Prostaglandin E2 (PGE2), similar to β-adrenergic receptor agonists, can protect airways from bronchoconstriction and resulting increase in airway resistance induced by a number of agents, including cholinergic receptor agonists and antigen. We examined the impact of sustained alterations in PGE2 pathways on changes in airway resistance. Genetic methods were utilized to alter PGE2 metabolism and signal transduction in the murine lung. PGE2 levels were elevated by generating mice lacking 15-hydroxyprostaglandin (Hpgd−/−), the major catabolic enzyme of PGE2, and by generating a transgenic line in which mouse PGE2 synthase (Ptges) expression is driven by a human lung-specific promoter, hSP-C. Conversely, to determine the impact of loss of PGE2 on airway reactivity, we examined mice lacking this synthase (Ptges−/−) and receptors that mediate the actions of PGE2, particularly the PGE2 EP2 receptor (Ptger2). Diminished capacity to produce and respond to PGE2 did not alter the response of mice to cholinergic stimuli. In contrast, the responsiveness to cholinergic stimulation was dramatically altered in animals with elevated PGE2 levels. The Hpgd−/− and hSP-C-Ptges transgenic lines both showed attenuated airway responsiveness to methacholine as measured by lung resistance. Thus, whereas compromise of the Ptges/PGE2/Ptger2 pathway does not alter airway responsiveness, genetic modulation that elevates PGE2 levels in the lung attenuates airway responsiveness.
- lung mechanics
airways are dynamic structures that can change their diameter in response to both endogenous and exogenous stimuli. For instance, airways can dilate when it is necessary to move large volumes of air, such as during exercise, and, conversely, upon exposure to toxic fumes or irritating gases, the airway can constrict to limit the exposure of the lungs (18). The ability of the airways to reversibly alter their size in response to stimuli has been termed airway responsiveness. An exaggerated bronchoconstrictor response to a provoking stimulus is termed airway hyperresponsiveness and is one of the cardinal characteristics of asthma (23).
Acetylcholine release by the parasympathetic airways stimulates Gq-coupled receptors on airway smooth muscle, and this pathway is largely responsible for maintaining airway tone (11). Although sympathetic innervations play less of a role in normal airway homeostasis, the constrictive actions of acetylcholine can be countered by the actions of epinephrine (adrenaline), which acts on β-adrenergic receptors expressed by smooth muscle cells (23). Activation of these Gs-coupled receptors relaxes airway smooth muscle, increasing lung ventilation. This response is the basis for the most common treatment of asthma: β-agonists. Airway tone and responsiveness can be modulated by other mediators, including prostaglandin E2 (PGE2).
PGE2 is produced from arachidonic acid (AA) by the sequential actions of a number of enzymes. AA is released from the membrane by phospholipase A2. The free AA is available to be metabolized by PTGS1 (Cox1) or PTGS2 (Cox2) to PGH2, the common precursor of all prostanoids. PGE2 is formed when PGH2 is further metabolized by a PGE synthase (PGES). Multiple enzymes capable of carrying out this response have been identified (1, 14, 25, 33–35, 38). However, to date, only one has been shown to contribute to the production of PGE2 in vivo (38).
PGE2 can activate four unique high-affinity seven-transmembrane G-coupled protein receptors commonly referred to as EP1–EP4 and encoded by four unique genes, Ptger1–Ptger4 (9, 20, 22). These receptors, through coupling to various G proteins, can affect a variety of sometimes opposing intracellular signaling pathways (22). The EP2 and EP4 receptors are coupled to Gs, and in all cells examined to date, activation of these receptors leads to an increase in cAMP. In contrast, EP1 and EP3 would be expected to contribute to smooth muscle contraction, because both can increase intracellular Ca2+ levels and/or decrease intracellular cAMP.
Exogenously administered PGE2 has been shown to have a complex and sometimes seemingly opposing effect on airways. In clinical studies, pretreatment with inhalated PGE has been shown to be bronchoprotective against a variety of challenges including allergen-, exercise-, and aspirin-induced bronchoconstriction (10, 21, 26, 28). However, in some trials after PGE inhalation, patients reported transient cough and retrosternal soreness (26). In addition, a small subset of patients, upon inhalation of PGE, experienced profound bronchoconstriction, sometimes requiring rescue with a β-agonist (5). In murine models of airway reactivity, PGE2 inhalation has been shown to cause a small but significant rise in Penh (37). This response was dependent on the expression of the EP1 or EP1 and EP3 receptors, depending on the mouse strain studied (37). At the same time, PGE2 inhalation has been shown to protect against methacholine-induced increase in lung resistance (RL) (29, 37). Furthermore, PGE2 could relax preconstricted tracheal rings. These actions were dependent on the expression of the EP2 receptor (8, 29, 37).
Studies of human airway smooth muscle cells have indicated that these cells express EP2 receptors (2, 3). The expression of EP2 by these cells is consistent with a bronchoprotective role of PGE2, raising the possibility that modulation of PGE2 levels might determine the responsiveness of the airways to provocative stimuli. However, a previous attempt to decrease airway responsiveness with chronic overexpression of Gs-coupled β2-adrenergic receptor (β2-AR) produced unexpected results (19). Mice with elevated β2-AR expression displayed an exacerbated response to methacholine, suggesting that an adaptive change had occurred, compensating for the constitutive increase in the activity of this Gs-coupled pathway. Mice deficient in β-adrenergic receptors surprisingly showed an attenuated response to constricting agents (19). It was important, therefore, to determine whether such compensatory changes would be observed upon alteration of the activity of other Gs-coupled receptors expressed by airways.
To determine the impact of alterations in PGE2 levels on airway responsiveness, we first examined mice in which the available PGE2 is reduced or mice that lack EP2. We next examined mice in which PGE2 levels are expected to be increased. The first mouse line lacks 15-hydroxyprostaglandin dehydrogenase (PGDH; Hpgd). Hpgd is expressed at high levels in the lung and metabolizes PGE2, its preferred substrate, to 15-keto-PGE2. In the second mouse line, PGE2 levels are increased by expression of a transgenic hSP-C-Ptges in the airway epithelia.
MATERIALS AND METHODS
All studies were approved by and conducted in accordance with the National Institutes of Health Guide for Care and Use of Laboratory Animals as well as the Institutional Animal Care and Use Committee guidelines of the University of North Carolina at Chapel Hill. Generation of Ptger2−/− (36), Ptges−/− (38), and Hpgd−/− (4) mice in this study has been previously reported. The Ptger2 mutation was maintained on a relatively inbred 129 background consisting of a mixture of 129/Ola and 129/SvEv. Hpgd −/− mice were originally generated on a mixed background of C57BL/6, DBA/2, and 129/SvEv. For the mixed background experiments, heterozygous siblings were mated to generate wild-type and Hpgd−/− mice. Inbred C57BL/6 mice homozygous for the mutant Hpgd allele were generated by repeatedly backcrossing the Hpgd mutation onto the C57BL/6 mouse strain for a minimum of four generations and then intercrossing the resulting N4 B6-Hpgd+/−mice. Intercrosses of B6 mice heterozygous for the disrupted Hpgd allele produced B6-Hpgd−/− and B6-Hpgd+/+ animals.
Generation of transgenic mouse line.
A full-length Ptges1 cDNA was generated by RT-PCR from total RNA isolated from murine lungs. (For a schematic representation of the transgene expression construct, see Fig. 7A.) The human surfactant protein C (hSP-C) promoter has been used extensively to generate lung-specific transgenic animals (12, 16, 39). This promoter fragment has been shown in mice to direct gene expression to distal epithelial bronchiolar cells (12, 16, 39). Generation of transgenic mice involved colipofection of the transgene construct along with a selectable marker into murine 129/Ola ES cells. Cells were selected for expression of the marker, and clones were genotyped for transgene integration by Southern blotting, using a fragment of the hSP-C promoter as a probe. Transgene-positive cells were injected into C57BL/6 embryos, which were then implanted into pseudopregnant recipient females. Chimeric animals were then bred with B6D2 mice, and the presence of agouti pups confirmed targeted ES cell transmission. Genomic DNA isolation and subsequent Southern blotting with the SP-C probe confirmed transmission of the transgene in pups. Transgene-positive mice were bred to B6D2 mice to generate wild-type and transgene-positive mice on the mixed background. Transgenic-positive animals were obtained in the expected Mendelian ratios consistent with a single genomic integration, developed normally, and had no overt phenotype. (See Fig. 7B, a Northern blot of tissues from both transgene-positive and wild-type littermates showing a lung-specific increase of the Ptges transcript in transgene-positive animals.) Congenic C57BL/6 mice carrying the hSP-C-Ptges transgene were generated by repeated crossing of transgenic mice with purchased C57BL/6 mice. Wild-type littermates were used as controls in these studies. Wild-type C57BL/6 and B6D2 mice were purchased from Jackson Laboratories.
Measurements of airway resistance in anesthetized mice.
Mice were anesthetized with 70–90 mg/kg pentobarbital sodium, tracheostomized, and mechanically ventilated at a rate of 300–350 breaths/min, a tidal volume of 150 μl, and a positive end-expiratory pressure (PEEP) of 3–4 cmH2O with a computer-controlled small animal ventilator (Scireq, Montreal, Canada). Once ventilated, mice were paralyzed with 0.8 mg/kg pancuronium bromide. Using the custom-designed software Flexivent (Scireq), we recorded airway pressure and airflow using a precisely controlled piston during a single inspiration and expiration with an amplitude of 150 μl and a period of 1 s (SnapShot). Applying this data to the single-compartment equation of motion (Eq. 1) allows determination of RL and dynamic compliance (CDYN): (1) where P(t) is pressure at time t, R is resistance, v̈(t) is flow at time t, C is compliance, V(t) is volume at time t, and P0 is resting pressure (PEEP).
Five separate measurements of lung parameters were taken before exposure to methacholine. These measurements were used to calculate the baseline RL and CDYN for each mouse. Aerosols of methacholine were delivered via a nebulizer (DeVilbiss) through a side port in the ventilator circuit for 30 s at a rate of 200 breaths/min. After methacholine exposure, RL and CDYN were measured every 10 s for 3 min, and the maximum value of RL and minimum value of CDYN for each dose were determined. For the indomethacin-pretreatment experiments, mice received an intravenous tail vein injection of 10 mg/kg indomethacin (Sigma) at least 4 h before the start of the experiment.
Quantitative real-time RT-PCR.
Total RNA was extracted from frozen pulverized tissue using RNAzolB (Teltest, Friendswood, TX) according to the manufacturer’s instructions. RNA was purified using Qiaprep RNAeasy columns (Qiagen), and cDNA was generated from 4 μg of purified RNA using Multiscribe reverse transcriptase (cDNA archive kit; Applied Biosystems, Foster City, CA). cDNA (2.5 ng) was used in amplification reactions with a SYBR green PCR master mix kit (Applied Biosystems) in accordance with the manufacturer’s instructions. All amplification were performed in quadruplicate on a Stratagene Mx300P cycler. Primers and probe set for the following genes were purchased from Assays on Demand (Applied Biosystems): Ptges, β-actin, Ptgs1, Ptgs2, Hpgd, Ptger2, and Pde4b. Differences in expression between tissues were determined using the comparative threshold cycle (ΔΔCT) method as suggested by the manufacturer (Applied Biosystems), normalizing each sample to β-actin levels. The expression of the transgene was expressed as the degree of change relative to wild-type littermates.
Lung tissue was homogenized in PBS/1 mM EDTA with 10 μg/ml indomethacin. Samples were applied to Amprep C18 octadecyl columns (Amersham) for extraction. PGE2 levels were assessed using the Correlate-EIA PGE2 kit (Assay Designs). Samples were run in duplicate and normalized to tissue weight.
All data are presented as means ± SE. Statistical significance for single data points was assessed using Student’s two-tailed t-test. Multivariate repeated-measures analysis of variance (MANOVA) was performed on all dose-response curves. The JMP IN statistical package was used to perform all statistical analyses (SAS, Cary, NC).
Basal tone and airway reactivity in mice with compromised PGE2 pathways.
We first determined whether either loss of the Gs-coupled EP2 receptor for PGE2 or diminished capacity to produce PGE2 due to loss of mPGES1 altered basal airway tone and/or airway reactivity in response to methacholine. To do this, we used a direct measurement of lung function. RL and CDYN in mice were calculated by determining resistance and compliance via the equation of motion (Eq. 1) as described in materials and methods. In anesthetized, mechanically ventilated mice, baseline lung parameters were established by averaging five separate measurements (SnapShot) for each mouse. Ptger2−/− and wild-type mice on the 129/SvEv background did not have significantly different baseline RL or CDYN values. Mean baseline RL values were as follows: Ptger2−/−, 1.33 ± 0.07 cmH2O·s·ml−1; and wild type, 1.32 ± 0.06 cmH2O·s·ml−1. Mean baseline CDYN values were as follows: Ptger2−/−, 0.035 ± 0.002 ml/cmH2O; and wild type, 0.034 ± 0.002 ml/cmH2O. Next, we determined whether loss of EP2-mediated PGE2 signaling in the lungs would alter the airway responsiveness to methacholine (a congenitor of acetylcholine that is metabolized less rapidly by acteylcholinesterases). Mice were exposed to doubling doses of aerosolized methacholine (3–100 mg/ml). Response to methacholine was plotted as the peak RL and minimum CDYN values obtained for each dose as a percentage of the baseline RL and CDYN values. Loss of EP2 did not alter airway responsiveness as measured by either RL or CDYN (Fig. 1). These experiments were repeated using a congenic C57BL/6 Ptger2−/− mouse pair, and again no difference in either baseline airway function or response to methacholine could be detected using these methods.
The Ptges−/− and wild-type mice on the DBA/1lacJ background, similar to the Ptger2−/−, did not have significantly different baseline RL or CDYN values. Mean baseline RL values were as follows: Ptges−/−, 1.43 ± 0.02 cmH2O·s·ml−1; and wild type, 1.44 ± 0.05 cmH2O·s·ml−1. Mean baseline CDYN values were as follows: Ptges−/−, 0.030 ± 0.001 ml/cmH2O; and wild type, 0.030 ± 0.002 ml/cmH2O. Next, we determined whether loss of mPGES1-mediated PGE2 production in the lungs would alter airway responsiveness to methacholine. Airway reactivity was evaluated as described for the Ptger2−/− cohort. Loss of mPGES1 did not alter airway responsiveness as measured by either RL or CDYN (Fig. 2). Thus neither loss of the EP2 receptor nor diminished capacity to produce PGE2 resulted in compensatory changes that alter the responsiveness of the airways to cholinergic stimulation.
Airway function in mice deficient in 15-PGDH.
We next determined whether elevated levels of PGE2 would alter baseline airway tone and/or diminish the response of the airways to methacholine. To do this, we utilized mice lacking 15-PGDH (Hpgd−/−). The inability of these mice to rapidly metabolize PGE2 in the perinatal period results in failure of the pups to remodel the ductus arteriosus. However, these animals can be rescued by a single injection of indomethacin in the perinatal period. We determined whether loss of this enzyme would lead to a measurable increase in PGE2 levels in the adult lung (Fig. 3). A small but significant increase was observed in levels of PGE2 in the lungs of Hpgd−/− animals. The Hpgd mutation was originally studied on a genetic background composed of alleles from three inbred strains of mice: C57BL/6, 129/SvEv, and DBA/2J. However, congenic C57BL/6 Hpgd−/− animals also were generated for these studies. No histological changes were apparent on examination of the lungs of either group of animals. We began these studies using a cohort of Hpgd−/− and wild-type littermates of mixed genetic background. Loss of PGDH significantly decreased airway resistance and increased CDYN in the unstimulated mouse. Mean baseline RL values were as follows: Hpgd−/−, 1.18 ± 0.09 cmH2O·s·ml−1; and wild type, 1.39 ± 0.08 cmH2O·s·ml−1 (P < 0.05, t-test). Mean baseline CDYN values were as follows: Hpgd−/−, 0.040 ± 0.003 ml/cmH2O; and wild type, 0.034 ± 0.002 ml/cmH2O. A cohort of N4 B6 Hpgd−/− and wild-type mice was generated. This cohort still demonstrated a significant difference in baseline RL and a trend toward higher CDYN values. Mean baseline RL values were as follows: Hpgd−/−, 1.33 ± 0.04 cmH2O·s·ml−1; and wild type, 1.55 ± 0.07 cmH2O·s·ml−1. Mean baseline CDYN values were as follows: Hpgd−/−, 0.037 ± 0.002 ml/cmH2O; and wild type, 0.033 ± 0.001 ml/cmH2O. However, these differences were not maintained when baseline parameters were measured in congenic B6-Hpgd−/− pairs.
We next determined whether Hpgd−/− mice have altered airway responsiveness to methacholine, again beginning our studies with mice of mixed genetic background. Hpgd−/− mice have a significantly lower reactivity to methacholine compared with wild-type mice as measured by percentage of baseline RL on the mixed background (Fig. 4; P < 0.05, MANOVA all points; P < 0.05, t-test for 100 mg/ml dose). There was no difference in percentage of baseline CDYN in response to methacholine between Hpgd−/− and wild-type mice at any dose (Fig. 4). Both the protection from methacholine constriction as measured by percentage of baseline RL and the lack of protection as measured by percentage of baseline CDYN were also observed when the studies were repeated using N4 B6 Hpgd−/− and wild-type mice (Fig. 5).
To verify that the difference in airway reactivity in the Hpgd−/− mice was due to altered prostanoid metabolism at the time of challenge, we pretreated a mixed background cohort of Hpgd−/− and wild-type mice with indomethacin (a potent cyclooxygenase inhibitor) before measuring their airway reactivity. Mice received 10 mg/kg indomethacin intravenously at least 4 h before the assessment. Pretreatment with indomethacin did not alter the difference in baseline airway tone. Mean baseline RL values were as follows: Hpgd−/−, 1.08 ± 0.06 cmH2O·s·ml−1; and wild type, 1.40 ± 0.11 cmH2O·s·ml−1. Mean baseline CDYN values were as follows: Hpgd−/−, 0.049 ± 0.002 ml/cmH2O; and wild type, 0.038 ± 0.002 ml/cmH2O. After indomethacin pretreatment, there was no significant difference between the Hpgd−/− and wild-type mice in their response to methacholine as measured by percentage of baseline RL or CDYN (Fig. 6).
Evaluation of airway responsiveness in mice overexpressing mPGES1.
Although 15-PGDH has been shown to preferentially metabolize PGE2 compared with other prostanoids, we cannot rule out the possibility that the protection against methacholine-induced constriction is mediated by change in the levels of other prostanoids, and not solely an increase in PGE2 levels. To begin to address this, we generated a transgenic mouse line carrying a Ptges transgene under the regulation of a lung-specific SP-C promoter. This human SP-C promoter fragment has been shown in mice to direct gene expression to distal epithelial bronchiolar cells (12, 16, 39).
A schematic representation of the transgene expression construct is shown in Fig. 7A. Preliminary analysis was carried out with cohorts of transgenic and wild-type mice on the mixed background described in materials and methods. Animals carrying the transgene were obtained in the expected Mendelian ratios consistent with a single genomic locus integration, developed normally, and had no overt phenotype. Histological examination of the lungs using hematoxylin-eosin or Masson trichrome blue stain reveal no gross abnormality in lung tissue (data not shown). Figure 7B is a Northern blot of tissues from both transgene-positive and wild-type littermates showing lung-specific increase of Ptges transcript in transgene-positive animals. A cohort of N3 B6 hSP-C-Ptges and wild-type animals were assessed for gene expression using real-time PCR. The transgene-positive mice had significantly more Ptges transcript than wild-type littermates as measured relative to wild type with the use of quantitative RT-PCR (Fig. 7C). Measurement of PGE2 in whole lung homogenates showed a trend toward higher levels in the hSP-C-Ptges mice; however, these did not achieve statistical significance (data not shown). To determine whether expression of the transgene resulted in altered expression of genes involved in PGE2 synthesis or in the actions of PGE2 after receptor binding, we carried out real-time PCR analysis on RNA prepared from the lungs of four transgenic animals and five nontransgenic littermates. No significant difference in expression of the genes Hpgd, Ptgs1, Ptgs2, Ptger2, Pde4b, or β-actin was detected.
Basal tone and airway reactivity in SP-C-Ptges transgenic mice.
A cohort of hSP-C-Ptges and wild-type littermates was generated by breeding a chimera generated from 129/SvEv ES cells carrying the transgene with B6D2 purchased mice. Assessment of baseline lung parameters reveals that the transgenic mice did not significantly differ from the wild-type mice in baseline RL or CDYN values. Mean baseline RL values were as follows: hSP-C-Ptges, 1.37 ± 0.05 cmH2O·s·ml−1; and wild type, 1.29 ± 0.05 cmH2O·s·ml−1. Mean baseline CDYN values were as follows: hSP-C-Ptges, 0.035 ± 0.005 ml/cmH2O; and wild type, 0.037 ± 0.005 ml/cmH2O. This lack of difference was recapitulated with a cohort of hSP-C-Ptges and wild-type littermates generated by backcrossing transgene-positive animals with C57BL/6 purchased mice for three generations. Mean baseline RL values were as follows: hSP-C-Ptges, 1.47 ± 0.06 cmH2O·s·ml−1; and wild type, 1.53 ± 0.12 cmH2O·s·ml−1. Mean baseline CDYN values were as follows: hSP-C-Ptges, 0.029 ± 0.001 ml/cmH2O; and wild type, 0.028 ± 0.002 ml/cmH2O.
Generation of a dose-response curve to methacholine as measured by percentage of baseline RL showed that the transgenic mice have a significantly attenuated response to methacholine. However, unlike the Hpgd−/− mice, we also noted a significant attenuation in the percentage of baseline CDYN change in these mice (Fig. 8). These changes in mice of mixed genetic background were recapitulated in studies of the mice in which the transgene had been moved onto a more homogenous genetic background (Fig. 9).
To verify that the attenuation in responsiveness in the transgenic mice on the mixed background is due to altered prostanoid metabolism at the time of challenge, we pretreated a cohort of transgenic and wild-type mice with indomethacin before measuring their airway reactivity invasively. Mice received 10 mg/kg indomethacin intravenously at least 4 h before the experiment. Assessment of baseline lung parameters reveals that the indomethacin pretreatment did not significantly alter baseline lung parameters in either group (data not shown). Generation of a dose-response curve to methacholine as measured by percentage of baseline RL and CDYN demonstrated no significant difference between the groups when pretreated with indomethacin (Fig. 10).
Previous studies have shown that PGE2 can protect against bronchoconstrictive challenges induced in humans (10, 21, 26, 28). Studies using mice lacking each of the four known PGE2 receptors showed that these protective actions were mediated in large part by the EP2 receptor (8, 29, 37). We have shown in the present study that loss of the EP2 receptor does not alter the basal tone of the airways and does not alter the response of the mice to methacholine. This is in stark contrast to the observation that loss of β-adrenergic receptors leads to decreased responsiveness to methacholine challenge (19). Similarly, we have found that both baseline tone and response to methacholine were not altered in mice homozygous for a mutation in Ptges. We cannot rule out the possibility that loss of mPGES1 has only a small effect on the capacity of specific cellular compartments within the lung to produce PGE2. mPGES1 is one of a number of gene-encoding proteins believed to be capable of converting PGH2 to prostaglandin. These include mPGES1, mPGES2, GSTM2, GSTM3, and cPGES/p23 (1, 14, 25, 33–35, 38). To date, however, mPGES1 is the only enzyme that has been shown to contribute to PGE2 production in vivo (38). Because the phenotype of the Ptges−/− mice does not recapitulate all of the phenotypes observed in the various PGE2 receptor-deficient mice, presumably other pathways for the production of PGE2 exist in vivo. High levels of expression of this synthase are observed in the lung, and it is therefore reasonable to expect that it normally contributes to the production of PGE2 by this organ, despite the fact that we were unable to measure change in PGE2 levels in the healthy airway.
We utilized two different approaches to increase PGE2 levels: mice lacking PGDH and mice expressing a lung-specific mPGES1 transgene. First, we examined mice lacking the enzyme PGDH. This enzyme is expressed at extremely high levels in the lung and has been shown to actively convert PGE2 into the less active metabolite 15-hydroxy-PGE2. Consistent with this, we detected a small but significant increase in PGE2 levels in the Hpgd−/− mice.
Initially, the studies of the Hpgd−/− mice were carried out on mice of mixed genetic background or on experimental and control littermates generated from N4 B6 Hpgd+/− animals. In these studies a subtle but significant change in baseline lung mechanics was observed. Interestingly, this difference was not sensitive to indomethacin, suggesting that it was not directly dependent on PGE2 production at the time of assessment. We propose the possibility that these differences were the result of 129 alleles that continued to cosegregate with the Hpgd mutant allele. This interpretation was supported by our failure to observe these differences in the N7 generation. In contrast, all Hpgd−/− mice studied showed an attenuated response to methacholine-induced bronchoconstriction, and this protection was indomethacin sensitive. In these animals it is expected that the relative production of PGE2 by the various populations of cells will not be altered spatially but, rather, that the survival of the metabolite will be extended. However, PGDH also has been shown to metabolize other eicosanoids with a 15-hydroxyl group (7). It is therefore possible that the protection observed was caused by an increase in levels of some other prostanoid. Alternatively, the protective actions of increased PGE2 in the Hpgd−/− mice might be limited by slight increases in metabolites with bronchoconstrictive actions.
It was therefore important to verify our results using mice that we expected to differ only in PGE2 levels in the lungs. We therefore generated mice carrying a lung-specific mPGES1 transgene. These mice showed remarkable similarity to the Hpgd-deficient mice. No change in baseline parameters was observed. Transgenic mice were protected against methacholine-induced reactivity, and the change in RL was of a similar magnitude to that observed in the Ptges−/− mice. However, further analysis revealed that the hSP-C-Ptges transgenic mouse line additionally shows significant protection from changes in CDYN in response to methacholine. The hSP-C promoter used in generation of this mouse line has been shown by others to direct transgene expression predominately in the distal bronchiolar regions of the adult murine lung (12, 16, 39). Changes in RL have been postulated to reflect changes in central/proximal airways, whereas changes in CDYN reflect changes in peripheral/distal airways (6, 13, 15, 17, 32). This protection to changes in CDYN is presumably due to augmented PGE2 synthesis in the distal regions of the airway.
The protective actions of PGE2 are mediated in a large part by activation of the EP2 receptor. It is interesting to speculate that the observation that difference in the impact of increased activity of the EP2 pathway versus the impact of constitutive increase in the activity of the β-adrenergic receptor pathway is related to intrinsic differences in regulation of EP2 and β2-AR signaling. Penn et al. (27), using smooth muscle culture, have shown that ligand binding to the β2-AR leads to rapid desensitization via β-arrestin-mediated receptor internalization. In contrast, the desensitization of the EP2 receptor is limited, perhaps related to the short cytoplasmic tail and lack of sites suitable to phosphorylation by kinases compared with multiple sites in the β2-AR and EP4 proteins (24, 27, 31). It is possible that the continual desensitization of the β2-AR leads to induction of intracellular signals that bring about long-term adjustments in cell physiology. Studies in other cell systems have suggested that the binding of β-arrestin to phosphorylated G protein-coupled receptors also leads to new signals via the activation of various mitogen-activated protein kinases (30). The lack of EP2 receptor phosphorylation stimulated desensitization might prevent these adaptive changes in smooth muscle cells.
Previous studies demonstrated that the protective actions of PGE2 could be distinguished from the bronchoconstrictive/irritant action: the protective action was mediated largely by the EP2 receptor, whereas the EP1 and EP3 receptors were shown to be responsible for restricting airflow (8, 29, 37). These results suggested that EP2-specific agonists would provide the beneficial actions of PGE2 without the complications of bronchoconstriction observed in some patients with PGE2 inhalation. Our present studies further support the potential of the EP2 receptor as a target in the treatment of restricted air flow, suggesting that the desensitization and compensatory changes observed with β-adrenergic receptor agonist might not occur upon stimulation of this pathway.
This work was supported by National Heart, Lung, and Blood Institute Grant HL-68141 (to B. H. Koller) and HL-071802 (to S. L. Tilley), Cystic Fibrosis Foundation Grant Koller00Z0 (to B. H. Koller), and American Heart Association Grant 0415427U (to A. K. Lovgren).
We thank Amy Pace for help in reviewing the manuscript, John Snouwaert for assembly of the hSP-C-Ptges transgene, and Anne Latour for genotyping.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society