pharmacol 101:55–60.

Leff P and Harper EA (1989) Do pharmacological methods for the quantification ofagonists work when the ternary complex mechanism operates? J Theor Biol140:381–397.

Leff P, Prentice DJ, Giles H, Martin GR, and Wood J (1990c) Estimation of agonistaffinity and efficacy by direct, operational model-fitting. J Pharmacol Methods23:225–237.

Meja K, Catley MC, Cambridge LM, Barnes PJ, Lum H, Newton R, and GiembyczMA (2004) Adenovirus-mediated delivery and expression of a cAMP-dependentprotein kinase inhibitor gene to BEAS-2B epithelial cells abolishes the anti-inflammatory effects of rolipram, salbutamol and prostaglandin E2: a comparisonwith H-89. J Pharmacol Exp Ther 309:833–844.

Narumiya S, Sugimoto Y, and Ushikubi F (1999) Prostanoid receptors: structures,properties, and functions. Physiol Rev 79:1193–1226.

Neubig RR, Spedding M, Kenakin T, and Christopoulos A (2003) International Unionof Pharmacology Committee on Receptor Nomenclature and Drug Classification.XXXVIII. Update on terms and symbols in quantitative pharmacology. PharmacolRev 55:597–606.

O’Shaughnessy TC, Ansari TW, Barnes NC, and Jeffery PK (1997) Inflammation inbronchial biopsies of subjects with chronic bronchitis: inverse relationship of CD8�

T lymphocytes with FEV1. Am J Respir Crit Care Med 155:852–857.Saetta M, Baraldo S, Corbino L, Turato G, Braccioni F, Rea F, Cavallesco G,

Tropeano G, Mapp CE, Maestrelli P, et al. (1999) CD8�ve cells in the lungs ofsmokers with chronic obstructive pulmonary disease. Am J Respir Crit Care Med160:711–717.

Saetta M, Mariani M, Panina-Bordignon P, Turato G, Buonsanti C, Baraldo S,Bellettato CM, Papi A, Corbetta L, Zuin R, et al. (2002) Increased expression of thechemokine receptor CXCR3 and its ligand CXCL10 in peripheral airways ofsmokers with chronic obstructive pulmonary disease. Am J Respir Crit Care Med165:1404–1409.

RO1138452 and IP-Receptors on Human Airway Epithelia 825

at ASPE

T Journals on July 1, 2016

jpet.aspetjournals.orgD

ownloaded from

Schneider J, Friderichs E, Kogel B, Seipp U, Stahlberg H-J, Terlinden R, andHeintze K (1993) Taprostene sodium. Cardiovasc Drug Rev 11:479–500.

Takahashi Y, Tokuoka S, Masuda T, Hirano Y, Nagao M, Tanaka H, Inagaki N,Narumiya S, and Nagai H (2002) Augmentation of allergic inflammation in pro-stanoid IP receptor-deficient mice. Br J Pharmacol 137:315–322.

Vanderheyden PML, Fierens FLP, and Vauquelin G (2000) Angiotensin II type 1receptor antagonists. Why do some of them produce insurmountable inhibition?Biochem Pharmacol 60:1557–1563.

Vane JR and Botting RM (1995) Pharmacodynamic profile of prostacyclin. Am JCardiol 75:3A–10A.

Vauquelin G, Van Liefde I, Birzbier BB, and Vanderheyden PML (2002) New in-sights in insurmountable antagonism. Fundam Clin Pharmacol 16:263–272.

Waud DR (1969) On the measurement of the affinity of partial agonists for receptors.J Pharmacol Exp Ther 170:117–122.

Waud DR, Son SL, and Waud BE (1978) Kinetic and empirical analysis of dose-response curves illustrated with a cardiac example. Life Sci 22:1275–1285.

Wise H and Jones RL (2000) Prostacyclin and Its Receptors, Kluwer AcademicPublishers, Hingham, MA.

Zhou W, Blackwell TS, Goleniewska K, O’Neal JF, FitzGerald GA, Lucitt M, BreyerRM, and Peebles RS (2007a) Prostaglandin I2 analogs inhibit Th1 and Th2 effectorcytokine production by CD4 T cells. J Leukoc Biol 81:809–817.

Zhou W, Hashimoto K, Goleniewska K, O’Neal JF, Ji S, Blackwell TS, FitzGeraldGA, Egan KM, Geraci MW, and Peebles RS (2007b) Prostaglandin I2 analogsinhibit pro-inflammatory cytokine production and T cell stimulatory function ofdendritic cells. J Immunol 178:702–710.

Address correspondence to: Dr. Mark A. Giembycz, Department of Phar-macology and Therapeutics, Airways Inflammation Group, Institute of Infec-tion, Immunity and Inflammation, University of Calgary, 3330 Hospital Dr.N.W., Calgary, AB, Canada T2N 4N1. E-mail: [email protected]

826 Ayer et al.

at ASPE

T Journals on July 1, 2016

jpet.aspetjournals.orgD

ownloaded from