Bloehimica et giophysim Acta, 11340992) 149-t$6 © 1092 Elsevier Science PublishersB.V, Aa rights reserved0t67-4889/92/$05.00


BBAMCR 13120

Purification and properties of the cGMP-inhibited cAMP phosphodiesterase from bovine aortic smooth muscle A n a R a s c 6 n ~, S a m L i n d g r e n b L a r s S t a v e n o w ¢, P e r B ¢ l f r a g e ", K a r I - E r i k A n d e r s s o n b, V i n c e n t C. M a n g a n i ¢ l l o d a n d E v a D e g e r m a n


Depcrzments of Medical und PIwsiolo~icat CheminrF~ ~ ClinWul Pharmacology, Lun~ and r Internal Mt~llclae, Malrnil ~:¢neml Plospltal, Land Uni*:ersity ~$.l~eden) and ": Laboratory of Cdlular Metabolism, Iffationa[ I-le~rt, Lung and Blood Institute, National Institutes oF Hea(lk. Bethcsda, MD (USA)

(Received 17 Seplemher 1991)

Key words: cGMP-inhibited~1i¢ nucleotideph~phodiestetage~,Smt~lh mnscle;Phospbodiesteraseinhihilor,(Bovineaorta) Pure cGMP-inhibited cAMP pbosphodiesterase (eGI-PDE) in/.Lg quantities was isolated from bovine aortic smooth muscle after mote than 5fl00-fold purification using DEAE iou-e~henge and affinity ¢hromatograp~ with a det'wative of the specific cGI-PDE inhibitor cilostamide ¢cojugatcd as a ligand to aminnethyl agarose (CIT-agawse). The cGI-PDE, which constituted about half of the high affinity cAMP-PDE activity of a tissue homogenate, was identified with a 105-kDa protein on SDS-PAGE through use of antibodies towards the human platelet, bovine cardiac and tar/the adipose tissue eGI-PDE in Western blot and inanunoprecipitation/immunoinectivation analysis. As obset~d during purification of the enzT/mefrom otbor tizst~s the enzyme protein wan ¢xquisilely sensitive to proteolytic nicking during purification, resulting in several 30--77-kDa l ~ i d e fragmenls. Rapid tmmunopmcipitation from fresh ti&~ae extracts was the only way found to partially prevent the prmeolysis. The native enzyln¢ had apparent molecular sizes of approx. I(KI000 or, mainly approx. 220000 by 8el chR"matograpby, presumably indicating the presence of mmomerie and dimerie forms. The enzyme hydroiyzcd cAMP and eGMP with normal Michaelis-Menten kinetics with K m of 0.16 and 0.09 pM, respectively, with V,~,~ for hydrolysis of cGMP of IL3 compared to 3.1 ~mol/min per mg protein for cAMP. The enzyme was potemly and selectively inhibited by cGMP (ir-'s0 -~ 0.25/tM) and the cardiotonic/vasndilatory drugs OIN2-3911 (a cilostamid¢ derivative), milrinone and C1~930 ( I C s ~ 0.05, 0.40 and 0.25 ~M, re,s ~ l y ) . "I'nc cGI-PDE was pbospbo~intcd by eAMP-depentknt protein kinasc as has been reported for the analelp:ms enzymes in heart, adipose/isaac and plalelets. The identification of a cGi-PDE in the aortic smooth muscle and its inhibitor spexifg'ity is consistent with the hypothesis that inhihilion of this enzyme is important in the mechanism through which these drugs imxluce vasorelaaation.

lalreductlea Vasorelaxant actions of the bipyridine milrinone, the cilo~tamide derivative OPC-3911 and the dihydropyradizone derivative C1-930 have been described Abbreviations: PDE. o~lic ,ucleotide plmsphodiesterasc;vGI-PDE, cGMP-inh~hed PDE; CIT-,Mamse,N-(2-isothioeyaaato)ethylderivativeof cilostarnldeconjugatedto amlaOethy[ailmmc; 5DS, radium dodeeyl ~lfate; PAGE, polyact~lamidegel ekctroplmresis; anti-cGlPDE, aatilaxlies (ItG fzactiea) a~imt "cGI-PDE~ cAMP-Prlg, cAMP-dependent protein Idnase; OPC-39II. H-c~lobe~t-N-2.hydt~yelh.vl-4-(6-(1,2-dlhl/dm-2-oxoquim~lo~ff))bul)'ramide;Ro 201724, 4-(3-butyo~.4-metho~benzyD-2-1midarolidong milriaon¢, 1,6-dlhy~O-2-m©thyl-6-md3,4'-hiplaidine]-5-carbonit rile; CI-930, 4,5.dlhydro-6-[4-(IH, imtdazol-l-yl~hcnyl]-,5-mcllffl-3(2H )-iryfiazQlle, Correspoadence: A. Resc6n, Department of Medical aad Physiological Chemistvj,Unlvemtv of Lond, P.O. Box 94. $-221 Q0 Lead~ Sweden.

in a number of in vitro and in v i m studies (1-7]. These drugs arc selective int~'bitors of a cGMP-inlfibited, knv K m cAMP phospbediestcrase (hereafter referred to as cGI-PDE), one of the multiple molecular forms of cyc'ic nudeotide phosphediestemses found in mammalian tissues [8,9]~ A correlation between the inhibitory and vasodilatory potency o f these drugs in intact vessel preparatiom has been found [6] and it has been suggested that cyclic A M P elevation through cGIPDE inhibition could be an important part of their mechanism of action [5]. Recent work from our laboratories with rat aorta preparations [10] has further supported this hypothesis by indicating that an enzyme activity with tim properties of a cGI-PDE was present in this tissue and that inhibition of this isozyme might be of primary importance for the observed relaxant effects of OPC-3911, CI-930 and milrinone on tat aorta rings contracted by phenylephtine, seromnia or K+-depolarization.

150 However, as in all previous studies with vascular smooth muscle preparations, the identification of the cGI-PDE in the aorta was based on chromatographic behaviour and inhibition properties. The enzyme abondance in all tissues studied is extremely low; insufficient amounts have prevented any molecular identification and characterization of the enzyme in the rat aorta. Obviously, isolation and characterization of the enzyme molecule are necessary prerequisites for further studies aimed at establishing its role in the vasoreiaxant effect of the drugs in question, as well as elucidating the molecular mechanisms by which the cGIPDE affects vascular smooth muscular contraction. The requirement for direct molecular characterization of the cGI-PDE is also underscored by the findings that cGI-PDE isoenzymes from various tissues differ significantly in basic properties, e.g., molecu!ar size [8,11-15] as well as eDNA-deduced amino acid sequence (unpublished results). For these reasons we have used the cilostamide affinity chromatography technique [8] to develop a procedure for purification of the enzyme from bovine aortic smooth muscle. Enough pure enzyme has been obtained to compare some of its molecular, immunological and enzynu31ogical properties with those of isoenzymes from other tissues.

Materials and Methods [2,8-3H]cAMP (36.4 Ci/mmol) and [&5-3H]cOMP (approx. 30 Ci/mmol) from New England Nuclear, Dreieich, Germany, were purified as described previously [8]. The heterogeneous non-ionic alkyl polyo~ethylene glycol detergent CI3Et 2 (abbreviated as C,E~, from the general ~ormula C,H2,÷, (OCH2 CHz)xON) was obtained from Bero[ Kemi, Stcnungsund, Sweden. Lyophilize& Staphylococcus aureus (a source of protein A), the catalytic suhunit of cAMP-dependent protein kinase (cAMP-PrK) (bovine heart) and cAMP-PrK inhibitor (rabbit s~quenee) were obtained from Sigma, St. Louis, MD. OPC-3911 (Ncyclohexyl-N-2-hydroxyet hyl-4-(6-(1,2-dihydro-2-oxequinolyloxy))butyramide) was generously supplied by Dr. H. Hidaka of Nagoya University, Japan; CIT-agarose (N(2-isothiocyanato)ethyl derivative of cilostamide, coupled to aminocthyl agarose) was prepared as described previously [8]; Re 20-1724 (4-(3-butyoxy-4methoxybenzyl)-2-imidazolidone) by Dr. M. Lin, National Institute of Health; CI-930 (4,5-dihydro-6-[4( 1H-imidazol-I -yl)phenyl]-5-methyl-3(2 H )-pyridazone) by Dr. R. Weishaar of Warner Lambert, Ann Arbor, MI; milrinone (t,6-dihydro-2-methyl-f-o×o~3,4'-bipyridine]-5-carboxinitrile) by Dr. A. Sofia of Sterling Winthrop Research Institute. Rensselaer, NY. The double antibody Proto Blot Western Blot AP System using alkaline phosphatasc conjugates was from P~omega (Madison, WI). Potyclonal antibodies against

bovine adipose tissue cGI-PDE and against human platelet eGI-PDE and the corresponding IgG fractions were prepared essentially as described [ll]. A monoelonal antibody against the bovine heart cGI-PDE was kindly supplied by Dr. Joseph Beavo, Department of Pharmacology, School of Medicine, University of Washington, Seattle, WA. All antibody preparations will be referred to as anti-eGI-PDE, specifying the source in each case. All other chemicals were of analytical grade and obtained from sources reported previously [8,1 l].

Phosphodiesterase assay High-affinity cAMP- and cGMP-PDE activities were assayed at 0.5 tim [3H]cAMP or [3H]cGMP as described in Ref. 8. For inhibition studies, the assay was performed in the presence of OPC-3911. CI-930, milrb none, eGMP or Re 20.1724. cGI-PDE activity was measured as the cAMP-PDE activity inhibited by 3 /~M OPC-391I. To monitor Ca2~/caimodulin stimulated PDE activity, the assay was also performed in the presence of CaZ+/calmodulin (2 raM, 0.2 ~g/ml). Isolation of cGI.PDE from aortic smooth muscle Bovine aortas, obtained from a local slaughter-house, were transported in 0.15 M NaCI at 4°C and processed as soon as possible (generally within 5 h). All purification steps were carried out at 4°C. Medial smooth muscle tissue (appmx. 200 g from approx. 2 k8 of aortas) was dissected out, cut into small pieces (5 X 5 mm) and homogenized in 200 ml of 20 mM Tris (pH 7.5) containing I mM EDTA, 250 mM sucrose plus leupeptin, antipain, and pepstatin (each 10 t+g/ml) with a knife homogenizer (15000 rpm for 5 x 15 s). The homogenate was centrifuged (4000 x g, 10 rain) and the resulting supematant was centrifuged again at 50000 x g , 1 h. More than 90% of total hmmogenate cAMP-PDE activity (assayed with 0.5/zM cAMP) was recovered in the 50000 x g supernatant and used for further purification. The supernatant was applied to a DEAE-Sepi'.acel column previously equilibrated with 50 mM sodium acetate buffer (pH 7.0) containing 10% glycerol (w/v), 1 mM EDTA, 1 mM dithioerythritol, 1 /zg/ml let,peptin, 1 ~ g / m l pepstatin and 10 ~tl/ml antipain (buffer A). After extensive washing, the eGI-PDE was eluted using a linear sodium acetate gradient as detailed in Fig. I. Fractions containing eGI-PDE activity were peeled and dialyzed (12 h × 2) against 50 eel. of 50 mM Tris (pH 7.5) containing 5 mM MI~CI2, 1 mM EDTA, 10% (w/v) glycerol, t ~ g / m l leupeptin and 1 p g / m l pepstatiu (buffer 13). The dialyzed material was applied to a 5 ml CIT.agarose affinily column, equilibrated with buffer B containing 0.03% (w/v) Cl3Et2 (buffer C); the inclusion of this low concentration of

151 detergent was important for maintaining stability of the enzyme (of. [8]). The column was first washed with 20 ml of buffer C containing 2 M NaBr and thereafter with 50 ml of buffer C containing 100 m M NaBr. To elute the enzyme, the gel was incubated at 4°C for 6 h with 10 ml of buffer C containing 100 m M NaBr and 50 mM cAMP. The eluate was collected and the gel washed with 30 ml of buffer C containing 100 mM NaBr. To remove the cAMP, the cluate was applied to a small (0.3 ml) DEAE-Sephacel column, equilibrated with buffer C containing 100 m M NaBr. After extensive washing the enzyme activity was eluted with buffer C containing 200 mM NaBr.

Polyacrylamide slab gel electro~.oresis and western blot analysis The cGI-PDE was subjected to electrophoresis under non-denaturing conditions as described by Davis [16] using 5.5% acrylamide gels. Slab gels were run at 100 V for 45 rain at 5°C before addition of the sample to remove persulfate and other oxidants, which otherwise inactivated tbc enzyme. At the end of the run, the lane was cut in 2 m m pieces, to each piece was added 200 /M of PDE-assay buffer and the assay was performed as described above, cGI-PDE was also subjetted to electrophoresis in 8% polyacrylamide slab gels (SDS-PAGE) ~ccording to Lacmmli [17]. The transfer of the protein to nitrocellulose and the subsequent immtmoblotting analysis using the human platelet, heart or bovine adipose tissue anti-cGI.PDE (diluted 5fl0-times) and visualization using alkalinephosphatase-labelled secondary goat anti-mouse or goat anti-rabbit lgG were per[onned as described by Towbin [18]. The antt"oodies were sometimes blocked

by preincubation with p u ~ human plat¢let eGI-PDE for 6 h -it 4°C.

lmmunoisolation of cGI-PDE A 50000 × g aorta smooth muscle supernatant was prepared within 5 h after slaughter. For immunoisolalion of cGI-PDE, freshi,j" prepared superuatanls were incubated with h u m a n plateiet anti-cG1-PDE (corresponding to 1 pl sca'um/pmol per min cGI-PDE) for 16 h at 4°C. S. sure,-bound protein A prepared and washed as described by Doolittle e t a l . [19] to reduce non-specific interactions, was added (10 pd of a 10% slurry/serum) and incubation continued for 30 rain at 4°C. ]mmunoprecipitates were isolated by centrifugation (5 min, Eppendorf mictocenttifuge. 4°(2). washed free times in phosphate-buffered saline (pH 7.3) containing 0.1% N-lam'yl sarcosine and resuspended in SDS-PAGE sample buffer. The cG1-PDE was incubated overnigkt at 4~C with different amounts of anfi..cGl-PDE. S. aureus-1~mmd protein A was added (2.5 p.g o | a 10% slurry//xg |gG) and after 30 rain the imm,.taoprecipitates were collected by centdfugation 5 rain in an Eppendorf microcentrifuge at 4°(2. Supernatants were a s s a ~ d for cGIPDE activity. When direct inhibition of cGI-PDE was studied, cOI-PDE assay was performed prior to addition of S. aureus-bound protein A. Controls with preimmune IgG were run in parallel to monitor inactivation.

eGI-PDE phosphowlation cGI-PDE it. crude tissue preparations or after purification was phosphor,~ated with the catalytic subuuit of eAMP-PrK and T32P]~TP as detailed in Fig. 4, I.S

p @ v k II


? \ t~





, 1.0




o o





noo~ Fr~tetion i~umlOer


Fig. l. DlEAE~Ccphac~i ebmmatciwaphy. 200 ml o~"5DO(XIx & supcrnalant from bovine aortic smooth mine,tie homogeneme was ~ d onto a 5>: 12 cm DEAE-Scphacol co[mrm, equilibrated at pl[ Z0 in b~ffer A (see Materials and Methods). Alter ~ the column with this buffer.

the PDE activitieswere elated (20 ml/h) with 3~0 ml of a linear sodium acetategradient from 50 to 1200mM as indk~ted. POE activity wes assayedin duplicatesusingcAMP (a) and cGMP (,~) as subslrates( pM) and protein mea~medacco~rdingto Bradfo~.

152 eAMP-PDE sensitive to this inhibitor cocluted with eGI-PDE, Pooled peak il was further affinity purified on CITagarose, a procedure known to be highly selective for eGI-PDEs [8,11]. The Ro-1724-sensitive en23,me did not bind to the column but all of the OPC-inhibired PDE activity in peak II did, as indicated by its absence in the fiery-through fraction. About 20% of the adsorted cGI-PDE activiW could be eluted with cAMP (Table I). This purification step resulted in an almost 3000-fold increase of specific cGI-PDE activity. The overall recovery of the cGI-PDE from the 50000 × g supernatant was 10% with approx. 5O00-fold purification. The purified en~me contained several poiypeptide species (105, 77/74, 53 and 30-kDa) revealed by SDSPAGE and silver staining (Fig. 2A). This was likely the result of proteolytie nicking of a I05-kDa native species during the purification, in analogy to what has been found in previous studies [8,11-14], as after PAGE was carried out under non-denaturing conditions, practically all (>90%) protein corresponding to that dee tected by silver stain exhibited PDE activity (data not shown) (cf. [8]). The identification of the eG1-PDE with the 105-30-kDa silver-stained bands on SDS-PAGE was further verified by their recognition with the polycionai human platelet and the monoclonal bovine cardiac anti-cGl°PDEs (Fig. 2B, C) (the 53- and 30-kDa poi~,peptides did not contain any epitope for recognition by the monoclooal antibody). The immunoreaetive bands in lanes B and C were absent when the antibodies had been blocked by preincubalion with pure human platelet cGI-PDE as described in Materials and Methods (data not illustrated), It was technically difficult to demonstrate the small amount of unprotcolyzed 105-kDa eOI-PDE in pure enzyme preparations. Therefore, cGI-PDE was also directly immunoprocip-

Protein determination Protein concentrations in crude enzyme fractions were estimated by the method of Bradford [20]. Because of the Iow amounts of ¢GI-PDE pure enzyme, protein was estimated according to the method of Stoschek [21] using colloidal gold to detect submicrogram amounts of protein (20-100 rig). Bovine serum albumin was used as standard protein. Values obtained with the latter method have been increased by a factor of 1.6 to compensate for the difference to cGI-PDE protein concentrations measured by the Lowry method [22], in order to allow comparisons with previously reported enzyme preparations. Results

Purification and properties of cGI-PDE fr~,n I~ovine aorta Supernatanls (50000 ×g) from 200 g of bm'ine aortic smooth muscle were used as starting material for purification of the cGI-PDE. This fraction contained > 90% of total smooth ['au~:.le homogenat- eAMP-PDE activity (assayed with 0.~/tM cAMP). About 50% of the supernatant enzyme activity could ee inhibited by OPC~3~tl (Table I), and by the bovine adipose tissue anti-cGI-PDE (data not shown), indicating that cGIPDE constituted half of the high-affinity cAMP-PDE activity of the tissue. DEAF_,-Sephace[chromatography of the 50000×g supernatant resulted in the separation of two peaks of PDE activity. (Fig. 1), one of which (peak I) preferentially hydrolyzed cGMP, was stimulated by Ca2+/calmodulia and insensitive to OPC-3911 (data not illustrated). The c~her (peak II), which preferentially hydro!y~e.d cAMP, was parti~.lly inhibited by 3 v M OPC-3911 (Table I) and by anti-cGI-PDE (about 50% maximal inhibition). The remaining cAMP.PDE activity could be inhibited by Ro-1724, indicating that a

TABLE 1 Purification of cOI-PDE from bovine aortic s m ~ h muscle 4200 g) Fraclion

Tolal protein (ms)

Total cAMPPDE activity " (am~/min)

OPC-inhtbiled activity b tnmol/min)

50000× g supernatanl


[ 865











DEAE-Sephace] (peak I!) CIT-agarose affinity chromatography (incl. dialysis and removal of cAMP)

0.037 d

Specific activity ~ (nmol/min per mg protein)

2 324

Purification (-fold)

5 ~,3

0.5 tiM [~IIIcAMP, b A mcasttrc of cGI-PDE aczivity calculated as: tactivity in absence of OPC-39t ))-(activi,'~' in presence of 3 p,M OPC-3911). 0PC-3911-inhibited activilY. '~ Value ~tainecl by lhe colloidal sold rnelhod, corrected to cnr~spond to Lowly values (se~ Matezlals ~nd Methods.).

Yield c (%) 100



153 Rated from the 500OO × g supcrnatants of fresh bovine aorta smooth muscie homogenate with the polyc[onal human platelet anti-eGl-PDE and S. au~us protein A followed by western blot of the immunoprceipitate (Fig. 2D). This procedure has also been the only way in our hands to partially prevent the protcol~ic nicking, in analogy with what has been found during isolation of the enzyme from other tissues [13], and allowed the clear demonstration of unproteolyzed 105-kDa cGIPDE. For unknown reasons the bovine adi]'cge tissue anti-cGi-PDE did not react with the pure cOl-PDIf on Wcstern blots, but inactivatcd/immunoprecipitated the native eaz'yme similarly to what has been previously found for the rat and bovine adipose tissue cGI-PDE (data not illustrated) (of. I8,111). On gel chromatography the pure cGI-PDE was elated in two peaks of enzyme activity with apparent molecular masses of a approx. 220 and 100-kDa (Fig. 3) (70% and 30% of the ¢GI-PDE activity, respectively), presumably representing di- and monomeric forms of the enzyme. The aorta smooth muscle cGI-PDE was a substrate for phospborylation by eAMP-PrK (Fig. 4). Intact 105kDa enzyme in fresh aorta extracts, identified by blocking the immunoprecipitating antibody with pure cGIPDE, was phosphorylated by the catalytic subunit of






U.4 0 Or^




0.$ "~t°'w'*k ~


0,1 log MY/ Fig. 3. Gel-filtration eh~atatolPrapby. Pule cGI-PDE was ¢hmmatographed on a Sepl,mcryl S-3~ column (IJ~X83 cm) equilibratted in buffer C containing 250 m~l NaBr, wilh a flow rate of IS ml/h. The apparent molecular sizes of the peaks of cGI-PDE activity wen: determined by comparison (using liaear lelPres,cion)~Rh standaxds: FEE, fe~ritin (Mr 440000);, 11~3,gamma globulin (Mr I f31][][]Ok bovine serum albumin (M, 670eOk OVA. ovailmmm (M, 43000). Inset- Elation palttem. Fractions o4' i ml were oa4k¢led and portions were assayed for cG1-PDE acibity.

the cAMP-PrK (Fig. 4.A, 13), with an estimated stoichiornetry of about 0.2 tool phosphate per tool ¢GIPDE (the amount of enzyme was calculated from the specific activity of the pure enzyme (Table !) and the cGI-PDE activity in the sample), it should be noted that this is likely to relprese~t a minimum value, since enzyme inactivation during thc purification would give the pure enzyme an e m m e ~ low specific activity compared to that ia fresh extracts. A 30-kDa protein present it, the crude tissue preparation was also pbesphorylated (Fig. 4A). However, its immunopre~pitazion was non.specific and occur~d also with a blocked antibody (Fig. 41]), indicating that R had no relation to the cGI-PDE. Pure cGI-PDE was phost~3~laled on the 105 and 77/74, but not on the 53 and 30-kDa pol~peptides (Fig 4C), indicating that omly the larger proteolytic fragments contain the pho~hor,/latcd sRc(s). The specific pcptide inhibitor of cAMP-PrK completely blocked the phosphorylation and no phos-

phorylation occurred fit the absence of added kinase





.~-.C'~ " " ~ . ~ , moisolated from a ,.xVF_~'~8" ~ ¢ m a I as . ~ t.~ to ~ .J'S-PAGE. Tile pro~ teias were diver stained [32] (A) ~r electro~loncd to nilv0cctlu]osc and incubated ~.'.t.h the pt~clopd human platetet anti cGI-PDE (13, D) or the raoaoclonal tovine cardiac anti-cGl-PDE (C). In lane D the presence of eres~Ractins proteins < 60-kDa could not he examined I~caasc d the psescacc (indicated by a star) of a large excess of lgG light and heaW chains, rec¢~,nized by the secondary antibody (goat antl.~bbit) coKiugat~ to a~,kaline u.scd to detect Ihe immunoree¢live protein ~ . ,~33].

(data not illustrated) (the = 40-kDa pltosphepeptide band in Fig, 4C rcwcscnts autoplmsphorylation of the catalytic subanit of the cAMP-i~'K added in equinmlar quantities with the cGI-PDE).

Kinetic properties and effect of ~ t o r s The pure cGI.PDE l~kob-zed cAMP and eGMP with linear Miehaelis-Mcnten kinet~ (not gluswated). The apparent K . values for cAMP and cGMP were similar whereas Vmx for cAMP hydrolysis was ten-fold that for cGMP (Table ID, In addition to cGMP and OPC-3911 the enzyme was sensitive to inhl"bifion by milrinone and CI-930 (IC~



Enzymologic~/ prol~nles of cGI.PDEs from diff~renJ tls~ues Bovine" ao~'ta smooth

Rat adipose=' tissue

Bovine= adipose tissue

Bovined heart

Bovine/human= ptat©lcts



Catalytic prolgrties Km(#M) cAMP cGMP

0.16 0.09

Vm= s (nmol/mJnper mR) cAMP 3.1 cGMP 0.3

Rat [ liver 2

0A i).3

0.3 0.8

0.18 0.02




0.3/28 10

8.5 2.0

2.5 1.6

6.0 O.b

3.0 0.3

6.2. 2.1

114/633 i 4|


Inhibitors(/tM} ]Cq I 0PC-3911






Re 20-1724 cGMP

> 30 h 0.25


0.6 0.4 1.90 0.2


0.005 u



0.26 +




1100 0.6

62 ~ 0.06 i


220 0.13


l ~0 2

• The present report,cG[-PDE activitywas assayedwith0.5 pM ['IHkAMPwithoutor with > 5 concentratio~nsof the [ndlcated inh~itors. ICso values were determined graphically. s From Degermanet al. [8~ = From Degemtanet al. [11]. d FgomHarrison¢t al. [13]. " FgomMcPhee el al.; Grant and Colt'nani23,12]+ r (I) From Bo'$es and Loten [15]; (2) Pyne ©t al. [14].

s ICsD-- concentraliort itihibiting 50% of the enzymeactivity. h The inhibition obtained with 30 tiM was tess than 20%. ZKt. J mU/rag. for OPC-39II < C]-930 ~ cGMP < milrinone), but not Re 20-1724 (Tabie ii). Discussion This report identifies a cGI-PDE in aorta smooth muscle, as previously predicted by indirect evidence in other studies [4,24-27]. Sufficient pure enzyme was isolated to allow comparison of its molecular and enzym o l n g i c a l properties with those o f the isoenzymcs from other tissues. The two m~jor problems encountered in attempts to isolate ¢GI-PDEs from various tissues are their extremely [ow tissue abundance and exquisite sensitivity to proteolytic nicking during the purification, From the present results, the tissue concentration can be ¢~timated to be in the order of about 1 ~ g / g smooth muscle (allowing for some inactivation during purification); thus, isolation of sufficient enzyme to allow at least partial molecular characterization required processing of kilogram amounts of bovine aorta, The extensive proteolyt[e nicking taking place during purification, with only a small fraction of the native 105-kDa enzyme protein remaining intact, made the identification of the purified proteins with eGI-PDE technically difficult. However, the immunological cross,reactivity with the cG1-PDE antibodies used in

western blots and in immunoinactivation experiments, the association of enzyme activity with practically all protein after non-denaturing PAGE, the d e a r demonstration of the 105-kDa protein after direct immunoprecipitation from fresh tissue extracts and the fact that this immunoprecipitation could be specifically prevented by blocking the antibody with pure platelet eGI.PDE, in combination, are strou~ evidence that the purified proteins constitute a ¢GI-PDE smooth muscle isoenzyme. As in provious work [13], rapid immunoprecipitation appeared to be the only way of avoiding some of the extensive proteo~ytic nicking which otherwise occurred during puriftcatiou [8,11-13,23]. The apparent molecular size of the unproteolyzed cGI.PDE molecule, 105-kDa, was smaller than the 135-kDa found for the rat adipose tissue enz'p~e [28], but similar ~ that of the human platelet enzyme and bovine cardiac [12,13]. Recent eDNA cloning and sequencing work in our laboratories has suggested the presence of multiple eGI-PDE isoforms. Previous attempts to describe the kinetic properties of vascular smooth muscle eGI-PDE were made with crude fractions [24-26] contaminated with other types of PDE, e.g., the R o 20-1724-sensitive enzyme which was separated from the cGI-PDE only by the CIT-agarose affinity chromatography step. The pure enzyme had enzy-

155 'cGI-FDE in isolated form as well as in fresh smooth muscle exlracts was phosphorylated by the catalytic subunit of cAMP-PrK. This is consistent with findings


with the c a r d i a c , a d i p o s e a n d platelet tissue isoen-



+~-,- "I106.





z y m e s [13,23,2930]. C a t e e h o l a m i n e s a n d insulin have b e e n s h o w n t o induce p h o s p h o r y l a t i o n a n d activation in isolated rat a d i p o c y t e s [28,31]. W h e t h e r a n a l o g o u s m e c h a n i s m s for h o r m o n a l c o n t r o l f o r c G I - P D E exist in v a s c u l a r s m o o t h muscle o r n o t is p r e s e n t l y b e i n g examined in isolated myocytes in o u r laboratories. A¢c¢.~q t o the p u r e e n z y m e a n d k n o w l e d g e o f its m o l ¢ c u ! a r p r o p erties a r e n e c e s s a r y p r e r e q u i s i t e s for this p u ~ .






" Y'





Fig, 4. Phospborylmion of eGI-PDE- A 50000× g aorta smooth muscle sapematant 180 pmol/min per ml cGI-PDE, A, B) or pu~e cGI-PDE (0.~ p.g/ml, C) was used as substrate for cAMP-PrK. Phosphotylation reactions were carried out for 2 h at 30"C with 20 mM Hepes (pH ZS)0 2 mM dilhiothreiml, 5 mM MgCI2, 170 nM eAMP-I~K catalytic subunit and 0.04 mM D.~PIATP (10000 cpm/pmol), W i t h the pure eGI-PDE reactions were stopl~d by adding SDS-PAGE sample heifer. Pltosphorylated cGI-PDE in aorta sepernalants was [mmunoisolaled as described in Materials and Methods, before SDS-PAGE and autoradioipapby of dried gels (A). In panel B, the platelet antl.e(3l-PDE was preincubated (btocked) with pure human platelet cGI-PDE, before addition to tbc supernatures. The catatytic subunit of the cAMP-PrK was aulophospho~ lated in the incubation as indicated in C (eqaimolar amounts of eAMP-PrK and pu~ cGI-PDE were used).

mological properties similar to those ~ported for the rat and bovine adipose tissue [8,11], bovine heart [13], bovine/human plat¢let [12,23] and rat liver enzymes [14,15] as illustrated in Table If. From the present results it is clear that the cGI-PDE from bovine aortic smooth muscle accounts for about half of the high-affinity cAMP-PDE activity in fresh tissue extracts and that it is a sensitive target for OPC-3911, milrinon¢ and CI-930. Functional studies with these inhibitors have been attempted but bovine aortic strips did not contract (or relax) in a tissue bath in our hands. However, in a parallel study [7] they were found to 13¢ potent relaxants of contracted rat aorta rings, in contrast to tolipram (or Ro 20-1724) which had negligible effects. The inhibition of cOI-PDE and the vasorelaxant effects r o u g h l y c o r r e l a t e d [10| w h i c h s u g g e s t e d that the inhibition of the cGI-PDE might have boen of primary importance for vasorelaxation of these inhibitors, at least in this system.

Mrs Zuz~na Wainicck and Ann-Kristin P~ilbrinkgave valuable technical assistance. We wish to thank Dr. J. Beavo for providing the monoclonal amibody directed against the bovine heart cGI-PDE and Dr. Elisabeth Meacci prepared the anti-human platelet cGI-PDE IgG. This work was suppotled by grants from the following sources: Diabetes Research Education Foun-

daticm, Edgewater, NJ; A_ P~lsson, Maim6; Nordisk Insulin, C o p e n h a g e n ; Swedish D i a b e t e s A s s o c i a t i o n ,

Swedish Hocehst, Stockholm, the Medical Faculty, University of Lurid and the Swedish Medical Research Council (project No. 3362), Aria Rasc6n is a visiting scientist from the Universidad Central de Vcnczu¢la. References l Caddy,RJ.. Muller. P.B, Kubo, $2t., Ru', H. and Leonard, D. (1986) Circulation 73, 124-129. 2 Ludm©r, P.L.. W~ght, R.F., An~k), M.O., Gent, P., Braumvakl, E. and Coluoci, W.S. (t9~) Cicculalion 73, 130-137. 3 Rnssing, T.H. and Drasen, J.M. (1986) J. PharmacoL F_.lp.Thor. 238, 874-879. 4 Kauffman, R.F., Scbeack, ICW., Une~ack. B.G, C'zow¢. V.G. and Cohen, M.I~ (lq~/) J. Phasmacol. Exp. Thef. 242, 864~872. 5 Silver, P.L, Lepme, R.E., O'Connor, EL, Letup, B.M., Hem¢l, I T . , Bentley, R.G. and Han"i~ A.L (1~8) J. Phmmac~. Exp. Ther. 247, 34-42. 6 Halris, A.L, Grant, A.M~ Silver, PJ., Evans, D.B. end Aloud, A.A. {1989) J. Caediovas~ IPhamtacol. 13, 238-244. 7 IAndgren, S., Aadersson, K.E., Belfrage, P., Dqlcrraan. E. and Manganiello, V.C (1989) Plmrmacol. TmicoL 64, 440-445. 8 Degemmn, E, Belfrage, P., Newa~n, A.H., Rice, K.C and Maaganicllo. V.C (lg6"/) J. BioL Chum. 262, 5797-5807. 9 Manganlcllo, V.C. Smith, CJ~ Degenvtan, F_ and ]klfra~. F 0990) in Cyclic Nucleolide Phosphodlesterases:Slmclure, Regulatioa and Drug Action (Beavo, J, and Houslay, M.D, ©d¢), pp. 87-116, John Wiley A-Soil& New York. tO Lindssen, S.. Rasc~, ~ , Anderssm, ICE., Mangaeielk~ V. and Degerman, E. (1991) Biochem. Pharn'm¢oL 42, 545-$b"L tl DoBerman, F_, Mmqpmiello, V.C.. Newman, A.H,, ~ice, ICE. and I~lf~ge, P. (1988)Second Mess. Phospho1~rot` 12, 171-182. I2 Grant, P.G. and ~ a n , FLW. (1984) Bi~¢hemisi~ 23. 18011807.

156 13 Harrison, S.A., Reifsnyde~, D.H., GalEs, B., Cadd, G.G. and IkavO, J,A, (1986) Mo|. pP,armacol. 29, 506-514. 14 ]~e+ NJ., Cooper, M.E. and Housley, M.D. (1~87) Biochem. J. 242, 33-42. 15 l ~ e s , f~ and Loten. E.G+(19,a8) Ear. J. Biochern. 174, 303-309. 16 Davis, B. (19641Ann_ N.Y. Aead. Sci. 121,404-407. ]7 Laemmli, U.K. (19701Nature (Load.) 227, 680-685. 18 Towbin, H, Staehelm, T. and Gordon, J. (19791Proc. Nat]./wad. Sei. USA 76, 4350--4354. 19 Don]itde. M.H,, Ma~tin. D.C. Davies, R.C., Reut~en, M~A. and E[o,,~on, J. (1991]. Anal. Biochem. 195, 364-368. 20 Bradford, M.M. (1~Y761Anal BiB:hem. 72, 248-2~4. 21 S1oschek, C.M. (19871Areal. Biochem. 160, 301-305. 22 Lowt,y, O.H., Rosebt'ough, N.J'., Farr, A.L. and Randall, R.J. (19511 J. Bin|. Chem. 193, 265-275. 23 MacPhee, C.H., Hardum, S.A, and Beavo, J.A. (3986) Proc. Natl. Acad+ ~ . USA 8~+6660-6663. 24 WeiEhaar, B.E~, Barrows, .[.D,~ Kol~,larz, D.C~ Quade, M,M. and Evans, r).B. (1986) Bi0chem. PhLrmacol. 35,787-800.

25 Utlgliier, C., ~;choeffzer, P., LeBec, A., Stroulhou, F- and Stoclet, .I.C. (19861 Biofhcm. Pharmacol, 35, 1743-175L 26 Silver, P,L,, Hemal, L.T., Perrone, M,H., Bentley, R,G,, Busllover, C.R. and Evans, D.B. (19881Eur. J. Pha,"ma¢ol. 150, 85-94. 27 Tanaka, T,, lshik~va, T., Hasiwara, M., Oaoda, IC, ]toh, H. and Hidaka, H. ([988) P]~arntacology36, 313-320. 28 Def,erman, E., Smith, C.J., Tornqvist, H., Vesta, V., Belfr~,g¢, P, and ManBaniel~o, V. (1990) Proc. Ned, .Acad. Sci, USA 87, 5.'33-537. 29 Macphee, CH,, Relfs~wder, D.H,, Moore, T.A., Letea, K.M. and Beavo, J.A. (19881J. Biol. Chem. 263, 103:53-10358. 30 Gram, P.G,, Manaarino. A.F, and Caiman, R.W. (19881 l~ro¢. Natl. Aead. Sci. USA 85, 9071-9075. 31 Smith, CJ., Vesta. V., D©Bewman,E, Belfrage, P. and Manianlello, V.C (1991},L Bi0]. Chem, 266, 13385-13390 32 RabJ]loud, T., Carpealier, G. and Tarmux, P. (lgSB) Eleclrophoresis 9, 288--291. 33 Wiser, MF. and Seh~.eiga, H.G. (19861 Anal. Biochem. 155, 71-77.

Purification and properties of the cGMP-inhibited cAMP phosphodiesterase from bovine aortic smooth muscle.

Pure cGMP-inhibited cAMP phosphodiesterase (cGI-PDE) in micrograms quantities was isolated from bovine aortic smooth muscle after more than 5000-fold ...
522KB Sizes 0 Downloads 0 Views