Thorax New drugsreview · phosphodiesterases are responsible for cyclic nucleotide hydrolysis in different tissues. Attempts are being made to increase tissue selectivity by synthesising

Thorax 1991;46:512-523

New drugs review

Phosphodiesterase inhibitors: new opportunitiesfor the treatment of asthma

Theodore J Torphy, Bradley J Undem

Theophylline has been a mainstay in the treat-ment of asthma for over 50 years. Its thera-peutic value stems from a combination of anti-inflammatory and bronchodilator activities inaddition to its ability to increase diaphrag-matic contractility.14 Although the phar-macological effects of theophylline have beenstudied extensively both in the laboratory andin the clinic, the molecular mechanisms res-

ponsible for its activity in asthma remain illdefined. In fact, several cellular activitiesprobably contribute to its action, includingcyclic nucleotide phosphodiesterase inhibi-tion,' 5 adenosine receptor antagonism,6stimulation of catecholamine release,7 and a

poorly understood ability to increase the num-ber and activity of suppressor T lym-phocytes.8Of theophylline's many activities, phos-

phodiesterase inhibition may be the mostimportant. Hypothetically, at least twotherapeutically beneficial effects could resultfrom inhibition of phosphodiesterase activityand the consequent rise in intracellular aden-osine 3',5'-monophosphate (cAMP) or guan-

Cyclic AMP

osine 3',5'-monophosphate (cGMP) concen-

trations in key cells concerned in the patho-physiology of asthma (figure). Firstly, bothcAMP and cGMP mediate airway smoothmuscle relaxation,9 10 so a rise in either of thesesecond messengers in airway smooth muscleshould result in bronchodilatation. Secondly,the ability of PDE inhibitors to increase thecAMP content of inflammatory cells leads toan inhibition of cell activation.5 1112 In princi-ple, then, PDE inhibitors should possess anti-inflammatory activity.Although it is useful in the treatment of

asthma, the value of theophylline is limited bya narrow therapeutic index and a wide range

of gastrointestinal, central nervous system,and cardiovascular side effects.`4 These sideeffects are generally ascribed to its lack ofselectivity. Of particular importance in thisrespect is theophylline's ability to antagoniseadenosine receptors and inhibit phosphodies-terase activity in inappropriate tissues (forexample, gastrointestinal, central nervous sys-

tem, cardiovascular).2 13 The development ofdrugs that couple the efficacy of theophylline

CyclicGMP|

Hormone

Receptor

AdenylateCyclase

Department ofPharmacology, SmithKline BeechamPharmaceuticals,King of Prussia,Pennsylvania 19406T J TorphyDivision of ClinicalImmunology, JohnsHopkins UniversitySchool of Medicine,Baltimore, Maryland21224, USAB J Undem

PDE

ATP o cAMP O 5'-AMP

cA-PK f. cA-PK(inactive) (active)

Hormone

NVD or EDRF Receptor

PDE GuanylateCyclase

5'-GMP cGMP GTP

cG-PK | cG-PK(active) (inactive)

INFLAMMATORY CELL ACTIVITY AIRWAY SMOOTH MUSCLE TONE

Role of cyclic nucleotides in the function of inflammatory cells and airway smooth muscle. See text for details. PDE-phosphodiesterase; cA-PK-cAMP dependent protein kinase; cG-PK-cGMP dependent protein kinase; NVD-nitrovasodilators; EDRF-endothelium derived relaxant factor.

I I

512

on August 28, 2020 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.46.7.512 on 1 July 1991. D

ownloaded from

http://thorax.bmj.com/

Phosphodiesterase inhibitors: new opportunities for the treatment of asthma

with an improved side effect profile and an

increased therapeutic index would be an

important advance in the treatment of asthma.Many pharmaceutical companies have attem-pted to improve the therapeutic profile oftheophylline by synthesising and evaluatingnew xanthine analogues without specificallyfocusing on the phosphodiesterase inhibitoryactivity. So far these efforts have not beensuccessful.An alternative approach toward developing

an "improved theophylline" has emergedrecently. This approach is based on thepremise that multiple, distinct forms of phos-phodiesterases exist and that differentphosphodiesterases are responsible for cyclicnucleotide hydrolysis in different tissues.Attempts are being made to increase tissueselectivity by synthesising inhibitors that are

targeted against the major phosphodiesteraseisozyme or isozymes responsible for regulat-ing cAMP and cGMP hydrolysis in airwaysmooth muscle and inflammatory cells.Although some of these inhibitors haveretained the xanthine nucleus, others have a

completely different chemical structure. It ishoped that the phosphodiesterase inhibitorsthat emerge will retain at least some of theanti-inflammatory or bronchodilator activity(or both) of theophylline but that they will beless likely to produce gastrointestinal, centralnervous system, and cardiovascular side effectsby virtue of their poor activity against adeno-sine receptors and the phosphodiesterasespresent in gut, brain, and heart. The purposeof this review is to explain the theoreticalfoundation for the new generation of phos-phodiesterase inhibitors in the treatment ofasthma.

Phosphodiesterase isozymesHydrolysis of the 3'-phosphoester bond on

cAMP or cGMP converts these second mes-

sengers to their inactive 5'-nucleotidemetabolites (figure). This reaction is catalysedby a heterogeneous group of isozymes collec-tively known as phosphodiesterases. The gen-eral characteristics of phosphodiesteraseisozymes are shown in table 1. It is importantto note that the kinetic values shown are

approximations. Actual values may vary sub-stantially, depending on the tissues, species,isolation procedure, and enzyme purity. Theenzymes differ nevertheless in their kinetic andphysical characteristics, substrate (cAMP or

cGMP) selectivities, sensitivity to endogenousactivators and inhibitors, susceptibility andresponse to phosphorylation by protein kin-ases, tissue distribution, and subcellular local-isation."4 5 Subclasses of isozymes may existfor each of the five families of phosphodies-terases shown in table 1.L' Recent evidencesuggests that the five families of phosphodies-terase isozymes are coded for by distinct butperhaps related genes.'l20

Phosphodiesterase (PDE)I, or the Ca2 /cal-modulin stimulated phosphodiesterase, as itsname implies, is stimulated by Ca'+/cal-modulin. At least two general forms of theenzyme have been identified; the first (PDE I,)

Table 1 Characteristics ofphosphodiesterase isozymes

K,,, (pmol/l)t

Family* Isozyme* cAMP cGMP

Il+ Ca2` /CaM stimulated 30 3I+ Ca2 /CaM stimulated 1 2II cGMP stimulated 50 50III cGMP inhibited 0-2 0.3IV cAMP specific 2 100V cGMP specific 150 1

*Nomenclature from ref 15.tNumbers represent approximate values (see ref 14); kineticcharacteristics vary substantially, depending on species, tissue,isolation procedure, and enzyme purity.+The subsets of PDE I are arbitrarily designated l and 1$.

-has a much greater affinity for cGMP (Km = 3pmol/l) than for cAMP (Km = 30 umol/l),whereas the second (PDE I,) has an equalaffinity (Km = 1-2 umol/l) for the two cyclicnucleotides. PDE II (cGMP stimulated phos-phodiesterase) has a high Km for both cAMPand cGMP, and displays positive cooper-ativity with respect to substrate. The activityof this enzyme against cAMP is stimulatedmanyfold by "physiological" concentrations(0 1-1-0 pmol/l) of cGMP. The isozyme withthe lowest Km for cAMP and cGMP (about0-3 umol/l for both cyclic nucleotides) is PDEIII. In contrast to PDE II, the ability of PDEIII to hydrolyse cAMP is inhibited by lowconcentrations (0-1-10 umol/l) of cGMP.PDE IV is termed the cAMP specific phos-phodiesterase because its affinity for cAMP(Km = 2 jymol/l) is much greater than itsaffinity for cGMP (Km = 100 ,umol/l). Thefinal isozyme, PDE V, is called the cGMPspecific phosphodiesterase. This enzymehydrolyses cAMP with a Km of more than 100umol/l and cGMP with a Km of 1 iimol/l.An important point is that, whereas certain

isozymes possess a degree of selectivity forcAMP or cGMP, all phosphodiesterases arecapable of hydrolysing both cyclic nucleo-tides. Certain isozymes may also, for variousreasons, be inactive in intact tissues. Thus, inthe absence of additional information, it isimpossible to predict which isozyme orisozymes are responsible for cyclic nucleotidehydrolysis in a tissue containing multiplephosphodiesterases simply by determiningwhich isozymes are present in tissue homo-genates.From the standpoint of drug discovery the

synthesis of compounds that possess a sub-stantial degree of selectivity for one isozymeby comparison with others has been critical.2'These compounds are valuable pharmaco-logical probes with which to assess the impor-tance of various phosphodiesterase isozymesin regulating cyclic nucleotide content inintact tissues. More importantly, isozymeselective phosphodiesterase inhibitors mayhave therapeutic advantages over non-selec-tive compounds (see below) and several iso-zyme selective phosphodiesterase inhibitorsare, or were, being assessed in clinical studiesfor various disorders. Some examples areshown in table 2.

513



j.com/


ownloaded from


Torphy, Undem

Table 2 Examples of isozyme selective phosphodiesterase (PDE) inhibitors

IC50Class Drug (/smol/l)* Source Indication or action Development staget

1. Calmodulin stimulated(PDE I) Vinpocetine 15 Wyeth-Ayerst Nootropic+ Phase III

2. cGMP inhibited (PDE III) Anagrelide (BL 4162A) 0 05 Bristol-Myers Squibb Antithrombotic Phase IICilostamide (OPC 3689) 0 005 Otsuka Vasodilator DiscontinuedEnoximone (MDL 17043) 1 Marion Merrell Dow Inotropic Phase IIIImazodan (CI 914) 6 Warner-Lambert Inotropic DiscontinuedMilrinone (WIN 47203) 0-3 Sterling Winthrop Inotropic Discontinued§Siguazodan (SK&F 94836) 1 SmithKline Beecham Inotropic Discontinued

3. cAMP specific (PDE IV) Denbufylline (BRL 30892) 1 SmithKline Beecham Multi-infarct dementia Phase IIIRo 20-1724 5 Hoffmann-LaRoche Research DiscontinuedRolipram (ZK 62711) 1 Schering AG Antidepressant DiscontinuedTibenelast (LY 186655) 20 Lilly Antiasthma Phase IIITVX 2706 3 Troponwerke GmbH Anti-inflammatory Discontinued

4. Mixed PDE III/IV Benzafentrine (AH 21-132) 1-3 Sandoz Antiasthma DiscontinuedZardaverine (B 842-90) 1-3 Byk-Gulden Antiasthma Phase II

5. cGMP specific (PDE V) Dipyridamole 1 Commercial Vasodilator MarketedMY-5445 0-5 Mitsubishi Kasei Antithrombotic PreclinicalSK&F 96231 1 SmithKline Beecham Bronchodilator PreclinicalZaprinast (M&B 22,948) 1 Rhone-Poulenc Rorer Antiallergic Discontinued

*Value represents IC,O against corresponding PDE isozyme. Each compound possesses a selectivity for the isozyme noted, although the degree of selectivity of thedifferent compounds ranges from fivefold to several orders of magnitude.tReflects development stage in the United States only.Nootropics represent a class of drugs that improve cognitive functions, such as learning, memory, reasoning, and understanding.§Development of oral form discontinued; an intravenous formulation was approved in the United States but has not yet been marketed.IC,O-concentration causing a 50% inhibition ofPDE activity.

Therapeutic hypothesisTwo important hypothetical considerationsunderpin the proposed use of isozyme selectivephosphodiesterase inhibitors to treat asthma.The first is that the tissue distribution ofphosphodiesterase isozymes is heteroge-neous,1422-25 so that the isozymes responsiblefor the regulation of cyclic nucleotide contentdiffers from tissue to tissue. Secondly, it shouldbe possible to develop organ selective phos-phodiesterase inhibitors by targeting the com-pounds at the predominant isozyme or iso-zymes in the tissue of interest,2'l 26 27 and this hasbeen achieved with the PDE III selectiveinhibitors.2"30 These compounds are directedagainst the major cAMP hydrolysing phos-phodiesterase present in the myocardium andvasculature and thus are intended as inotropespossessing a modest degree of vasodilatoractivity. Although a positive effect of PDE IIIinhibitors on long term survival in patientswith congestive heart failure has yet to beshown, these compounds have a pronouncedbeneficial effect on haemodynamics.2930 In con-trast to previous agents that lacked isozymeselectivity (such as theophylline), the PDE IIIinhibitors produced few, ifany, side effects thatcould be ascribed to phosphodiesterase inhibi-tion in non-cardiovascular tissues.25 This sup-ports the proposal that the side effect profile ofphosphodiesterase inhibitors can be improvedby synthesising compounds that possess a highdegree of selectivity for the appropriateisozyme. An improved therapeutic index maypermit higher doses of isozyme selectiveinhibitors to be administered-that is, dosesthat produce a substantially greater inhibitionof the relevant isozyme than do non-selectiveinhibitors. The clinical efficacy of isozymeselective phosphodiesterase inhibitors maythen be greater than that of their non-selectivecounterparts.The second factor to consider with regard to

the efficacy of PDE inhibitors is that theirability to increase cyclic nucleotide content isrelated to the basal activity of adenylate orguanylate cyclase in target tissues, a con-sequence of the synergism between phos-phodiesterase inhibitors and activators ofadenylate cyclase (for example, beta adren-oceptor agonists) or guanylate cyclase (forexample, nitrovasodilators). The effect of theo-phylline in asthma has been attributed topotentiation of the effects of endogenousactivators of adenylate cyclase."l' The impor-tance of phosphodiesterase inhibition in thetherapeutic activity of theophylline has,however, been questioned,213" primarilybecause the plasma concentrations of unboundtheophylline that elicit bronchodilatation invivo produce only a modest (10-20%) inhibi-tion of phosphodiesterase activity in humanlung extracts.'3'4 This degree of inhibition inthe presence of endogenous adenylate orguanylate cyclase activators may, however, besufficient to produce a substantial increase incyclic nucleotide content in airway smoothmuscle and inflammatory cells.532 This proposalis supported by the fact that therapeutic con-centrations of theophylline act synergisticallywith exogenously administered activators ofadenylate cyclase to inhibit human polymor-phonuclear leucocyte activation ex vivo." 35

In contrast to the results with human poly-morphonuclear leucocytes, most studies inasthmatic subjects have detected only anadditive bronchodilator effect when a betaadrenoceptor agonist and theophylline aregiven together,3638 though a synergisticinteraction has been observed on occasion.'9One possible explanation for the failure todetect a synergistic interaction in vivo is thattheophylline produces bronchodilatation byseveral cellular mechanisms in addition tophosphodiesterase inhibition.

Several endogenously released agents are

514



j.com/


ownloaded from



capable of stimulating adenylate cyclaseactivity in inflammatory cells and airwaysmooth muscle, and could thus act to potentiatethe responses to phosphodiesterase inhibitors.These include (1) circulating catecholaminesf-rom the adrenal medulla4'; (2) prostaglandinE2 (PGE2) from monocytes, alveolar macro-phages,41 and airway epithelial cells42; (3) pros-tacyclin (PGI2) from vascular endothelialcells43; and (4) neurotransmitters from pulmon-ary adrenergic neurones and inhibitory non-adrenergic non-cholinergic (i-NANC)neurones (for example, vasoactive intestinalpolypeptide).4 45 The role of these endogenousadenylate cyclase activators in regulating theresponsiveness of target cells to phosphodies-terase inhibitors is speculative. Local concen-trations ofautacoids such as PGE2 and PGI2 areincreased in areas of inflammation and afterantigenic stimulation,46 so the functional res-ponse to phosphodiesterase inhibitors may begreater in the inflamed lungs of patients withasthma than in non-inflamed tissues. Cyclo-oxygenase inhibitors and beta adrenoceptorantagonists could reduce the therapeuticactivity of phosphodiesterase inhibitors,though to our knowledge neither of thesepossibilities has been tested.

Airway smooth muscleROLE OF CYCLIC NUCLEOTIDESThe mechanisms by which cyclic nucleotidesmediate smooth muscle relaxation has been thesubject of several recent reviews.' 10 4749 An indepth discussion of this topic is beyond thescope of this review, but a brief overview of theproposed roles ofcAMP and cGMP in regulat-ing airway smooth muscle tone is given; furtherdetails can be obtained elsewhere.9 10 47 48As depicted schematically in the figure,

intracellular cyclic nucleotide content isincreased by increasing the rate at which cyclicnucleotides are formed or by decreasing therate at which they are degraded. Activators ofadenylate cyclase (such as beta adrenoceptoragonists and prostaglandins E2 and 12) bind tocell surface receptors and, through a guaninenucleotide binding protein, increase the rate atwhich Mg2+-ATP is converted to cAMP. Incontrast to the regulation of adenylate cyclase,agents such as the nitrovasodilators (for exam-ple, sodium nitroprusside, nitroglycerin) andendothelium derived relaxant factor (probablynitric oxide) activate soluble guanylate cyclaseactivity directly without interacting with a cell

Table 3 Summary ofphosphodiesterase (PDE) isozymes in airway smooth muscle

Species Tissue Isozymes Reference (s)

Canine Trachealis I/, II, III, IV, V 51, 52Bovine Trachealis I, II, IV*t 53, 54Guinea pig Trachealis III, IV+ 51Human Bronchus I, III, IV, V*§ 55Human Trachealis II, 1,, II, III, IV, V** 56

*Subclassification of PDE I (that is, s or 11) has not yet been determined.tFraction designated PDE I may also contain PDE V.+At least two additional fractions of phosphodiesterase activity eluting from anion-exchangecolumns were not characterised.§PDE II may be present but not detected.**Results based on preliminary characterisation.

surface receptor. A distinct particulate guany-late cyclase is stimulated by the atriopeptinsthrough a receptor mediated mechanism that isnot dependent on a guanine nucleotide bindingprotein.The intracellular target enzymes for cAMP

and cGMP are cAMP dependent and cGMPdependent protein kinases. As cellular concen-trations ofcAMP orcGMP rise the appropriatecyclic nucleotide binds to and activates itscorresponding protein kinase. The activatedprotein kinases then mediate the physiologicalresponses to cyclic nucleotides by phos-phorylating and thus changing the activity ofkey substrates (such as enzymes and ion trans-port systems) concerned in the regulation ofsmooth muscle tone. Several relevant sub-strates are likely to be phosphorylated bycAMP dependent and cGMP dependentprotein kinases in airway smooth muscle.9 10 Inbrief, the increases in cAMP or cGMP contentcan produce airway smooth muscle relaxationby two general mechanisms. Firstly and per-haps most importantly, an increase in cyclicnucleotide concentrations leads to a decrease inthe cytosolic free Ca2' concentration by reduc-ing the mobilisation of Ca2+ from intracellularstores, inhibiting the influx of extracellularCa2', stimulating Ca2+ efflux or increasing Ca2+sequestration into intracellular storage sites.Secondly, activation of the cAMP or cGMPprotein phosphorylation cascades may inhibitthe activity of contractile proteins directly.Several of these biochemical pathways arelikely to be activated simultaneously and act ina coordinated fashion to reduce airway smoothmuscle tone.

Bronchodilators that act through secondmessenger pathways-for example, beta ad-renoceptor agonists and phosphodiesteraseinhibitors-should relax airway smooth muscleregardless of the mediator or mediators respon-sible for inducing bronchoconstriction. Thisshould be an advantage in a disease such asasthma in which several mediators may act inconcert to elicit bronchoconstriction.

PHOSPHODIESTERASE ISOZYMES IN AIRWAYSMOOTH MUSCLEOver the last 15 years, and especially the lastfive, several reports have described phos-phodiesterase isozyme profiles in airwaysmooth muscle from various species (table3).5056 A few caveats must be considered in anydiscussion of the results of these studies.Firstly, kinetic analyses ofenzyme activity havegenerally been carried out with crude or at bestpartially purified preparations. To our know-ledge, there have been no reports of studies inwhich every phosphodiesterase isozymepresent in airway smooth muscle has beenpurified to homogeneity. Secondly, airwaysmooth muscle preparations are not homogen-eous in terms of cellular content and at leastsome of the phosphodiesterase activity presentin homogenates of airway smooth muscle willcome from submucosal glands, capillary endo-thelial cells, fibroblasts, neurones, mast cells,etc. Thirdly, as discussed later, the presence ofa phosphodiesterase isozyme in a tissue homo-

515



j.com/


ownloaded from


Torphy, Undem

genate does not necessarily indicate a

physiological role for that isozyme. Fourthly,only recently have highly selective and potentinhibitors of phosphodiesterase isozymes beenavailable as pharmacological probes; many

published studies on the functional responses

of airway smooth muscle to phosphodiesteraseinhibitors have used compounds that are not

isozyme selective. Finally, virtually all studiesdescribing phosphodiesterase activity in airwaysmooth muscle have concentrated on trachealsmooth muscle. We do not yet know whetherisozyme profiles differ between central andperipheral airways.

Phosphodiesterase isozyme profiles havebeen studied more extensively in the caninetrachealis than in any other type of airwaysmooth muscle. Polson and colleagues usedanion exchange chromatography to separatephosphodiesterases from the soluble fraction ofcanine trachealis homogenates almost a decadeago.50 Five apparently distinct peaks of phos-phodiesterase activity eluted from the column,though only one of these was characterisedkinetically. The phosphodiesterase in this peakwas not stimulated by Ca +/calmodulin andhydrolysed cAMP with a Km of 0O6 ,mol/l andcGMP with a Km of 3 pmol/l. On the basis ofthe characteristics outlined in table 1, thesedata suggest that at least one of the isozymes incanine trachealis is PDE III. More recently, at

least two groups have used conventionalchromatographic techniques to separate caninetracheal phosphodiesterases and have thenidentified the isozymes present by means ofkinetic analyses, isozyme specific modulators ofenzyme activity (for example, cGMP, Ca" /calmodulin), and isozyme selective phospho-diesterase inhibitors.5' 52 These studies haveshown five distinct phosphodiesterase iso-zymes, including phosphodiesterases I:, II,III, IV, and V. There appears to be a differen-tial subcellular distribution of these isozymesin that PDE II is absent from high speedmembrane fractions, whereas PDE III is en-

riched in membrane fractions.52 Although fiveisozymes are present in canine trachealis, somemake a minimal contribution to total cyclicnucleotide phosphodiesterase activity in tissuehomogenates.9

Bovine and guinea pig tracheal phosphodies-terase isozymes have also been isolated andcharacterised. Cytosolic fractions of bovinetrachealis homogenates contain phosphodies-terases I, II, and IV.53 54 In view of the substan-tial inhibitory potency of zaprinast against theenzyme described as phosphodiesterase I"35perhaps, as in canine trachealis,5'2 the enzyme

preparation designated PDE I in these studiesalso contained PDE V. PDE III was conspic-uously absent from the soluble fraction ofbovine trachealis homogenates; whether it ispresent in the membrane fraction of this tissueis not yet known. Isolation and partial charac-terisation of phosphodiesterases from guineapig tracheal smooth muscle indicate thepresence of both PDE III and PDE IV.5" Atleast two other peaks of phosphodiesteraseactivity were eluted from anion exchangecolumns, but the enzyme activity in thesefractions were not characterised.5'

Finally, a limited amount of work has beendirected toward defining the phosphodiesteraseisozyme profile of human airway smooth mus-cle. In an early study by Bergstrand andLinquist,55 at least four distinct peaks of PDEactivity were isolated from human bronchialtissue by anion exchange chromatography. Thekinetic characteristics of the different activitiessuggest the presence ofPDEs I, III, IV, and V.The fact that this study was carried out withbronchial tissue, a preparation containing ahighly heterogenous cell population, raises thequestion of whether one or more of the PDEsidentified were present in contaminating cellsrather than in the smooth muscle. The resultsare supported, however, by a recent prelimin-ary report on human tracheal smooth muscle,56which appears to contain all the phosphodies-terases present in bronchial tissue as well asPDE I,56 an enzyme that would probably nothave been detected by the techniques used inthe earlier study.

ACTION OF ISOZYME SELECTIVEPHOSPHODIESTERASE INHIBITORS ON AIRWAY

SMOOTH MUSCLEIdentification of a PDE isozyme in a tissue doesnot provide information on the relative impor-tance of the isozymes in regulating cyclicnucleotide content in vivo. One approach todefining the physiological role of phosphodies-terase isozymes, and hence appropriatemolecular targets for new therapeutic agents, isto examine the biochemical and functionaleffects of isozyme selective inhibitors in intacttissues. Polson and colleagues determined thepotency of various xanthines as inhibitors oftwo distinct phosphodiesterase activitiesisolated from canine trachealis: a low Km cAMPphosphodiesterase, probably PDE III, and asecond phosphodiesterase activity, perhapsrepresenting a mixture of isozymes, whichpreferred cGMP as a substrate.50 57 Whenpotencies of the various xanthines as inhibitorsof the two phosphodiesterase activities werethen related to their potencies as relaxants ofmethacholine contracted trachealis strips aremarkably good correlation was observed,50 57

suggesting that the xanthines examined wererelaxing canine trachealis by inhibiting theactivity of a specific cAMP or cGMP phos-phodiesterase. The results of these studies, thefirst to describe a correlation between inhibitionof distinct phosphodiesterase activities andrelaxation of airway smooth muscle, cannot beinterpreted unequivocally. One problem wasthe use of alkylxanthines that inhibited cAMPand cGMP phosphodiesterase activities non-selectively. Consequently, the potency order ofthe compounds for inhibition of the cAMPphosphodiesterase activity50 was identical tothat for inhibition of the cGMP phosphodies-terase activity.57 This and the lack of informa-tion on the effect of these compounds on cyclicnucleotide content in intact trachealis stripsmakes it difficult to determine whether therelaxant response of the trachealis to the phos-phodiesterase inhibitors was due to inhibitionof either the cAMP or the cGMP phosphodies-terase activity, or both.

516



j.com/


ownloaded from



More recently, the physiological role ofvarious phosphodiesterase isozymes in caninetrachealis has been examined by correlating thefunctional response (relaxation) of the tissue toisozyme selective phosphodiesterase inhibitorswith the biochemical response (cyclic nucleo-tide content) to these compounds.58 59Siguazodan (SK&F 94836), a highly selectivePDE III inhibitor,5260 produced a concentra-tion dependent relaxation of methacholine or

histamine contracted canine trachealis stripsthat was associated with an increase in cAMPcontent and an activation of cAMP dependentprotein kinase.58 On the other hand, siguazodanhad no effect on the functional or biochemical(that is, cGMP accumulation) responses tosodium nitroprusside,58 a guanylate cyclaseactivator that is thought to relax airway smoothmuscle via a cGMP mediated mechanism.'0These results provide strong evidence to sup-

port a major role for PDE III in regulatingcAMP content but not cGMP content in intactcanine trachealis.A similar approach was taken in another

study aimed at defining the roles of PDE III,IV, and V in canine trachealis.59 Theseexperiments assessed the ability of variousisozyme selective PDE inhibitors to potentiatethe functional and biochemical responses toisoprenaline or sodium nitroprusside. Thephosphodiesterase inhibitors used were SK&F94120, Ro 20-1724, and zaprinast, all of whichhad been shown to selectively inhibit caninetracheal PDE III, PDE IV, and PDE V respec-

tively.52 Isoprenaline induced relaxation andcAMP accumulation were potentiated by bothSK&F 94120 and Ro 20-1724, but these agentshad no effect on the response to sodiumnitroprusside. In contrast, zaprinast poten-tiated sodium nitroprusside induced relaxationand cGMP accumulation, but did not poten-tiate the responses to isoprenaline. Theseresults suggest that PDE III and PDE IVhydrolyse cAMP in intact canine trachealis,whereas PDE V hydrolyses cGMP.59

Several studies have shown the potent relax-ant activity ofPDE III inhibitors on the guineapig isolated trachea, providing further supportfor a role for this isozyme in regulating airwaysmooth muscle cAMP content.6'"" Harris andcolleagues found an excellent correlation be-tween the ability of a series of compounds torelax guinea pig trachea and their ability toinhibit PDE III isolated from guinea pigtrachealis.6' An analogous series ofexperimentsconducted with a different set of compoundsshowed no relation between relaxation andPDE IV inhibition but, inexplicably, relaxa-tion did correlate with the ability of thesecompounds to compete with a high affinityrolipram binding site.6' The nature and func-tion of this high affinity rolipram binding sitehave yet to be defined,'M but it may represent a

discrete subtype of PDE IV. There was pro-nounced synergism between the relaxantactivity of PDE III inhibitors and compoundsthat bind to the high affinity rolipram bindingsite. For example, the EC50 for rolipram was

reduced nearly 25 fold in the presence of a fixedconcentration of CI-930, a PDE III inhibitor.6"

Finally, a role for PDE V in regulating cGMPcontent in guinea pig airway smooth muscle issuggested by the observation that SK&F96231, a selective PDE V inhibitor, relaxesguinea pig isolated trachea in a concentrationdependent manner.65The biochemical response of guinea pig

trachealis to isozyme selective phosphodies-terase inhibitors has not been studied in greatdetail. The PDE III inhibitor SK&F 94120and the mixed PDE III/IV inhibitor benzafen-trine (AH 21-132) increase cAMP content inmuscle rich guinea pig tracheal preparations,though detectable increases in cAMP contentare observed only with concentrations of thesecompounds that produce near maximal relaxa-tion.6' 6 Much lower concentrations of ben-zafentrine substantially potentiated the abilityof a threshold concentration of forskolin toelevate cAMP content in this preparation.Benzafentrine,' but not SK&F 94120,61 alsoincreases cGMP content in guinea pig tra-chealis. SK&F 96231, a PDE V inhibitor,increased cGMP content in guinea pig lungparenchymal strips.65 Although the strips con-tain many cell types in addition to airwaysmooth muscle, this observation implies thatinhibition ofPDE V in intact tissues can lead toa rise in cGMP content.The bronchorelaxant activity of isozyme

selective phosphodiesterase inhibitors in theguinea pig is not limited to in vitro prepara-tions. Various PDE III, PDE IV, mixed PDEIII/IV, and PDE V inhibitors have substantialbronchodilator activity in anesthetised guineapigs63 65 67-71 and, as in experiments in vitro, thebronchodilatation appears to occur regardlessof the spasmogen used to induce tone.The effects ofisozyme selective phosphodies-

terase inhibitors in bovine airway smoothmuscle have not been examined as extensivelyas in guinea pig or canine airway smoothmuscle and there are no data correlating relax-ant responses with changes in cyclic nucleotidecontent. Preliminary data suggesting thatrolipram is a potent relaxant of bovine tra-chealis imply an important role for PDE IV inregulating cAMP content in this tissue.5472 Onthe other hand, the role of PDE III in bovinetrachealis is somewhat ambiguous. Althoughthe PDE III inhibitors SK&F 94120 andmilrinone relax this tissue, the potency andefficacy of both compounds are distinctlylow.54 72 The poor relaxant activity of PDE IIIinhibitors, coupled with the failure to detectPDE III in the soluble fraction of bovinetrachealis homogenates,53 54 raises doubt aboutthe functional role of PDE III in this tissue.The functional role of PDE V in the bovinetrachea appears to be minor as zaprinast haslittle or no relaxant activity in this tissue.54 1 73

Functional effects of isozyme selective phos-phodiesterase inhibitors on human isolatedairway smooth muscle have not yet been repor-ted, though several groups are actively engagedin this area of research and results of thesestudies should be available soon.One final caveat concerns the interpretation

of the data discussed in this section: no phos-phodiesterase inhibitors possess absolute

517



j.com/


ownloaded from


Torphy, Undem

isozyme specificity. When present in high con-centrations all isozyme selective inhibitors willinhibit phosphodiesterases other than the onesagainst which they are targeted. Certainisozyme selective phosphodiesterase inhibitorsmay also regulate physiological processes bymechanisms that are unrelated to phosphodies-terase inhibition. For example, the ability ofrolipram to inhibit agonist induced generationof inositol phosphates in bovine trachealis maybe unrelated to phosphodiesterase inhibitionand cAMP accumulation.73 Thus conclusionson the role of individual phosphodiesteraseisozymes in intact tissues drawn from experi-ments with isozyme selective phosphodies-terase inhibitors are somewhat tenuous, parti-cularly in the absence of corroborative bio-chemical data.

Inflammatory cellsROLE OF CYCLIC NUCLEOTIDESThe vast number of published papers leaveslittle doubt that cAMP acts as an inhibitorymessenger in inflammatory cells.5" 12 In addi-tion to inhibiting inflammatory mediatorproduction and release, it is evident that anincrease in cellular cAMP inhibits other func-tions of inflammatory cells, including chemo-taxis, cytotoxicity and cell aggregation. Thusincreases in cAMP content have a generalisedsuppressive influence on human inflammatorycells (figure).More than 20 years ago Lichtenstein and

coworkers suggested that an increase in cellularcAMP led to an inhibition of antigen inducedhistamine release from human basophils.74This speculation was based on the observationthat beta adrenoceptor agonists and theo-phylline were effective inhibitors of antigeninduced histamine release from human mixedleucocyte preparations, and was consistentwith the observations made by Schild some 30years earlier that epinephrine is a potentinhibitor of anaphylactic histamine releasefrom lung tissue.75 Various compounds thateither stimulate adenylate cyclase or inhibitcyclic nucleotide phosphodiesterase (see

Table 4 Summary ofphosphodiesterase (PDE) isozymes in human inflammatory cells

PredominantPDE Effect ofPDE

Cell type isozyme(s) t inhibitors+ Reference(s)

Basophil IV lMediator release 88Mast cell IV iMediator release 32Neutrophil IV lMediator release 90, 99

iRespiratory burstEosinophil* IV ISuperoxide formation 76Monocyte IV iMediator release 77, 91, 93, 103, 104

ICytokine formationLymphocyte III, IV iCytotoxicity 91, 92, 98, 100, 101

lIgE formationICytokine formationiBlastogenesis

Platelet IIITAggregation 58, 89, 94, 105, 106iMediator release

*So far only studies with guinea pig eosinophils have been conducted.tOnly the major cAMP metabolising enzyme or enzymes are listed.$Represents the effect of either isozyme selective or non-selective phosphodiesterase inhibitors;although a few of the cited studies were conducted with non-selective compounds, the overallpharmacological effect observed is likely to have been due to inhibition of the majorphosphodiesterase isozyme present in the cell.

below) are now known to have inhibitory effectson many inflammatory cell types, includingeosinophils, neutrophils, monocytes, platelets,basophils, and mast cells.5 112 76-79Although many pharmacological studies

have shown that increasing cAMP contentinhibits inflammatory cell activity, there is littleinformation on the specific mechanisms mediat-ing this inhibitory effect, though it appears to bedue to multiple mechanisms. As with thestudies on smooth muscle discussed above,evidence indicates that at least one of themechanisms by which cAMP suppresses in-flammatory cell activity includes an inhibitoryeffect on stimulus induced increase in cytosolicCa2+ .78 80-85 Because an increase in cytosolicCa2+ is a trigger for several cellular activities,including mediator production and release,inhibiting this event would be expected tocause a general inhibition of inflammatory cellfunction. There is also evidence, however, thatcAMP can inhibit mediator release in theabsence of gross changes in cytosolic Ca2+concentrations.84 In mast cells a rise in cAMP ismuch more effective in inhibiting antigeninduced eicosanoid biosynthesis than in inhi-biting the rise in cytosolic Ca2+ or histaminerelease.8486 In the case of eicosanoid biosyn-thesis, it has been suggested that activation ofthe cAMP cascade may cause direct inhibitionof an enzyme or enzymes in the biosyntheticpathway (for example, phospholipase A2).84Much less is known about the role ofcGMP

in the regulation of inflammatory cell function.An increase in cGMP may enhance antigeninduced histamine release from human lungmast cells.87 These studies, however, werecarried out on lung parenchymal fragments, sothe biochemical data are difficult to interpret.Neither sodium nitroprusside nor dibutyrylcGMP inhibited immunological mediatorrelease from human basophils88 or isolatedhuman lung mast cells in our studies (unpub-lished observations) Although the precise roleofcGMP in regulating inflammatory cell func-tion is uncertain, its effect appears to be modestby comparison with the profound inhibitoryeffect of cAMP.

PHOSPHODIESTERASE ISOZYMES ININFLAMMATORY CELLSThe profile of phosphodiesterase isozymes ininflammatory cells has been obtained onlyrecently (table 4). Chromatographic analysescombined with kinetic and pharmacologicalcharacterisations show that PDE IV is themajor phosphodiesterase isozyme responsiblefor catabolisingcAMP in several human inflam-matory cell types, including mast cells, baso-phils, monocytes, and neutrophils. In contrast,PDE III is the predominant isozyme in plate-lets and a combination ofPDE III and PDE IVis present in mixed lymphocyte preparations.The next section looks at the identity of

the predominant cAMP phosphodiesterase ac-tivity in various inflammatory cells. In virtuallyall the cells studied so far, however, minoractivity from one or more phosphodiesteraseshas been found in addition to the majoractivity. In the basophil and platelet, for exam-

518



j.com/


ownloaded from


Phosphodiesterase inhibitors: new opportunitiesfor the treatment of asthma

ple, the additional phosphodiesterase is selec-tive for cGMP and is inhibited by zaprinast,88"8suggesting that it is PDE V. On the basis of thesubstrate specificity this isozyme seems unlikelyto have a physiological role in the catabolism ofcAMP. Thus selectively inhibiting this iso-zyme is unlikely to lead to a rise in cellularcAMP.

Basophil and mast cellIn the human basophil most of the cAMPhydrolysis activity is accounted for by PDE IV(Km = 1 6 jmol/1).88 A small portion (under20%) of the activity is attributed to PDE III(Km = 0 3 1M), on the basis of the observationsthat the low Km isozyme was inhibited bycGMP and the selective PDE III inhibitorSK&F 95654. The phosphodiesterase isozymeprofile in human mast cell populations is notknown. Mouse mast cells derived from mousebone marrow contain a single cAMP hydrolys-ing isozyme with kinetic and pharmacologicalcharacteristics of PDE IV.32

NeutrophilThe cAMP PDE activity in the human neutro-phils is eluted from DEAE anion exchangecolumns in a single peak, has a Km of0 7 pmol/l,and is sensitive to inhibition by the selectivePDE IV inhibitors rolipram and Ro 20-1724.9oThese findings suggest that the phosphodies-terase isozyme responsible for hydrolysingcAMP in the human neutrophil is predomi-nantly if not exclusively PDE IV.

LymphocyteSystematic evaluation of the PDE isozymeprofile in the various subtypes of lymphocyteshas not been performed to our knowledge.Thompson and coworkers reported that amixed lymphocyte preparation (99% lym-phocytes) contains a single phosphodiesteraseisozyme with a low Km (0 40 ymol/l) for cAMPhydrolysis and a much higher Km for cGMPhydrolysis.9' The elution profile and kineticcharacteristics suggest that the isozyme is PDEIII. Both PDE III and PDE IV may be presentin T lymphocytes because in preparationsenriched in T lymphocytes two high affinitycAMP selective phosphodiesterases exist, oneof which is inhibited by Ro 20-1724 andanother which is not.92 The different isozymeprofile observed in these cells versus mixedlymphocytes raises the possibility that differentsubsets of lymphocyte may contain differentphosphodiesterases .92

MonocytePurified monocyte preparations isolated fromperipheral blood contain a single cAMP selec-tive phosphodiesterase with a Km of 1-3,mol/l19; this isozyme appears to be PDE IV.This is supported by a preliminary study inwhich the major cAMP phosphodiesterase inmonocytes was partially purified and character-ised93 and shown to have the kinetic character-istics (cAMP Km = 1-8 ymol/l) and pharmaco-logical properties (inhibited by rolipram andRo 20-1724) consistent with its being PDE IV.

PlateletThe predominant phosphodiesterase isozymein the human platelet is PDE III... .4. Thisisozyme has a Km for both cAMP and cGMP ofabout 0-4 imol/l, is selectively inhibited byseveral PDE III inhibitors, including cilo-stamide, milrinone, and siguazodan, and isinhibited by cGMP (IC,, - 0-1 imol/l), adistinct characteristic ofPDE III.

ACTION OF ISOZYME SELECTIVE INHIBITORS ONINFLAMMATORY CELL FUNCTIONIn view of the individual profile of phos-phodiesterase isozymes in inflammatory cells,and knowledge that a rise in cAMP has potentanti-inflammatory activity, it is reasonable topropose that isozyme selective inhibitors maybe useful as anti-inflammatory agents in thetreatment of asthma. Information on the func-tional consequence of selective inhibition ofphosphodiesterase isozymes in human inflam-matory cells is summarised below. Becausemuch of the work is recent some conclusionsare drawn from data that are unpublished orpublished only as abstracts. Appropriatecaution in these areas is warranted. In generalthe PDE isozyme profile in the respective celltype appears to be a good predictor of thepharmacology of isozyme-selective inhibitors(table 4).

Mast cellSK&F 95654 and zaprinast, inhibitors of PDEIII and PDE V respectively, have little or noeffect on immunologically stimulated mediatorrelease from human isolated lung mast cells.32In contrast, rolipram and Ro 20-1724 inhibitantigen induced mediator release from thesecells, and potentiate the inhibitory effect ofadenylate cyclase activators.32 The effect ofisozyme selective inhibitors on human lungmast cells is mimicked in the mouse bonemarrow derived mast cell. In these cells PDEIV inhibition reduces antigen induced his-tamine release and leukotriene C4 production,whereas zaprinast and siguazodan are withouteffect.32 Interestingly, the rat peritoneal mastcell appears to respond differently to isozymeselective inhibitors from both the human lungmast cell and the mouse bone marrow derivedmast cell. In this cell type submicromolarconcentrations of zaprinast effectively inhibitimmunologically induced histamine release,whereas rolipram has little effect.95 Whether thisreflects a species difference or is a result ofmicroenvironment dependent mast cellheterogeneity is unknown.

BasophilInhibiting PDE IV with rolipram or Ro 20-1724 effectively inhibits antigen-induced his-tamine and leukotriene C4 release from humanbasophils.889596 Rolipram also potentiates theinhibitory effect offorskolin on mediator releasefrom these cells.95 Neither zaprinast nor SK&F95654 influences antigen induced mediatorrelease from the human basophil, but SK&F95654 potentiates the inhibitory effect ofrolipram. This potentiation is observed bothfor inhibition of mediator release and increase

519



j.com/


ownloaded from


Torphy, Undem

in cellular cAMP.95 Thus there appears to be asynergistic interaction between the inhibitorsof PDE III and PDE IV in the basophil.Combinations ofPDE III and PDE IV inhibi-tors have also been reported to act synergis-tically in guinea pig trachealis, rat ventricle,and human lymphocyte.639798 Examination ofthe effects of various combinations of isozymeselective phosphodiesterase inhibitors on in-flammatory cell function deserves furtherattention.

NeutrophilThe neutrophil respiratory burst is inhibitedby the PDE IV inhibitors, rolipram, and Ro 20-1724.9 The effectiveness of these inhibitors isequal to that of the relatively non-selectivephosphodiesterase inhibitors theophylline andenprofylline. The phosphodiesterase profile ofthese cells suggests that PDE IV is the onlyphosphodiesterase isozyme regulating cAMPcontent in neutrophils. Thus selective concen-trations of zaprinast and various PDE IIIinhibitors are without effect on the respiratoryburst in this cell type. Inhibition of phos-phodiesterase and increase in cAMP have alsobeen found to inhibit lysozyme release fromneutrophils and the synthesis of plateletactivating factor and LTB,.7999

EosinophilNo information is available on the effect ofisozyme selective phosphodiesterase inhibitorson human eosinophil function. In a prelimin-ary report Dent and coworkers reported thatPDE IV selective inhibitors reduce superoxideformation in guinea pig eosinophils.76

LymphocyteThe PDE IV inhibitor Ro 20-1724 inhibits theactivity of human cytotoxic T lymphocytes'°°and interleukin-2 production by human T-lymphocytes."'0 In addition, both Ro 20-1724and CI-930, a PDE III inhibitor, cause amodest inhibition of blastogenesis in T lym-phocytes.98 These agents have a much morepronounced inhibitory effect on blastogenesiswhen used in combination, however, suggest-ing that both PDE III and PDE IV activitiesare important in regulating cAMP content inthese cells.9 98 Selective inhibitors of PDE IVmay also inhibit antibody production by Blymphocytes. Ro 20-1724 inhibits IgE produc-tion by peripheral blood mononuclear leuco-cytes and most, but not all, of this activity canbe attributed to a direct inhibitory effect on theB cells.'02

MonocyteInhibition of monocyte phosphodiesterase andthe resulant rise in cellular cAMP is associatedwith inhibition of arachidonic acid metabolismin zymosan treated cells." Theophylline andother agents that increase cAMP content alsoinhibit the production of tumour necrosisfactor in human monocytes and rat macro-phages,'03104 probably by a cytokine specificmechanism as phosphodiesterase inhibitorshave no effect on the production ofinter-leukin-1.103 Effects of isozyme-selective phosphodies-

terase inhibitors on monocyte function havenot been reported.

PlateletConsistent with its phosphodiesterase isozymeprofile, platelet function is profoundly inhibitedby selective PDE III inhibitors. Sub-micromolar concentrations of several selectivePDE III inhibitors, including anegralide, SKF94120, Ro 15-2041, and siguazodan, are effec-tive inhibitors of agonist induced calciummobilisation, aggregation, and secre-tion.60 89 1051' Inhibition of PDE IV or V withrolipram and zaprinast respectively has little orno effect on platelet responses.89

Clinical experienceTHERAPEUTIC EFFECTSClinical experience with the use of isozymeselective phosphodiesterase inhibitors for thetreatment of asthma is limited. To our know-ledge, the first such compound to be examinedwas zaprinast, an inhibitor ofthecGMP specificphosphodiesterase (PDE V). In an initialplacebo controlled, double blind crossover trialoral zaprinast (10 mg) reduced exercise in-duced bronchoconstriction in adult asthmaticpatients significantly,'07 whereas it failed toinhibit bronchoconstriction induced by his-tamine. The reason that zaprinast was activeagainst exercise induced bronchoconstrictionbut not histamine induced bronchoconstrictionis not obvious. On the basis ofthe proposed roleof cGMP in airway smooth muscle, we mightspeculate that the efficacy ofzaprinast stemmedfrom bronchodilator activity. This explanationdoes not appear to be viable, however, in viewof the failure of zaprinast to abrogate histamineinduced bronchoconstriction. On the otherhand, zaprinast might have altered neuralreflexes or mediator release, perhaps through amechanism unrelated to PDE inhibition. But ithad no effect on exercise induced bronchocon-striction in a subsequent study with asthmaticchildren.'08 Thus the positive results obtainedin the earlier study were not repeated in adifferent group of asthmatic patients.At least a dozen PDE III inhibitors have

been evaluated for inotropic or antithromboticactivity in man but bronchodilator activity hasbeen assessed only for one, enoximone.'09 Pul-monary mechanics were studied before andafter an intravenous infusion of enoximone (3mg/kg) in 19 patients with decompensatedchronic obstructive pulmonary disease. Enox-imone reduced pulmonary resistance andincreased dynamic lung compliance in bothspontaneously breathing and artificially ven-tilated patients. Although long term studieshave not been conducted in patients withasthma, these preliminary results suggest thatenoximone can reverse bronchoconstriction inman.

Preliminary results on the activity of tiben-alast,"10 a PDE IV inhibitor, and benzafen-trine,"1' a mixed PDE III/IV inhibitor, alsohave been reported. The results with tibenalastwere equivocal. A single oral 150 mg dose oftibenalast caused a slight but non-significant

520



j.com/


ownloaded from



increase in FEV, in 40 asthmatic subjects.Because of the putative anti-inflammatoryactivity of PDE IV inhibitors, it would be ofinterest to assess the effect of tibenalast over alonger time. Inhalation of benzafentrine (2-32mg) by non-asthmatic subjects reversed metha-choline induced bronchoconstriction, asmonitored by changes in specific airways con-ductance, in a dose dependent manner."' Apreliminary report indicates that intravenouslyadministered benzafentrine (40 and 80 mg)produced a modest and transient broncho-dilatation, whereas the compound was in-effective when administered by the oral route(9_90 mg)."2

SIDE EFFECTSWith the possible exception of the PDE IIIinhibitors, clinical experience with isozyme-selective phosphodiesterase inhibitors is notyet extensive enough to draw conclusions aboutthe side effect profile of the different classes ofcompounds. The potential for certain sideeffects can be predicted based on the anticipatedactivity of different isozyme selective inhibitorson various organs.97 For example, by virtue oftheir cardiotonic activity, PDE III inhibitorswould be expected to produce an increase inmyocardial contractility and modest vasodila-tion.'>30 Whether such effects would bedetrimental in subjects with asthma who wereotherwise healthy is unknown. Perhaps ofmoreconcern is the arrhythmogenic potential ofthese compounds."3 If such activity is theresult of PDE III inhibition per se, it couldrepresent a major limitation to the use of theseagents to treat asthma.The most obvious concern regarding the

PDE IV inhibitors stems from their apparentantidepressant activity."s"6 The anti-depres-sant activity of rolipram and, presumably,other compounds of its class appears to belinked to inhibition of PDE IV in the brain."7Whether PDE IV inhibitors will produce un-acceptable central nervous system effects inindividuals who do not suffer from affectivedisorders remains to be answered. Gastrointes-tinal disturbances, such as pyrosis, nausea, andvomiting, are frequently reported side effectsassociated with the use of rolipram.11"l6Rolipram has also been reported to produce arapid and pronounced fall in plasma osmo-lality, probably as a result of an antidiureticeffect; this resolves after seven days' treat-ment.'8 It is not clear whether these effects aredue to PDE IV inhibition in general or whetherthey are unique to rolipram.

Cyclic GMP is purported to have a criticalphysiological role in mediating vasodilationand PDE V appears to be important in regulat-ing cGMP content in vascular smooth mus-cle.471'9 120 Thus PDE V inhibitors may havecardiovascular activity.

Because of the dearth of information con-cerning the clinical activity of isozyme-selec-tive phosphodiesterase inhibitors, most of theside effects discussed above are speculative.Individual agents are also likely to produce sideeffects that are not related to phosphodiesteraseinhibition and cannot therefore be predicted.

ConclusionsUnderstanding of the potential role of isozymeselective phosphodiesterase inhibitors in thetreatment of asthma is far from complete.Preliminary data are beginning to provide anoverall sense of the potential value of thesecompounds. Most evidence so far suggests thatPDE III inhibitors, and perhaps PDE IV andV inhibitors, possess bronchodilator activity.In addition, inhibitors of PDE IV may haveactivity against the inflammatory aspects ofasthma. Little is known about the potentialutility of inhibitors ofPDE I and PDE II.Attempts to broaden the therapeutic activity

of isozyme selective phosphodiesteraseinhibitors have been made by synthesisinghybrid compounds and at least two mixed PDEIII/IV inhibitors, zardaverine and benzafen-trine, have undergone preliminary clinicalstudies. These compounds may provide anoptimal combination of bronchodilator andanti-inflammatory activities. Hypothetically, acombination of PDE IV and PDE V inhibitorswould have a similar therapeutic profile.Although the concept of hybrid phosphodies-terase inhibitors is superficially attractive, anyloss in isozyme selectivity is likely to lead tomore side effects, thus subverting one of themajor proposed advantages of isozyme selec-tive phosphodiesterase inhibitors over non-selective compounds.

After having been moribund for many years,interest in phosphodiesterase inhibitors asagents for asthma has undergone a strikingresurgence, particularly with respect to thepotential therapeutic advantages of isozymeselective phosphodiesterase inhibitors. Anunderstanding of the range of activity of thesecompounds should emerge over the nextdecade as results from clinical trials becomeavailable.

We gratefully acknowledge the help of Dr Carol Kulp forconducting literature searches and Ms Dotti Lavan for hercareful preparation of this manuscript.

1 Jenne JW. Physiology and pharmacodvnamics of the xan-thines. In: Jenne JW, Murphy S, eds. Drug therapy forasthma: research and clinical practice. (Lung biology inhealth and disease series.) New York: Dekker, 1987:297-334.

2 PerssonCGA.Overviewofeffectsoftheophylline. JAllergyClin Immunol 1986;78:780-7.

3 Persson CGA, Erjefilt I, Gustafason B. Xanthines-symptomatic or prophylactic in asthma? Agents Actions1988;23(Suppl):S 1 37-55.

4 Weinberger M. Treatment of chronic asthma with theo-phylline. ISI Atlas Sci Pharm 1988;2:53-61.

5 Kuehl FA, Zanetti ME, Soderman DD, Miller DK, HamEA. Cyclic AMP-dependent regulation of lipid mediatorsin white cells. A unifying concept for explaining theefficacy of theophylline in asthma. Am Rev Respir Dis1987;136:210-3.

6 Fredholm BB. Are methylxanthine effects due to antag-onism of endogenous adenosine? Trends Pharmacol Sci1980;1: 129-33.

7 Poisner AM. Direct stimulant effect of aminophylline oncatecholamine release from the adrenal medulla. BiochemPharmacol 1973;22:469-76.

8 Fink G, Mittelman M, Shokat B, Spitzer SA. Theo-phylline-induced alterations in cellular immunity in asth-matic patients. Clin Allergy 1987;17:313-6.

9 Torphy TJ. Action of mediators on airway smooth muscle:functional antagonism as a mechanism for bronchodilatordrugs. Agents Actions 1 988;23(suppl):S37-53.

10 Torphy TJ, Hay DWP. Biochemical regulation of airwaysmooth muscle tone: an overview. In: Townley RG,Agrawal DK, eds. Airway smooth muscle: modulation ofreceptors and response. Boca Raton: CRC Press, 1990:39-68.

521



j.com/


ownloaded from


Torphy, Undem

11 Bourne HR, Lichtenstein LM, Melmon KL, Henney CS,Weinstein Y, Shearer GM. Modulation of inflammationand immunity by cyclic AMP. Science 1974;184:19-28.

12 Plaut M, Marone G, Thomas LL, Lichtenstein LM. Cyclicnucleotides in immune responses and allergy. AdvCyclicNucleotide Res 1980;12:161-72.

13 Church MK, Featherstone RL, Cushley MJ, Mann JS,Holgate ST. Relationships between adenosine, cyclicnucleotides, and xanthines in asthma. J Allergy ClinImmunol 1986;78:670-5.

14 Beavo JA. Multiple isozymes of cyclic nucleotidephosphodiesterase. Adv Second Messenger PhosphoproteinRes 1988;22:1-38.

15 Beavo JA, Reifsnyder DH. Primary sequence of cyclicnucleotide phosphodiesterase isozymes and the design ofselective inhibitors. Trends Pharmac Sci 1990;1 1:150-5.

16 Colicelli J, Birchmeier C, Michaeli T, O'Neill K, Riggs M,Wigler M. Isolation and characterization of a mammaliangene encoding a high-affinity cAMP phosphodiesterase.Proc Natl Acad Sci USA 1989;86:3599-603.

17 Henkel-Tigges J, Davis RL. Rat homologs of the Droso-phila dunce gene code for cyclic AMP phosphodiesterasesensitive to rolipram and RO 20-1724. Mol Pharmacol1990;37:7-10.

18 Pittler SF, Baehr W, Wasmuth JJ, McConnell DF,Champagne MS, van Tuinen P, Ledbetter D, Davis RL.Molecular characterization of human and bovine rodphotoreceptor cGMP phosphodiesterasex-subunit andchromosomal localization of the human gene. Genomics1990;6:272-83.

19 Swinnen JV, Joseph DR, Conti M. Molecular cloning of rat

homologues of the Drosophila melanogaster dunce cAMPphosphodiesterase: evidence for a family of genes. ProcNatI Acad Sci USA 1989;86:5325-9.

20 Livi GP, Kmetz P, McHale MM, et al. Cloning andexpression of cDNA for a human low-K,,, rolipram-sensitive cyclic AMP phosphodiesterase. Molec CellBiol1990;10:2678-86.

21 Weishaar RE, Cain MH, Bristol JA. A new generation ofphosphodiesterase inhibitors: multiple molecular forms ofphosphodiesterase and the potential for drug selectivity.JMedChem 1985;28:537-45.

22 Beavo JA, Hardman JG, Sutherland EW. Hydrolysis ofcyclic guanosine and adenosine 3',5'-monophosphates byrat and bovine tissues.J Biol Chem 1970;245:5649-55.

2'3 Appleman MM, Avaro MA, Takemato DJ, Whitson RH.Cyclic nucleotide phosphodiesterases. In: Nathanson JA,Kebalian JW, eds. Handbook of experimental pharma-cology. Vol 5812. Berlin: Springer, 1982:262-300.

24 Thompson WJ, Appleman MM. Characterization of cyclicnucleotide phosphodiesterases of rat tissues. J Biol Chem1971;246:3145-50.

25 Wells JN, Hardman JG. Cyclic nucleotide phosphodies-terases. Adv Cyclic Nucleotide Res 1977;8:119-43.

26 HidakaH, Tanaka T, Itoh H. Selective inhibitors of cyclicnucleotide phosphodiesterases. Trends Pharmacol Sci1984;5:237-9.

27 Weiss B. Differential activation and inhibition of multipleforms of cyclic nucleotide phosphodiesterase. Adv CyclicNucleotide Res 1975;-5:195-21 1.

28 Colucci WS, Wright RF, Braunwald E. New positiveinotropic agents in the treatment of congestive heartfailure: mechanisms of action and recent clinical develop-ments. N EnglJ Med 1986;314:349-58.

29 Hood WB. Controlled and uncontrolled studies of phos-phodiesterase III inhibitors in contemporary cardio-vascular medicine. Am J Cardiol 1989;63:46-53A.

30 Wood MA, Hess ML. Review: long-term oral therapy ofcongestive heart failure with phosphodiesteraseinhibitors. AmJMed Sci 1989;297:105-13.

31 Nielson CP, Crowley JJ, Cusack BJ, Vestal RE.Therapeutic concentrations of theophylline andenprofylline potentiate catecholamine effects and inhibitleukocyte activation. J Allergy Clin Immunol 1986;78:660-7.

32 Torphy TJ, Livi GP, Bacarek JM, White JR. Chilton FH,Undem BJ. Therapeutic potential of isozyme-selectivephosphodiesterase inhibitors in the treatment of asthma.Second Messenger Phosphoprotein Research (in press).

33 Bergstrand H. Phosphodiesterase inhibition and theo-phylline. Eur J Respir Dis 1980;10(suppl):37-44.

34 Polson JB, Krzanowski JJ, Goldman AL, Szentivanyi A.Inhibition of human pulmonary phosphodiesteraseactivity by therapeutic levels of theophylline. Clin ExpPharmacol Physiol 1978;8:535-9.

35 Nielson CP, Crowley JJ, Morgan ME, Vestal RE.Polymorphonuclear leukocyte inhibition by therapeuticconcentrations of theophylline is mediated by cyclic-3',5'-adenosine monophosphate. Am Rev Respir Dis 1988;137:25-30.

36 Bitling B, Dahlqvist R, Garli M, Hornbland Y, Ripe E.Separate and combined use of terbutaline and theo-phylline in asthmatics. Effects related to plasma levels. EurJ Respir Dis 1982;63:399-409.

37 Handslip PDJ, Dart AM, Davies BH. Intravenous salbu-tamol and aminophylline in asthma: a search for synergy.

Thorax 1981;36:741-4.38 Svedmyr K, Svedmyr N. Does theophylline potentiate

inhaled /l, agonists? Allergy 1982;37:101-10.39 Svedmyr N, Svedmyr K. In vitro and in vivo effects of

theophylline and (J-adrenostimulants in combination. EurJ Respir Dis 1980;109(suppl):83-91.

40 Barnes PJ. Endogenous catecholamines and asthma.Allergy Clin Immunol 1986;77:791-5.

41 Godard P, Chaintreuil J, Damon M, et al. Functionalassessment of alveolar macrophages: comparison of cellsfrom asthmatics and normal subjects. J Allergy ClinImmunol 1982;70:88-93.

42 Keikauf GD, Ueki IF, Nadel JA, Widdicombe JH. Brady-kinin stimulates chloride secretion and prostaglandin E2release by canine tracheal epithelium. AmJ Physiol 1985;248:F48-55.

43 Johnson AP. Human pulmonary endothelial cells in cul-ture: activities of cells from arteries and cells from veins.Clin Invest 1980;65:841-50.

44 Laitinen LA, Laitinen A. Innervations of airway smoothmuscle. Am Rev Respir Dis 1987;136:S38-42.

45 Palmer JBA, Barnes PJ. Neuropeptides and airway smoothmuscle function. Am Rev Respir Dis 1987;136:S50-4.

46 Schulman ES, Adkinson NF, Jr, Newball HH. Cyclooxy-genase metabolites in human lung anaphylaxis: airway vs.parenchyma.J Appl Physiol 1982;53:589-95.

47 Lincoln TM. Cyclic GMP and mechanisms of vasodilation.Pharmac Ther 1989;41:479-502.

48 Murray KJ. Cyclic AMP and mechanisms of vasodilators.Pharmac Ther 1990;47:329-45.

49 Krall JF. Receptor-mediated regulation of tension insmooth muscle cells. In: Jenne JW, Murphy S, eds. Drugtherapy for asthma: research and clinical practice. NewYork: Dekker, 1987:97-128.

50 Polson JB, Krzanowski JJ, Szentivanyi A. Inhibition of ahigh affinity cyclic AMP phosphodiesterase and relaxationof canine tracheal smooth muscle. Biochem Pharmacol1983;31:3403-6.

51 Silver PJ, Hammel LT, Perrone MH, Bentley RG,Bushover CR, Evans DB. Differential pharmacologicsensitivity of cyclic nucleotide phosphodiesteraseisozymes isolated from cardiac muscle, arterial and airwaysmooth muscle. Eur J Pharmacol 1988;150:85-94.

52 Torphy TJ, Cieslinski LB. Characterization and selectiveinhibition of cyclic nucleotide phosphodiesteraseisozymes in canine tracheal smooth muscle. MolecPharmacol 1990;37:206-14.

53 Elliott KRF, Berry JL, Bate AJ, Foster RW, Small RC.The isoenzyme selectivity of AH 21-132 as an inhibitor ofcyclic nucleotide phosphodiesterase activity. J EnzymeInhibition 1990;4:245-5 1.

54 Nicholson CD, Shahid M, van Amsterdam RGM,Zaagsma J. Cyclic hucleotide phosphodiesterase (PDE)isoenzymes in bovine tracheal smooth muscle, theirinhibition and the ability of isoenzyme inhibitors to relaxprecontracted preparations. Eur J Pharmacol 1990;183:1097-8.

55 Bergstrand H, Lundquist B. Partial purification and charac-terization of cyclic nucleotide phosphodiesterase fromhuman bronchial tissue. Molec Cell Biochem 1978;121:9-15.

56 Cieslinski LB, Reeves ML, Torphy TJ. Cyclic nucleotidephosphodiesterase (PDEs) in canine and human trachealsmooth muscle (TSM) [abstract]. FASEB J 1988;2:A 1065.

57 Polson JB, Krzanowski JJ, Szentivanyi A. Correlationbetween inhibition of a cyclic GMP phosphodiesteraseand relaxation of canine tracheal smooth muscle. BiochemPharmacol 1985;34:1875-9.

58 Murray KJ, England PJ, Hallam TJ, et al. The effects ofsiguazadon, a selective phosphodiesterase inhibitor, onhuman platelet function. Br J Pharmacol 1990;99:612-6.

59 Torphy TJ, Burman M, Huang LBF, Tucker SS. Inhibi-tion of the low K, cyclic AMP phosphodiesterase in intacttrachealis by SK&F 94836: mechanical and biochemicalresponses. J Pharmacol Exp Ther 1988;246:843-50.

60 Torphy TJ, Zhou H-L, Burman M, Huang LBF. Role ofcyclic nucleotide phosphodiesterase isozymes in intactcanine trachealis. Molec Pharmacol 1991;39:376-84.

61 Bryson SE, Rodger IW. Effects of phosphodiesteraseinhibitors on normal and chemically-skinned isolatedairway smooth muscle. Br J Pharmacol 1987;92:673-81.

62 Rossing TH, Drazen JM. Effects of milrinone on contractileresponses of guinea pig trachea, lung parenchyma andpulmonary artery. J Pharmacol Exp Ther 1986;238:874-9.

63 Harris AL, Connell MJ, Ferguson EW, et al. Role of lowK, cyclic AMP phosphodiesterase inhibition in trachealrelaxation and bronchodilation in the guinea pig. JPharmacol Exp Ther 1989;251:199-206.

64 Schneider HH, Schmiechan R, Brezinski M, Seidler J.Stereospecific binding of the antidepressant rolipram tobrain protein structures. Eur J Pharmacol 1986;127:105-15.

65 Murray KJ, Eden RJ, England PJ, et al. Potential use ofselective phosphodiesterase inhibitors in the treatment ofasthma. Agents Actions (in press).

66 Small RC, Berry JL, Boyle JP, et al. Biochemical andelectrical aspects of the tracheal relaxant action ofAH 21-132. Eur JPharmacol 1991;192:417-26.

67 Gristwood RW, Sampford KA. Inhibition of histamine-induced bronchoconstriction by SK&F 94836, salbutamoland theophylline [abstract]. Br J Pharmacol 1987;92:631P.

68 Osborn RR, Ferracone JD, Novak LB, Hand JM. Effects ofrolipram (RO) enantiomers on in vitro and in vivo guineapig pulmonary models [abstract]. FASEB J 1990;4:A615.

69 Bewlev JS, Chapman ID. AH 21-132, a novel relaxant of

522



j.com/


ownloaded from



airway smooth muscle [abstract]. Br J Pharmacol 1988;93:52P.

70 Beume R, Eltze M, Kilian U. Bronchodilation due to

inhibition of cAMP-PDE III: preclinical results on thepyridazinone derivative B 842-90. Arch Pharmacol 1987;335(suppl):R76.

71 Ho PPK, Esterman M, Wang L, et al. LY 186655, 5,6-diethoxybenzo(B)thiophene-2-carboxylic acid, sodiumsalt, a new orally active bronchodilator [abstract].Pharmacologist 1987;29:173.

72 Hall IP, Walker D, Hill SJ, Tattersfield AE. Effect ofisozyme selective phosphodiesterase inhibitors on bovinetracheal smooth muscle tone. Eur J Pharmacol1990;183: 1096-7.

73 Hall IP, Donaldson J, Hill SJ. Modulation of carbachol-induced inositol phosphate formation in bovine trachealsmooth muscle by cyclic AMP phosphodiesterase inhibi-tors. Biochem Pharmacol 1990;39:1357-63.

74 Lichtenstein LM, Margolis S. Histamine release in vitro:inhibition by catecholamines and methylxanthines.Science 1968;161:902-3.

75 Schild H. Histamine release and anaphylactic shock inisolated lungs of guinea-pigs. Q J Exp Physiol 1936;26:165-79.

76 Dent G, Rabe K, Giembycz MA, Barnes PJ. Inhibition ofeosinophil respiratory burst activity by type IV-, but not

type 111-selective cAMP phosphodiesterase inhibitors[abstract]. Am Rev Respir Dis 1990;141:A878.

77 Godfrey RW, Manzi RM, Gennaro DE, Hoffstein ST.Phospholipid and arachidonic acid metabolism inzymosan-stimulated human monocytes: modulation bycAMP. J Cell Physiol 1987;131:384-92.

78 Takenawa T, Ishitoya J, Nagai Y. Inhibitory effect ofprostaglandin E5, forskolin, and dibutyryl cAMP on

arachidonic acid release and inositol phospholipidmetabolism in guinea pig neutrophils. J Biol Chem 1986;261:1092-8.

79 Zurier RB, Weissmann G, Hoffstein S, Kammerman S, TaiHH. Mechanisms of lysosomal enzyme release fromhuman leukocytes. II. Effects ofcAMP and cGMP auton-

omic agonists and agents which effect microtubule func-tion. J Clin Invest 1974;53:297-309.

80 Foreman JC, Hallett MB, Mongar JL. The relationshipbetween histamine secretion and 45-calcium uptake bymast cells. J Physiol (London) 1977;271:193-214.

81 Nielson C. f,-Adrenergic modulation of the polymorpho-nuclear leukocyte respiratory burst is dependent upon themechanism of cell activation. J Immunol 1987;139:2392-7.

82 Rasmussen H, Good DBP. Relationships between calciumand cyclic nucleotides in cell activation. Physiol Rev1977;57:421-509.

83 Undem Bj, Graziano FM, Buckner CK. Effects of isopro-terenol on histamine release induced from monodispersedguinea pig lung cells by different secretagogues. J Phar-macol Exp Ther 1985;233:617-22.

84 Undem BJ, Torphy TJ, Goldman D, Chilton FH. Inhibi-tion by adenosine 3':5'-monophosphate of eicosanoid andplatelet-activating factor biosynthesis in the mouse PT- 18mast cell. J Biol Chem 1987;265:6750-8.

85 White JR, Ishizaka T, Ishizaka K, Sha'afi RK. Directdemonstration of increased intracellular concentration offree calcium as measured by quin-2 in stimulated rat

peritoneal mast cell. Proc Natl Acad Sci USA 1984;81:3978-82.

86 Undem BJ, Peachell PT, Lichtenstein LM. Isoproterenol-induced inhibition ofimmunoglobulin E-mediated releaseof histamine and arachidonic acid metabolites from thehuman lung mast cell. J Pharmacol Exp Ther 1988;247:209-17.

87 Kaliner M, Orange RP, Austen KF. Immunological releaseof histamine and slow reacting substance of anaphylaxisfrom human lung. IV. Enhancement by cholinergic andalpha adrenergic stimulation. J Exp Med 1972;136:556-67.

88 Peachell PT, Undem BJ, Schleimer RP, Lichtenstein LM,Torphy TJ. Action of isozyme-selective phosphodies-terase (PDE) inhibitors on human basophils [abstract].FASEB J 1990;4:A639.

89 Simpson AWM, Reeves ML, Rink TJ. Effects of SK&F94120, an inhibitor of cyclic nucelotide phosphodiesterasetype III, on human platelets. Biochem Pharmacol1988;37:231 5-20.

90 Neilson CP, Vestal RE, Sturm RJ, Heaslip R. Effects ofselective phosphodiesterase inhibitors on the poly-morphonuclear leukocyte respiratory burst. J Allery ClinImmunol 1990;86:801-8.

91 Thompson WJ, Ross CP, Pledger WJ, Strada SJ, BannerRL, Hersh EM. Cyclic adenosine 3':5'-monophosphatephosphodiesterase. Distinct forms in human lymphocytesand monocytes. J Biol Chem 197f 251:4922-9.

92 Robicsek SA, Krzanowski JJ, Szentivanyl A, Polson JB.High pressure liquid chromatography of cyclic nucleotidephosphodiesterase from purified human T-lymphocytes.Biochem Biophys Res Commun 1989;163:554-60.

93 White JR, Torphy TJ, Christensen SB IV, Lee JA, MongS. Partial purification of the rolipram sensitive phos-phodiesterase from human monocytes [abstract]. FASEBJ 1990;4:A1987.

94 Macphee CH, Harrison SA, Beavo JA. Immunologicalidentification of the major platelet low-K,, cAMP phos-phodiesterase: probable target for antithrombic agents.Proc Natl Acad Sci, USA 1986;83:6660-3.

95 Frossard N, Landry Y, Pauli G, Ruckstuhl M. Effects of

cyclic AMP- and cuyclic GMP-phosphodiesteraseinhibitors on immunological release of histamine and onlung contraction. Br J Pharmac 1981;73:933-8.

96 Busse WW, Anderson VL. The granulocyte response to thephosphodiesterase inhibitor Ro 20-1724 in asthma.J Allergy ClGn Immunol 1981;67:70-4.

97 Nicholson CD, Challiss RAJ, Shahid M. Differentialmodulation of tissue function and therapeutic potential ofselective inhibitors of cyclic nucleotide phosphodiesteraseisoenzymes. Trends Pharmacol Sci 1991;12:19-27.

98 Robicsek SA, Blanchard DK, Djeu J, Krzanowski JJ,Szentivanyl A, Polson JB. Attenuation of human T-lym-phocyte blastogenesis by the selective inhibition ofcAMP-phosphodiesterase [abstract]. FASEB J 1990;4:A337.

99 Chilton FH, Schmidt D, Torphy T, Goldman D, UndemB. cAMP inhibits platelet activating factor (PAF) bio-synthesis in human neutrophil [abstract]. FASEBJ 1989;3:A308.

100 Plaut M, Marone G, Gillespie E. The role of cyclic AMP inmodulating cytotoxic T lymphocytes. II. Sequentialchanges during culture in responsiveness of cytotoxiclymnphocytes to cyclic AMP-active agents. J Immunol1983;131 :2945-52.

101 Averill LE, Kammer GM. Inhibition of interleukin 2production is mediated by a cyclic adenosine mono-phosphate (cAMP)-dependent pathway (abstract]. ClinRes 1985;33:839A.

102 Cooper KD, Kang K, Chan SC, Hanifin JM. Phosphodies-terase inhibition by Ro 20-1724 reduces hyper-IgE syn-thesis by atopic dermatitis cells in vitro. J Invest Dermatol1985;84:477-82.

103 Lee JC, Votta B, Dalton BJ, Griswold DE, Bender PE,Hanna N. Inhibition ofhuman monocyte IL- I productionby SK&F 86002. IntlJ Immunother 1990;6:1-12.

104 Renz H, Gong J-H, Schmidt A, Nain A, Gemsa D. Releaseof tumor necrosis factor-x from macrophages. Enhance-ment and suppression are dose-dependently regulated byprostaglandin E, and cyclic nucleotides. J Immunol 1988;141:2388-93.

105 Muggli R, Tschopp TB, Mittelholzer E, Baumgartner HR.7-Bromo-1 ,5-dihydro-3,6-dimethylimidazo(2,1 -b)quin-azolin-2(3H)-one (Ro 15-2041), a potent antithromboticagent that selectively inhibits platelet cyclic AMP-phos-phodiesterase. J Pharmacol Exp Ther 1985;235:212-9.

106 Seiler S Arnold AJ, Grove RI, Fifer CA, Kelly SL, StantonHC. Effects of anagrelide on platelet cAMP levels, cAMP-dependent protein kinase and thrombin-induced Ca++fluxes. J Pharmacol Exp 1987;243:767-74.

107 Rudd RM, Gellert AR, Studdy PR, Geddes DM. Inhibi-tion of exercise-induced asthma by an orally absorbedmast cell stabiliser (M&B 22,948). Br J Dis Chest 1983;77:78-86.

108 Reiser J, Yeang Y, Warner JO. The effect of zaprinast (M&B22,948, an orally absorbed mast cell stabilizer) on exercise-induced asthma in children. Br J Dis Chest 1986;80:157-63.

109 Leeman M, Lejeune P, Meot C, Naeije R. Reduction inpulmonary hypertension and in airway resistance byenoximone (MDL 17,043) in decompensated COPD.Chest 1987;91:662-6.

110 Israel E, Mathur PN, Tashkin D, Drazen JM. LY 186655prevents bronchospasm in asthma of moderate severity[abstract]. Chest 1988;94:71S.

111 Foster RW, Rakshi K. Bronchodilator potency, effective-ness and time course of inhaled nebulised AH 21-132 innormal human subjects [abstract]. Br J Pharmacol 1990;99: 193P.

112 Small RC, Foster RW, Berry JL, Chapman ID, ElliottKRF. The bronchodilator action of AH 21-132. AgentsActions (in press).

113 Naccarelli GV, Goldstein RA. Electrophysiology of phos-phodiesterase inhibitors. Am J Cardiol 1989;63:35-40A.

114 Horowski R, Sastre-Y-Hernandez M. Clinical effects of theneurotropic selective cAMP phosphodiesterase inhibitorrolipram in depressed patients: global evaluation of thepreliminary reports. Curr Ther Res 1985;38:23-9.

115 Bobon D, Breulet M, Gerard-Bandenhove MA, et al. Isphosphodiesterase inhibition a new mechanism ofantidepressant action? A double blind double-dummystudy between rolipram and desipramine in hospitalizedmajor and/or endogenous depressives. Eur Arch PsychiatrNeurol Sci 1988;238:2-6.

116 Eckmann F, Fichte K, Meya U, Sastre-Y-Hernandez M.Rolipram in major depression: results of a double-blindcomparative study with amitriptyline. Curr Ther Res1988;43:291-5.

117 Wachtel H. Potential antidepressant activity of rolipramand other selective cyclic adenosine 3',5'-monophosphatephosphodiesterase inhibitors. Neuropharmacology 1983;22:267-72.

1 18 Sturgess I, Searle GF. The acute effect ofthe phosphodies-terase inhibitor rolipram on plasma osmolality. J ClinPharmacol 1990;29:369-70.

119 Lugnier C, Schoeffter P, LeBec A, Strouthou E, Stoclet JC.Selective inhibition of cyclic nucleotide phosphodies-terases ofhuman, bovine and rat aorta. Biochem Pharmacol1986;35:1743-51.

120 Souness JE, Brazdil R, Diocee BK, Jordan R. Role ofselective cyclic GMP phosphodiesterase inhibition in themyorelaxant actions of M&B 22,948, MY-5445, vinpoce-tine and 1-methyl-3-isobutyl-8-(methylamino) xanthine.Br J Pharmacol 1989;98:725-35.

523



j.com/


ownloaded from


Documents

Thorax New drugsreview · phosphodiesterases are responsible for cyclic nucleotide hydrolysis in different tissues. Attempts are being made to increase tissue selectivity by synthesising