bronchorelaxation, including studies with human bronchus

Transcription

1 Br. J. Pharmaol. (1993), 18, '." Mamillan Press Ltd, 1993 Investigation into the role of phosphodiesterase IV in bronhorelaxation, inluding studies with human bronhus *J. Cortijo, J. Bou, J. Beleta, I. Cardeluis, J. Llenas, *E. Morillo & 'R.W. Gristwood Division of Biologial Sienes, Laboratorios Almirall, Cardener 68-74, 824-Barelona and *Department of Pharmaology, University of Valenia, 461-Valenia, Spain 1 We have investigated the role of yli nuleotide phosphodiesterase IV (PDE IV) in the relaxation of human bronhus and guinea-pig trahea in vitro and in guinea-pigs in vivo. 2 Funtional studies showed that the seletive PDE IV inhibitors, rolipram and denbufylline, relaxed human and guinea-pig preparations in vitro. 3 Two linially used xanthine non-seletive PDE inhibitors, theophylline and pentoxifylline, were also effetive in these preparations, but were muh less potent than the seletive agents used. 4 The rank order of poteny for the four PDE inhibitors in both speies was similar. 5 Biohemial studies indiated that PDE IV was the major PDE isoform present in the human bronhial tissue. PDEs I, II and V were also identified. 6 Theophylline and pentoxifylline were, as expeted, non-seletive inhibitors of the human enzymes, but there was a good orrelation between PDE IV inhibitory and bronhorelaxation potenies, suggesting that PDE IV inhibition is important for the linial bronhodilator ativities of the two xanthine ompounds. 7 We have onfirmed the ability of seletive PDE IV inhibitors to ause bronhodilatation in guineapigs in vivo. 8 We onlude that our study has provided further evidene that seletive PDE IV inhibitors ould at as bronhodilators in the lini. Keywords: Phosphodiesterase isoenzymes; PDE IV; human bronhus; theophylline; pentoxifylline; rolipram; denbufylline; bronhorelaxation Introdution There is urrently interest in the possible use of seletive inhibitors of yli nuleotide phosphodiesterases (PDEs) in the treatment of asthma (see e.g. Torphy & Undem, 1991; Giembyz & Dent, 1992). There are known to be at least five distint families of PDE isoenzymes, PDEs I to V, whih differ in their substrate speifiity and affinity as well as in their regulatory properties and tissue distribution (see Beavo & Houslay, 199). Of these, PDE IV inhibitors are reeiving speial interest as anti-asthmati agents due to evidene that these an at as both anti-inflammatory agents and bronhodilators in animal speies (see Niholson et al., 1991; Giembyz & Dent, 1992). There are, however, few linial data generally available onerning seletive PDE IV inhibitors in asthma. Theophylline, urrently widely used linially in asthma, is a weak non-seletive PDE inhibitor, but it is believed that PDE inhibitory ativity (inluding PDE IV inhibition) -may ontribute to both its bronhodilator and anti-inflammatory ativities in man (see e.g. Torphy & Undem, 1991; Gristwood et al., 1991). Theophylline, therefore, may be an important link ompound between in vitro human and animal pharmaologial studies and the potential linial effiay of PDE IV inhibitors. The purpose of the present investigation was to ompare, using human in vitro studies, the bronhorelaxant and PDE inhibitory ativities of two seletive PDE IV inhibitors, rolipram (Reeves et al., 1987) and denbufylline (Niholson et al., 1989), and two linially used xanthine non-seletive PDE inhibitors, theophylline (Persson, 1986) and pentoxifylline (Ward & Clissold, 1987). Isoprenaline was inluded as a standard bronhodilator agent. ' Author for orrespondene. A further important aim of the study was to ompare the human in vitro bronhodilator ativities of these agents with their in vitro and in vivo bronhodilator ativities in the guinea-pig, a ommonly used animal for models of human asthma. Methods Relaxant ativity in human isolated bronhus Lung tissue was obtained from 19 patients (18 males and 1 female, 35 to 78 years old, mean age 61.7 ± 2.8 years) who were undergoing surgery (Hospital de La Fe, Valenia, Spain) for lung arinoma. None of the patients had a history of asthma. After the resetion of one or more lung lobes, a piee of marosopially normal tissue was ut free and submerged in Krebs solution (for omposition see drugs and solutions) at 4 C for transport to the laboratory. One in the laboratory (Valenia) parts of the bronhus were disseted free from parenhymal lung tissue and preparations ut (3-4 mm length x 2-3 mm internal diameter) as previously desribed (Cortijo et al., 1992). Preparations were stored in Krebs solution, equilibrated with 5% CO2 in 2 at 4'C until used. Experiments were routinely ompleted within 24 h of initiating storage. For experiments, the bronhial rings were suspended on tissue hooks in 1 ml organ baths ontaining Krebs solution, gassed with 5% CO2 in 2 at- 37 C (ph 7.4). Eah preparation was onneted to a fore displaement transduer (Grass FTO3) and isometri tension hanges reorded on a Grass polygraph (model 7P). The preparations were equilibrated for 6-9 min with hanges in bath Krebs solution every 2 min before adding drugs.

2 PDE IV INHIBITORS A BRONCHORELAXATION 563 A load of 2 g was maintained throughout the equilibrium period and a stable resting level of tone was present immediately prior to drug administration. The effets of test drugs were investigated by adding umulative onentrations of these to the baths. Responses were allowed time to stabilize prior to inreasing the bath drug onentration (usually within 15 min of addition). Only one onentration-effet urve was onstruted with eah preparation. Experiments were terminated by the addition of theophylline, 1 x IO-I M, the effet of whih was taken to represent the maximum relaxation possible in the tissue. Changes in fore were measured from isometri reordings and expressed in g. The maximum response indued with a relaxant agent (Em.x) and the molar onentration required to produe 25% (EC25) and 5% (EC5) of maximal relaxation was alulated by linear regression from the individual onentration-effet urves. Relaxant ativity in guinea-pig isolated trahea Guinea-pigs (male Dunkin-Hartley) of weight range 4-6 g were maintained without food but with free aess to water for 16 h before experimentation. Animals were then killed by ranial perussion, the trahea disseted free and plaed into Krebs solution at 4C. The trahea was ut into preparations onsisting of 2 artilaginous rings whih were then ut through the artilaginous zones. The preparations were suspended in 3 ml organ baths ontaining Krebs solution at 37TC gassed with 5% CO2 in 2. These were allowed at least 1 h to equilibrate during whih time the resting tension was maintained at 1 g. A ontrol response to isoprenaline (1 x 1-7 M) was then obtained, to indiate the maximum relaxation possible in this tissue. The tissues were then washed and allowed 3 min to re-equilibrate before one of the drugs under study was added by a umulative onentration proedure. Responses were allowed to stabilize (usually within 15 min of drug addition). Tension was reorded with isometri transduers (Letia TRl 1) onto a Letia polygraph (model 4). Changes in tension were measured, and EC25 and EC5 values alulated, as desribed for human preparations. Bronhodilator ativity in anaesthetized guinea-pigs Animals of weight range 4-5 g were anaesthetized with sodium pentobarbitone 6mgkg-', i.p. A traheotomy was performed and a polyethylene tube inserted into the ut trahea, onneted to a respirator (Ugo Basile 725) and a side arm onneted to a pressure transduer (Letia TRA21). The left arotid artery was annulated for blood pressure and heart rate reordings and the right jugular vein for drug administration. After a 15 min stabilization period, a ontinual infusion (Braun infusion pump) of histamine (1 pg ml-', 2-6 ml h-') suffiient to elevate insufflation pressure by 15-2% was administered via the left jugular vein and maintained throughout the experiment. The histamine infusion was ontinued for 45 min, by whih time a steady state bronhoonstriaor response had been observed, before one of the test drugs was given by i.v. bolus injetions using an asending dose sheme. The maximum bronhodilatation produed by eah drug was measured and expressed as the % inhibition of the histamine-indued bronhoonstritor response. Extration, separation and haraterization of PDE isoenzymes from human bronhus Individual human bronhi, weighing g were homogenized for 6 s, with an Ultraturrax at 9 r.p.m. in 5 volumes of ie-old buffer A (2 mm Bis Tris, ph 6.5, ontaining 5 mm sodium aetate, 2 mm benzamidine, 2 mm EDTA, 5 mm P-meraptoethanol and 5 mm phenylmethylsulphonylfluoride (PMSF)). The homogenate was entrifuged at 15 g for 1 min and the lear supernatant was filtered through.22 1m Millex filters. The sample was injeted into a MONO-Q HR 5/5 olumn (1 ml of gel bed, Pharmaia) attahed to an FPLC hromatography system and equilibrated in the same buffer. After washing with 15 ml of buffer A, the PDEs were eluted by developing a 2 ml linear sodium aetate gradient from 5 to 1 mm in buffer A. Flow rate was 1 ml min-' throughout. Frations of.5 ml were olleted, analyzed and stored as previously desribed (Gristwood et al., 1992). Moleular weights of purified PDEs were determined by gel filtration on a Superose 12 HR 1/3 olumn (Pharmaia) attahed to an FPLC system. The olumn was equilibrated at room temperature in 2mM Bis Tris, 15mm sodium aetate, 2 mm EDTA, 5 mm P-meraptoethanol, 2 mm benzamidine ph 6.5 buffer. The olumn was alibrated with proteins of known moleular weight by running.1 ml samples at a flow rate of.4 ml min-'. Elution volumes were determined by following the absorbane at 28 nm on line, and the obtained Kay values were orrelated with the logarithm of their moleular weights. Column void volume was determined with Blue Dextran 2 under the same onditions as the standards. Phosphodiesterase samples obtained from the ion exhange hromatography were also run in a similar manner and.2 ml frations were olleted. Elution volumes for the different enzymes were determined by assaying the enzyme ativity in the frations. Moleular weights were alulated by interpolation of the Kav values on the regression line obtained for the standards. Cyli nuleotide PDEs were assayed following the proedure of Thompson & Strada (1984). Inhibition assays were run in dupliate at 3 C for 2 min at a substrate onentration of.25 JM. The substrate was adenosine 3':5'-yli monophosphate (yli AMP) unless otherwise stated. PDE II was assayed in the presene of 5 fim unlabelled guanosine 3':5'-yli monophosphate (yli GMP). For eah of the assayed drugs, 6 to 8 different onentrations were tested in at least two different enzyme preparations. ICW values were obtained by non-linear regression using InPLot, GraphPad Software on an IBM omputer. Drugs, reagents and solutions The following drugs were used: zaprinast (a gift from Rhone- Poulen Rorer U.K.), SK&F 9412 (a gift from SK&B Ltd, Welwyn, U.K.), isoprenaline sulphate, obtained from Boehringer Ingelheim (Germany) and theophylline, purhased from Sigma-Aldrih Quimia S.A. (Madrid, Spain). Denbufylline, rolipram and pentoxifylline were synthesized in the Department of Chemial Synthesis, Almirall. Calmodulin was obtained from Boehringer Mannheim (Barelona, Spain). Benzamidine, histamine, PMSF, yli GMP and yli AMP were from Sigma-Aldrih Quimia, S.A. (Madrid, Spain). The low and high moleular weight markers and Blue Dextran 2 were from Pharmaia Iberia (Barelona, Spain). [8-3H]-adenosine 3':5'-yli monophosphate and [8-3H]-guanosine 3':5'-yli monophosphate were from Amersham International (U.K.). The drugs used for the biohemial studies were dissolved in dilute NaOH, dimethylsulphoxide (DMSO) or water. Drug vehiles at onentrations employed did not affet enzyme ativities. For the pharmaologial studies, isoprenaline was dissolved in distilled water ontaining.57 mm asorbi aid. Stok solutions of denbufylline were prepared in 4% polyethyleneglyol 3 and rolipram in 2% polyethyleneglyol 3. Theophylline and pentoxifylline were dissolved in water. Subsequent dilutions of drugs were made in the Krebs solution whih had the following omposition (mm): NaCl 118, KCl 4.7, CaCl2 2.5, MgSO4 1.6, NaHPO4 1.2, NaHCO3 25 and gluose 11.

3 564 J. CORTIJO et al. Statistis Values are given as mean ± s.e. mean. Statistial analysis of results was arried out by analysis of variane (ANOVA) followed by Dunan's test or by Student's t test for paired or unpaired data as appropriate. Results Bronhorelaxant ativity in human bronhus All drugs tested aused onentration-dependent inhibition of the spontaneous tone of human bronhi as shown in Figure 1. All produed full relaxation, i.e. the maximal relaxation was not signifiantly different from that obtained with theophylline, 1 x 1-3 M, see Table 1. The threshold onentrations for isoprenaline, rolipram, pentoxifylline and denbufylline were similar at between 1O-9M and 1-8 M, whereas the threshold for theophylline was higher, as shown in Figure 1. Considering both EC25 and EC5 values, the poteny order of drugs for relaxation was isoprenaline > rolipram = denbufylline > pentoxifylline> theophylline, as indiated in Table 1. One further important observation was that in preparations with spontaneous tone, rolipram, denbufylline, and pentoxifylline all produed lear biphasi onentrationresponse urves (see Figure 1). Bronhorelaxant ativity in guinea-pigs in vitro All drugs produed onentration-dependent relaxation of the guinea-pig traheal preparations, as shown in Figure 2. The maximum observed relaxations were as follows; isoprenaline.67 ±.3 g, rolipram.44 ±.4 g, denbufylline.4 ±.4 g, theophylline.66 ±.3 g and pentoxifylline.34±.7g, n=5-12. In terms of threshold onentrations there was a lear order of isoprenaline <denbufylline = rolipram < pentoxifylline = theophylline. EC25 and EC~v values for the ompounds on guinea-pig trahea are shown in Table 2. For Eq5 values the poteny order was isoprenaline > rolipram = denbufylline :' pentoxifylline> theophylline. For EC3 values the order was now isoprenaline W denbufylline = rolipram > pentoxifylline> theophylline. Of all agents tested only rolipram produed a lear biphasi onentration-response urve (Figure 2). Bronhodilator ativity in anaesthetized guinea-pigs o Or x (U E w Iom CC 'm 6 8-1L og1o drug onentration (M) Figure 1 Relaxant effets of theophylline (A), pentoxifylline (U), denbufylline (), rolipram (V) and isoprenaline () in human bronhi with spontaneous tone. Points are means ± s.e.mean (vertial bars). n values are indiated in Table 1. x E w I- F._ x a) : All drugs produed dose-dependent inhibitions of the histamine-indued bronhospasm, as shown in Figure 3. In terms of threshold doses the order was isoprenaline> rolipram> denbufylline > pentoxifylline = theophylline. In order to ompare drug potenies, doses ausing a mean derease of 5% of the histamine-indued bronhoonstrition were al- 4 _ 6 _ 8-1L I I I loglo drug onentration (M) Figure 2 Relaxant effets of theophylline (A), pentoxifylline (U), denbufylline (), rolipram (V) and isoprenaline () in guinea-pig traheal ring preparations. Points are means ± s.e.mean (vertial bars). n values are indiated in Table 2. Table 1 Relaxant effets of yli nuleotide phosphodiesterase preparations Theophylline Pentoxifylline Denbufylline Rolipram Isoprenaline - log EC5 -log EC25 p/n (M) (M) 1/6 11/6 9/5 9/6 8/ ±.13* 5.3 ± ± ± ±.3t 4.98 ±.3$ 6.1 ± ± ± ±.2t (PDE) inhibitors on spontaneous tone of human bronhial Emu (g) 1.53 ± ± ± ± ±.22 Maximal relaxation (g) indued by theophylline (lx 1-3 M) 1.52 ± ± ± ±.17 Values are means ± s.e.mean. p represents the number of preparations and n the number of patients. *P <.5 from the other agents; tp <.5 from theophylline and pentoxifylline; $P <.5 from the other agents exept pentoxifylline.

4 PDE IV INHIBITORS A BRONCHORELAXATION 565 Table 2 Relaxant effets of yli nuleotide phosphodiesterase (PDE) inhibitors on spontaneous tone of guinea-pig traheal preparations Theophylline Pentoxifylline Denbufylline Rolipram Isoprenaline -log EC5 -log EC25 n (M) (M) ± ± ±.17* 5.3 ±.2* ±.13* 8.1 ±.24* ±.23* 8.7 ±.17* ±.4* 9.75 ±.5* *P<.5 ompared with theophylline. Cl (A LU al ) (U E o C._._ ) C U._ L._C Co r I- 21 Fration number Figure 4 Representative elution profile of yli nuleotide phosphodiesterase (PDE) ativities from human bronhus on a MONO Q ion exhange olumn. The low speed entrifugation supernatant (1ml) from one individual sample was hromatographed as desribed in the text. Frations (.5 ml) were olleted and assayed for PDE ativity in the following onditions: I gm yli AMP (@), 1 gm yli AMP plus 1 gm old yli GMP (U) and 1 JM yli GMP (A). The 4 peaks olleted were: frations 37 to 53 (A), frations 56 to 62 (B), frations 64 to 68 (C) and frations 76 to 85 (D). For the ordinate sale, I unit of enzyme ativity is defined as the amount hydrolysing 1 pmol of substrate per min. I5 /1 /; of its ativity using kineti and regulatory properties suggested that peak B orresponded to a form of PDE IV. IC5 values of 1.1 and 1.3 gm were obtained for this peak using denbufylline and rolipram respetively. The low moleular weight obtained for peak B suggested that it ould have been derived from native PDE IV by limited proteolysis. This, and the small amounts of peak B obtained, led us to exlude this enzyme from the biohemial o studies with the rest of inhibitors Peak C represented 16.7 ± 3.4% of the total ativity mea- Dose (,ug kg-1, i.v.) sured with 1 JAM yli AMP as substrate and the data pree 3 Inhibition of histamine-indued bronhoonstrition in sented in Table 3 indiate that this enzyme is the yli GMP Figur anaesthetized guinea-pigs by isoprenaline (), rolipram (V), den- stimulated form (PDE II). bufyllmine (), pentoxifylline (-) and theophylline (A). Points are Peak D was the major peak in all samples (56 ± 1.7% of mean, s ± s.e.mean (vertial bars). n = 4-6. total). Its haraterization, shown in Table 3, indiates that this peak is also PDE IV. Gel filtration studies indiated a moleular weight of 823 ± 56, as well as the homogeneity of the enzyme, with little ontamination by other PDE isoenzymes (see Figure 5). ulat;ed. These were as follows in pg kg' i.v.: isoprenaline No lear evidene of yli GMP inhibited PDE (PDE III).45: ±.1, rolipram, 4 ± 1, denbufylline 69 ± 2, pentoxifyl- was obtained either from the hromatograms or from the use line 4534 ± 86 and theophylline 5842 ± 911 (n = 4-6). of the seletive PDE III inhibitor SK&F 9412 (Gristwood et j Isolation and haraterization of the human bronhial phosphodiesterases A representative hromatogram of the PDEs isolated from human bronhi by ion exhange is shown in Figure 4. In 5 out of 6 individual preparations, 4 peaks of PDE ativity were observed. The haraterization of the peaks is shown in Table 3. When assayed with 1 JAM yli AMP as substrate, peak A was the most variable in size, ranging from 8 to 21% of total PDE ativity. When submitted to gel filtration, peak A was resolved into two distint peaks. This, taken together with its differential sensitivity to zaprinast when yli GMP was used as substrate instead of yli AMP and its marginal ativation by Ca2+/almodulin (Table 3) indiates that peak A is a mixture of PDE I and PDE V. Peak B was always the smallest of all and ould not be learly identified in one of the samples. The haraterization al., 1986) on seleted PDE frations. Effets of the drugs on the isolated isoenzymes The effets of the PDE inhibitors on the three major peaks of human bronhial PDE ativities are shown in Table 4. As expeted, rolipram and denbufylline displayed seletivity for Peak D, whereas theophylline and pentoxifylline inhibited all three peaks. The poteny order of the inhibitors on peak D was denbufylline = rolipram >- pentoxifylline > theophylline. Disussion The results from this study have onfirmed the relaxant ativity of theophylline in human respiratory musle in vitro. This is in agreement with previous findings (Cortijo et al., 1992). Furthermore, we have shown that pentoxifylline, another linially used xanthine, also has relaxant properties

5 566 J. CORTIJO et al. Table 3 Charaterization of the human bronhial phosphodiesterases separated by ion exhange hromatography Km yli AMP Km yli GMP Ca2+/Calmodulin (% ativation) Cyli GMP 5pm (% ativation) Mol. weight (KD) Zaprinasta (IC5) Zaprinastb (IC5) SK&F 9412 (IC5) Peak A Peak B Peak C Peak D >5 > >5 77* 9.6** > > >5 All onentrations are in gm. acyli AMP as substrate; bcyli GMP as substrate., not determined., no effet. * indiates %s beause of non hyperboli kinetis. **Km value for yli AMP in the presene of yli GMP. Dispersion values, lower than 1% in all ases, have been omitted for larity. 6 t.4 UV Fration number Figure 5 Gel filtration profile on Superose 12 of an aliquot of the 4th peak obtained from the ion exhange hromatography (peak D). The arrow indiates olumn void volume. See text for details. For the ordinate, 1 unit of enzyme ativity is defined as the amount hydrolysing 1 tsmol of substrate per min. in human bronhial tissue in vitro. Although pentoxifylline is urrently used linially as a haemorheologial agent (Ward & Clissold, 1987), linial studies have shown that the drug has bronhodilator ativity in man (Nolte, 1971). We onsider that the onentrations of both theophylline and pentoxifylline used in this study are relevant to the plasma onentrations of these drugs ahieved linially. Thus, in the ase of theophylline, therapeuti plasma onentrations are between 5 and 1 LM, 6% of whih is protein bound (Svedmyr, 1988). Therefore, free plasma onentrations would not be very different from the in vitro EC25 (;1IOM) and EC% (,8 gm) values found in this study. For pentoxifylline, mean peak plasma onentrations in man (following a standard 4 mg dose) are approximately 5 LM (plasma binding is low, see Ward & Clissold, 1987), again a value lose to the in vitro EC25 (;1 gm) and EC5 (;5 gm) found for pentoxifylline in this study. Considering responses in human preparations at the EC5 level, both theophylline and pentoxifylline were relatively weakly ative on the tissues, whereas the seletive PDE IV inhibitors tested, rolipram and denbufylline, were muh more potent relaxants. Indeed rolipram, at the response level of EC25, was lose to isoprenaline in its poteny. These data onfirm the bronhodilator potential of PDE IV inhibitors in man, partiularly in terms of poteny and maximal effet. Consistent with the bronhorelaxant ativities of denbufylline and rolipram, our biohemial studies onfirmed the presene of PDE IV in human bronhi and furthermore, indiated that it is the major PDE isoform present in this tissue. There were in fat two peaks of ativity that ould be haraterized as PDE IV, (Peaks B and D) although we believe, based on their properties, that Peak B ould be a proteolysed form of the enzyme. Thus, whereas peak D eluted in the position typial for PDE IV both from human (Reeves et al., 1987) and animal soures (Gristwood et al., 1992) and had a moleular weight lose to that of loned human PDE IV (Livi et al., 199) peak B was smaller and eluted before PDE II in the hromatogram. A omparison of the inhibitory potenies of the four PDE inhibitors on PDE IV (Peak D ativity) and relaxant ativity in human bronhi (Tables 1 and 4) indiated a good orrelation, and this suggests that PDE IV inhibitory ativity for both theophylline and pentoxifylline is at least partly responsible for their in vitro bronhorelaxant ativities in human tissue. We annot, of ourse, exlude the possibility that for the xanthine ompounds, inhibition of the other PDE isoenzymes present, ontributed to their relaxant ativities, but no orrelation ould be found for the tested ompounds between inhibition of the other peaks and their relaxant potenies, partiularly in the ase of rolipram. Reently Shahid et al. (1992) have shown that up to six different peaks of PDE ativity ould be resolved by ion exhange hromatography in human bronhi. Although the hromatographi profile desribed by these authors appears to be more omplex than that found in the present paper, no major disrepanies an be pereived. As with our samples, PDE IV was the main ativity hydrolyzing yli AMP in the study of Shahid et al. (1992). PDE V and two slightly Table 4 Effet of the ompounds under study on the phosphodiesterase peaks isolated by ion exhange hromatography Peak D Theophylline Pentoxifylline Denbufylline Rolipram Peak A 3.75 ± ± ± ±.5 Peak C 3.26 ± ± ± ± ± ± ± ±.1 For eah drug 6-8 onentrations were tested in dupliate for are at least two different samples. Values are -log IC5 in molar onentration ± s.e.mean.

6 PDE IV INHIBITORS A BRONCHORELAXATION 567 different forms of PDE I ould be deteted by the authors eluting at low salt onentrations. These math the enzymes that were identified in our peak A. PDE II was also deteted. The main differene found was the presene of a minor form of PDE IV eluting at a different position from that of peak B in the study of Shahid et al. (1992). It was interesting that another major PDE isoenzyme, PDE III, was not identified in our biohemial experiments. This is in agreement with reent data (Shahid et al., 1992). Consistent with this, it has reently been shown that seletive PDE III inhibitors, unlike seletive PDE IV inhibitors, did not indue inreases in intraellular yli AMP ontent in human ultured traheal smooth musle ells (Hall et al., 1992). These data taken together suggest that PDE III inhibitors may not have important bronhodilator ations in man. As in human preparations, all PDE inhibitors indued onentration-dependent relaxation of guinea-pig, traheal preparations with similar maximal effets. Furthermore, the rank order of poteny of the drugs was similar for the two speies. The only major quantitative differene between guinea-pig and human tissues was that isoprenaline was almost 2 orders of magnitude more potent on guinea-pig preparations. The potenies of the PDE inhibitors, however, appeared similar. We did not investigate PDE enzymes in guinea-pig trahea in the present study. Biohemial studies have been previously reported in the literature and it is known that PDE IV is present in this tissue (Silver et al., 1988; Takagi et al., 1992). Unlike human bronhus, guinea-pig trahea also has PDE III present (Silver et al., 1988), and PDE III inhibitors are effetive bronhodilators in the guinea-pig both in vitro (Bryson & Roger, 1987) and in vivo (Gristwood & Sampford, 1987). It is therefore, possible that both PDE III and IV inhibition ontributed to the bronhorelaxant ativity of theophylline and pentoxifylline in the guinea-pig. Rolipram and denbufylline both demonstrated potent dose-related bronhodilator ativity in anaesthetized guineapigs. This onfirms that seletive PDE IV inhibitors an at as bronhodilators in vivo. In this in vivo preparation there was a differene in poteny between rolipram and denbufylline that was not indiated by the guinea-pig in vitro studies. The reason for this is not known, but may have been due to Referenes BEAVO, J.A. & HOUSLAY, M.D. (199). Isoenzymes of Cyli Nuleotide Phosphodiesterases. Chihester: John Wiliey & Sons. BRYSON, S.E. & ROGER, I.W. (1987). Effets of phosphodiesterase inhibitors on normal and hemially skinned isolated smooth musle. Br. J. Pharmaol., 92, CORTIJO, J., SARRIA, B., PEDROS, C., PERPIIA, M., PARIS, F. & MORCILLO, E. (1992). The relaxant effets of romakalim (BRL 34915) on human isolated airway smooth musle. Naunyn- Shmiedebergs Arh. Pharmaol., 346, GIEMBYCZ, M.A. & DENT, G. (1992). Prospets for seletive nuleotide phosphodiesterase inhibitors in the treatment of bronhial asthma. Clin. Exp. Allergy, 22, GRISTWOOD, R.W., BELETA, J., BOU, J., CARDELUS, I., FERNAN- DEZ, A.G., LLENAS, J. & BERGA, P. (1992). Studies on the ardia ations of flosequinan in vitro. Br. J. Pharmaol., 15, GRISTWOOD, R.W., EDEN, R.J., OWEN, D.A.A. & TAYLOR, E.M. (1986). Pharmaologial studies with SK&F 9412 a novel positive inotropi agent with vasodilator ativity. J. Pharm. Pharmaol., 381, GRISTWOOD, R.W., LLUPIA, J., FERNAEZ, A.G. & BERGA, P. (1991). Effets of theophylline ompared with prednisolone on late phase airway leukoyte infiltration in guinea-pigs. Int. Arh. Allergy Appl. Immunol., 94, GRISTWOOD, R.W. & SAMPFORD, K.A. (1987). Inhibition of histamine indued bronhoonstrition by SK&F 94836, salbutamol and theophylline in the anaesthetised guinea-pig. Br. J. Pharmaol., 92, 631P. HALL, I.P., TOWNSE, P., DAYKIN, K. & WIDDOP, S. (1992). Control of tissue yli AMP ontent in primary ultures of human airway smooth musle ells. Br. J. Pharmaol, 15, 71P. pharmaokineti onsiderations. Nevertheless, it is lear that both seletive PDE IV inhibitors were muh more potent than either pentoxifylline or theophylline in vivo. The onsisteny of the bronhorelaxant effiay of the PDE IV inhibitors and the xanthines in human bronhus and guinea-pig trahea in vitro oupled with their effiay in the guinea-pig in vivo, and evidene that theophylline and pentoxifylline at as bronhodilators in man, all lead us to the onlusion that seletive PDE IV inhibitors will have bronhodilator ativity in man in vivo. One additional finding of interest was that in human preparations, rolipram, denbufylline, pentoxifylline and theophylline all produed onentration-response urves whih appeared somewhat biphasi in nature. This was partiularly so for the rolipram urve whih had evidene of a plateau between 1-6M and 1-5M. A biphasi onentration-response urve for rolipram was also evident in guinea-pig traheal preparations, where there was evidene of a plateau between 1-8 and 1-6 M. Although a definite explanation for these findings annot be given, the possibility of the involvement of another PDE isoenzyme or isoenzymes in the relaxation of human tissue annot be exluded. If that were the ase, the most seletive ompound of those tested (rolipram) might be expeted to have the most evident plateau, as indeed is the ase. In onlusion, our study has indiated the importane of PDE IV in the relaxation of human bronhial smooth musle in vitro. The data obtained indiate that the inhibition of this isoenzyme is important for the bronhorelaxation ativities of rolipram, denbufylline as well as theophylline and pentoxifylline both in guinea-pig and man and, further, that seletive PDE IV inhibitors ould be linially useful bronhodilators in man. The present work was supported in part by grants from C.I.C.Y.T. (FAR-9-68; FAR ) of Ministerio de Industria y Energia (Spain). The authors are indebted to Dr F. Paris (Head of the Servie of Thorai Surgery, Hospital La Fe, Valenia, Spain) and to the surgial team for making the human lung tissue available to us. The authors also wish to thank C. Cabello, R. Caballer and J. Mafie for tehnial assistane and A. Estelles for typing the manusript. LIVI, G.P., KMETZ, P., MEGAN, M., MCHALE, M.M., CIESLINSKI, L.B., SATHE, G.M., TAYLOR, D.P., DAVIS, R.L., TORPHY, T.J. & BALCAREK, J.M. (199). Cloning and expression of DNA for a human low-km, rolipram-sensitive yli AMP phosphodiesterase. Mol. Cell. Biol., 1, 6: NICHOLSON, C.D., CHALLIS, R.A.J. & SHAHID, M. (1991). Differential modulation of tissue funtion and therapeuti potential of seletive inhibitors of yli nuleotide phosphodiesterase isoenzymes. Trends Pharmaol. Si., 12, NICHOLSON, C.D., JACKMAN, S.A. & WILKE, R. (1989). The ability of denbufylline to inhibit yli nuleotide phosphodiesterase and its affinity for adenosine reeptors and the adenosine re-uptake site. Br. J. Pharmaol., 97, NOLTE, D. (1971). Lungenfunktionsuntersuhungen zur bronhospasmolytishen Wirkung von 3,7-Dimethyl-1(5-oxo-hexyl)-xanthin. Arzneimittel-Forshung., 21, PERSSON, C.G.A. (1986). Overview of effets of theophylline. J. Allergy Clin. Immunol., 78, REEVES, M.L., LEIGH, B.K. & ENGLA, P.J. (1987). The identifiation of a new yli nuleotide phosphodiesterase ativity in human and guinea-pig ardia ventrile. Biohem. J., 241, SHAHID, M., PHILPOT, A.J., DE BOER, J., ZAAGSMA, J. & NICHOL- SON, C.D. (1992). Cyli nuleotide phosphodiesterase (PDE) isoenzyme ativities in human peripheral bronhi. Br. J. Pharmaol., 15, 129P.

7 568 J. CORTIJO et al. SILVER, P.J., HAMEL, L.T., PERRO, M.H., BENTLEY, R.G., BUSH- OVER, C.R. & EVANS, D.B. (1988). Differential pharmaologi sensitivity of yli nuleotide phosphodiesterase isoenzymes isolated from ardia musle, arterial and airway smooth musle. Eur. J. Pharmaol., 15, SVEDMYR, N. (1988). Asthma: Basi Mehanisms and Clinial Management. ed. Barnes, P.J., Rogers, I.W. & Thompson, N.C. pp London: Aademi Press. TAKAGI, K., OGAWA, K., TANAKA, H., SATAKE, T., WATANABE, Y., CHIJIWA, T. & HIDAKA, H. (1992). Relaxant effets of various xanthine derivatives. Relationship to yli nuleotide phosphodiesterase inhibition. Adv. Seond Messenger Phosphorylation Res., 25, THOMPSON, W.J. & STRADA, S.J. (1984). Cyli nuleotide phosphodiesterase (PDE). In Methods of Enzymati Analysis. ed. Bergmayer, H.U. 3rd., Vol. IV, pp Weinheim Verlag Chemie. TORPHY, T.J. & UEM, B.J. (1991). Phosphodiesterase inhibitors: new opportunities for the treatment of asthma. Thorax, 416, WARD, A. & CLISSOLD, S.P. (1987). Pentoxifylline. A review of its pharmaodynami and pharmaokineti properties, and its therapeuti effiay. Drugs, 34, (Reeived July 1, 1992 Revised September 21, 1992 Aepted Otober 14, 1992)