Báo cáo khoa học: Characterization of inhibitors of phosphodiesterase 1C on a human cellular system potx

13 426 0
  • Loading ...
    Loading ...
    Loading ...

Tài liệu hạn chế xem trước, để xem đầy đủ mời bạn chọn Tải xuống

Tài liệu liên quan

Thông tin tài liệu

Ngày đăng: 07/03/2014, 05:20

Characterization of inhibitors of phosphodiesterase 1C ona human cellular systemTorsten R. Dunkern and Armin HatzelmannBiochemistry 2 Inflammation, ALTANA Pharma AG, Member of the Nycomed Group, Konstanz, GermanyKeywordsglioblastoma cell line; 8-methoxymethyl-3-isoabutyl-1-methylxanthine;phosphodiesterase 1C; SCH51866;vinpocetineCorrespondenceT. R. Dunkern, Department of BiochemistryInflammation RPR ⁄ B2, ALTANA PharmaAG, Member of the Nycomed Group,Byk-Gulden-Str. 2, 78467 Konstanz,GermanyFax: +49 7531 84 2712Tel: +49 7531 84 3121E-mail: torsten.dunkern@altanapharma.com(Received 25 April 2007, revised 17 July2007, accepted 19 July 2007)doi:10.1111/j.1742-4658.2007.06001.xDifferent inhibitors of the Ca2+⁄ calmodulin-stimulated phosphodiester-ase 1 family have been described and used for the examination of phospho-diesterase 1 in cellular, organ or animal models. However, the inhibitorsdescribed differ in potency and selectivity for the different phosphodiester-ase family enzymes, and in part exhibit additional pharmacodynamicactions. In this study, we demonstrate that phosphodiesterase 1C isexpressed in the human glioblastoma cell line A172 with regard to mRNA,protein and activity level, and that lower activities of phosphodiesterase 2,phosphodiesterase 3, phosphodiesterase 4 and phosphodiesterase 5 are alsopresent. The identity of the phosphodiesterase 1C activity detected was ver-ified by downregulation of the mRNA and protein through human phos-phodiesterase 1C specific small interfering RNA. In addition, the measuredKmvalues (cAMP, 1.7 lm; cGMP, 1.3 lm) are characteristic of phospho-diesterase 1C. We demonstrate that treatment with the Ca2+ionophoreionomycin increases intracellular Ca2+in a concentration-dependent waywithout affecting cell viability. Under conditions of enhanced intracellularCa2+concentration, a rapid increase in cAMP levels caused by the ade-nylyl cyclase activator forskolin was abolished, indicating the involvementof Ca2+-activated phosphodiesterase 1C. The reduction of forskolin-stimu-lated cAMP levels was reversed by phosphodiesterase 1 inhibitors in a con-centration-dependent way. Using this cellular system, we compared thecellular potency of published phosphodiesterase 1 inhibitors, including8-methoxymethyl-3-isobutyl-1-methylxanthine, vinpocetine, SCH51866, andtwo established phosphodiesterase 1 inhibitors developed by Schering-Plough (named compounds 31 and 30). We demonstrate that up to 10 lm8-methoxymethyl-3-isobutyl-1-methylxanthine and vinpocetine had noeffect on the reduction of forskolin-stimulated cAMP levels by ionomycin,whereas the more selective and up to 10 000 times more potent phosphodi-esterase 1 inhibitors SCH51866, compound 31 and compound 30 inhibitedthe ionomycin-induced decline of forskolin-induced cAMP at nanomolarconcentrations. Thus, our data indicate that SCH51866 and compounds 31and 30 are effective phosphodiesterase 1 inhibitors in a cellular context, incontrast to the weakly selective and low-potency phosphodiesterase inhibi-tors 8-methoxymethyl-3-isobutyl-1-methylxanthine and vinpocetine. A172cells therefore represent a suitable system in which to study the cellulareffect of phosphodiesterase 1 inhibitors. 8-Methoxymethyl-3-isobutyl-1-methylxanthine and vinpocetine seem not to be suitable for the study ofphosphodiesterase 1-mediated functions.AbbreviationsIL, interleukin; 8MM-IBMX, 8-methoxymethyl-3-isobutyl-1-methylxanthine; PDE, phosphodiesterase; si, small interfering.4812 FEBS Journal 274 (2007) 4812–4824 ª 2007 The Authors Journal compilation ª 2007 FEBSThe superfamily of cyclic nucleotide-hydrolyzing phos-phodiesterase (PDE) enzymes comprises 11 familymembers (PDE1–PDE11). Within each family, differ-ent genes encode different PDE subtypes, and alterna-tive splicing of the messenger RNAs generates multipleisoforms. The entire PDE superfamily consists ofabout 50 proteins [1]. The historical classification intodifferent PDE families is based on their substrate spec-ificity for cAMP, cGMP or both, and mechanisms ofactivation and inhibitor sensitivity. The PDE familyPDE1, which consists of the subtypes PDE1A, PDE1Band PDE1C, each encoded by one gene, becomes acti-vated upon Ca2+⁄ calmodulin binding. Thus, PDE1enzymes enable signaling crosstalk between the twoimportant cellular second messengers Ca2+and cyclicnucleotides, translating changes in Ca2+concentra-tions into changes in cAMP ⁄ cGMP concentrations.PDE1A is highly expressed in the brain as well as inspermatozoa. PDE1B1 mRNA is present in neuronalcells of the cerebellum, caudate, and hippocampus,and in Purkinje cells [2]. In addition, PDE1B1 wasfound in the heart and skeletal muscles [3]. PDE1C,for which the three splicing variants PDE1C1,PDE1C2 and PDE1C3 have been described [4], isexpressed in human arterial smooth muscle cells [5],nociceptive sensory neurons [6], the myocardium [7],and beta TC3 insulinoma cells [8].PDE1 enzymes have been reported to be involved inheterogeneous functions such as smooth muscle cellproliferation (PDE1C [5]), Cl–secretion by humancolonic epithelial cells [9], interleukin (IL)-13 produc-tion by CD3 ⁄ CD28-stimulated human T-lymphocytes[10], the acrosome reaction of human spermatozoa[11], lipopolysaccharide-induced IL-6 release from alve-olar epithelial cells [12], glucose-induced insulin secre-tion from insulinoma cells [8], contraction of ureteralsmooth muscle cells [13], hypoxic pulmonary remodel-ing [14], granulocyte–macrophage colony-stimulatingfactor-induced human monocyte-to-macrophage differ-entiation (PDE1B [15]), and gall bladder relaxation[16].Until now, knowledge on the function of PDE1 indifferent cell types, organs or even animal models hascome from the experimental use of more or less weakinhibitors with relatively low PDE1 selectivity, such asvinpocetine and 8-methoxymethyl-3-isobutyl-1-methyl-xanthine (8MM-IBMX) [with the exception of somePDE1-targeting small interfering (si)RNA approaches][17]. Unfortunately, changes in cyclic nucleotide con-centrations, which are indicative of PDE-inhibitoryaction, were often not investigated in such studies.This gives rise to the question of the validity of inter-pretations based on results obtained with the use ofthese compounds, especially as other modes of actionbesides PDE inhibition have been documented for thenootropic drug vinpocetine as well as for 8MM-IBMX,such as modulation of ion currents [18–20] and adeno-sine receptor inhibition [21], respectively.Because of the lack of a cellular model that allowsseparate monitoring of PDE1 activity, no comparisonof the cellular potency of PDE1 inhibitors described inthe literature has been performed up to now. This isimportant, because the cellular efficiency of PDEinhibitors does not only depend on their potency. It isalso affected by their physicochemical properties, suchas lipophilicity, which influences membrane permeabil-ity, and the subcellular localization. Furthermore, cel-lular potency is influenced by the concentrations of thesubstrates in the subcellular local environment, due tocompetition with the PDE inhibitors.Thus, we developed a cellular model using PDE1C-expressing A172 glioblastoma cells that allows us toquantitatively monitor the effects of the PDE1-inhibit-ing compounds 8MM-IBMX, vinpocetine, SCH51866,compound 30, and compound 31 (Fig. 1) on PDE1Cactivity under conditions of enhanced intracellularCa2+concentration.ResultsHuman glioblastoma A172 cells expressCa2+/calmodulin-activated PDE1CAs shown in Fig. 2, we identified high Ca2+⁄ calmodu-lin-induced cAMP ⁄ cGMP PDE activity in the cellularextract of A172 cells; this therefore must be consideredto be PDE1 enzyme activity. In comparison to thedescribed PDE1 activity of other cell types, such aseosinophils, macrophages, CD4+⁄ CD8+T-lympho-cytes and A549 epithelial cells [22], the measuredPDE1 activity in A172 cells was up to 10 000-foldhigher. The similar PDE activity with regard to therate of cAMP and cGMP hydrolysis was a first indica-tion that this might be PDE1C activity. This enzyme iswell known to hydrolyze both cyclic nucleotides withsimilar efficiency, in contrast to PDE1A and PDE1B.Our assumption was confirmed by determining thecorresponding Kmvalues, which were very similarfor both cAMP (1.7 lm) and cGMP (1.3 lm) hydro-lysis (published data for the other PDE1 sub-types: PDE1A2, Km[cAMP ⁄ cGMP] ¼ 112.7 ⁄ 5.0 lm;PDE1B1, Km[cAMP ⁄ cGMP] ¼ 24.3 ⁄ 2.7 lm) and werealso similar to the published data for PDE1C [4]. Inaddition, lower activities of PDE5 and PDE4 andminor activities for PDE2 and PDE3 were measured.In accordance with the data for PDE activity, weT. R. Dunkern and A. Hatzelmann Characterization of inhibitors of PDE1CFEBS Journal 274 (2007) 4812–4824 ª 2007 The Authors Journal compilation ª 2007 FEBS 4813confirmed the expression of PDE1C protein in A172cells by western blotting (Fig. 3A), in which recombi-nant PDE1C protein served as a molecular mass con-trol. More precisely, we detected two protein bandsthat probably reflected the PDE1C splice variants ⁄ iso-forms PDE1C1 and presumably PDE1C3 (based onthe determined molecular masses of 72 and 75 kDa).The polyclonal antibody used has been reported todetect different isoforms of PDE1C. We confirmedthat this antibody shows only negligible crossreactivitywith PDE1A and PDE1B (data not shown). To verifythe identity of the two observed protein bands andthus the specificity of the antibody used, we performedtransfections with PDE1C (isotype-unspecific) siRNAsmart pools. As shown in Fig. 3B, 48 h after transfec-tion of A172 cells with 10 nm PDE1C-targeted siRNA,PDE1C mRNA expression was strongly reduced byabout a factor of 5. Consequently, 72 h after siRNAtransfection, expression of the two protein bands wasalso reduced by about 60% in comparison to negativecontrol siRNA-transfected cells (Fig. 3C), which ulti-mately proved them to be identical to PDE1C. Usingthe same PDE1C siRNA smart pool, we were also ableto reduce PDE1C mRNA and protein expression inA549 cells, which further demonstrates the specificityof the siRNA (data not shown).Ionomycin treatment increases intracellular Ca2+concentrations and activates PDE1C in A172 cellsTo monitor Ca2+-mediated PDE1C activation, wetreated A172 cells with different concentrations of iono-mycin. As shown in Fig. 4A, ionomycin (2–10 lm)increased intracellular Ca2+concentration over 30 minin a dose-dependent manner.It was important to assess the concentrations of iono-mycin that cells could tolerate over a 30 min period.Cells were therefore preincubated for 30 min with arange of ionomycin concentrations up to 10 lm, andthen with 10 lm ionomycin plus 5 mm Ca2+to elicit amaximum Ca2+response (indicated by a dashed arrowin Fig. 4A). This response no longer occurred when theionomycin concentration in the 30 min preincubationNNNNNNOOOHHVinpocetin8MM-IBMXNNNNNHOFFFSCH51866NNNNNONNNNNO3130OOFig. 1. Molecular structures of the PDE1inhibitors vinpocetine, 8MM-IBMX,SCH51886, and compounds 30 and 31.PDE1 [cAMP]PDE1 [cGMP]PDE2 [cGMP]PDE3 [cAMP]PDE4 [cAMP]PDE5 [cGMP]010000200003000040000PDE activitypmol x min-1 x 108 cells-1Fig. 2. Expression of family-specific PDE activities in A172 cells.Cells were analyzed for PDE1 (cAMP and cGMP), PDE2, PDE3,PDE4 and PDE5 hydrolysis activity. The mean values and standarddeviations of 10 independent experiments are shown. The sub-strates (cAMP ⁄ cGMP) used to determine each PDE activity areshown in parentheses.Characterization of inhibitors of PDE1C T. R. Dunkern and A. Hatzelmann4814 FEBS Journal 274 (2007) 4812–4824 ª 2007 The Authors Journal compilation ª 2007 FEBSphase was higher than 6 lm, which suggests a cyto-toxic effect of ionomycin at these higher concentra-tions. The addition of 50 mm EGTA (as a control) atthe end of the experiment (indicated by a solid arrowin Fig. 4A) decreased the Ca2+signal. On the basis ofthese results, 6 lm ionomycin (in the presence of 5 mmCa2+in the medium, 6 min incubation time) was usedin further experiments to rapidly increase intracellularCa2+concentrations (Fig. 4B). Under these condi-tions, the elicited Ca2+signal measured at 6 min wasabout 40% of the maximal possible Ca2+signal andthus not in the saturation phase. To further verify thatthese treatment conditions did not affect cell viability,cytotoxicity assays were performed. In contrast to vari-ous concentrations of Triton X-100, which served aspositive controls, treatment with 6 lm ionomycin for15 min did not affect the viability of A172 cells(Fig. 4C).As shown in Fig. 5, treatment of A172 cells with theadenylyl cyclase activator forskolin (10 lm) for 1, 5and 20 min time-dependently increased intracellularcAMP concentrations up to three-fold in comparisonto mock-treated controls. However, pretreatment ofthe cells with ionomycin for 5 min to increase intracel-lular Ca2+concentrations abolished or attenuated theforskolin-induced increase in cAMP, which we consid-ered to be due to the activation of PDE1C in thesecells under these conditions.Inhibitors of PDE1 block the ionomycin-induceddecrease in forskolin-induced cAMPconcentrationsVarious inhibitors of PDE1 are described in the litera-ture, including vinpocetine, 8MM-IBMX, SCH51866,and the Schering-Plough compounds 31 and 30 [23].PDE1CLipidERK2ControlPDE1CsiRNALipidNegative control siRNAPDE1C siRNA0500010000PDE1C protein expression(arbitrary densitometric units)PDE1CABC2010301 2.5A172 cellsSf21/PDE1C163.2 kDa82.2 kDa011.20.80.60.40.2Negative control siRNALipidPDE1C siRNAPDE1C mRNA(relative expression) µgFig. 3. PDE1C protein and mRNA areexpressed in A172 gliobastoma cells. (A) 10,20 and 30 lg of A172 cell extracts wereanalyzed for PDE1C protein expression bywestern blotting. Cell lysate from Sf21insect cells (1 and 2.5 lg) transfected withhuman PDE1C1 was used as a control.Molecular mass markers are shown to theright of the blot. (B, C) PDE1C siRNA (smartpool; 10 nM) ⁄ negative control siRNA wastransfected into A172 cells. Forty-eighthours later, the cells were harvested formRNA generation and subsequent cDNAsynthesis. The expression of PDE1C mRNAwas analyzed by real-time PCR in relation tolipid-treated cells (B). The mean values andstandard deviations of three experimentsare shown. Otherwise, cells were harvested72 h after transfection, and analyzed forPDE1C protein expression by western blot-ting and densitometric analysis (C). To con-firm equal protein loading, the blot wasadditionally incubated with an antibody toERK2.T. R. Dunkern and A. Hatzelmann Characterization of inhibitors of PDE1CFEBS Journal 274 (2007) 4812–4824 ª 2007 The Authors Journal compilation ª 2007 FEBS 4815We investigated these compounds with regard topotency against PDE1A3, PDE1B1 and PDE1C1 andoverall PDE selectivity (Table 1). The rank order ofpotency of these compounds against PDE1C1 iscompound 31 > compound 30 > SCH51866 > 8MM-IBMX > vinpocetine. Indeed, the frequently usedPDE1 inhibitors vinpocetine and 8MM-IBMX have lowpotency (IC50values 100 lm and 10 lm, respectively)and selectivity vs. other PDEs. The IC50values that wehave determined for inhibition of the PDE1 enzymes bythese compounds agree with those in the literature[23–26].Next, we analyzed the effect of these inhibitors onthe ionomycin-induced decrease in forskolin-stimulatedcAMP concentrations in A172 cells. Figure 6A showsthat SCH51866 reversed the ionomycin-induceddecrease in cAMP concentrations in a concentration-dependent manner. Even 10 nm of this PDE1 inhibitorproved to be significantly effective. At this concentra-tion, no other PDE expressed in A172 cells other thanPDE1C was inhibited by this compound (Table 1),and thus the observed effect must be attributed exclu-sively to the inhibition of PDE1C. In contrast, withoutaddition of ionomycin and thus with basal intracellularCa2+concentrations, SCH51866 (up to 100 nm) didnot potentiate the increase in cAMP induced by fors-kolin (Fig. 6B). Compounds 31 and 30 proved to beinhibitors with higher enzymatic potency (Table 1).Accordingly, as shown in Fig. 7A, these compoundsalso proved to be cellularly strongly effective. Thus,compound 30 became significantly effective at 10 nm,and compound 31 at 1 nm, in modulating intracellularcAMP concentrations. The effect of these two com-pounds developed by Schering-Plough proved also tobe concentration dependent. However, at higher con-centrations, these compounds (as also observed forRFU6000110001600021000260003100036000ControlIonomycin51015Time (min)AB0.005%0.010%0.015%0.020%Ionomycin050100Triton X-100Death cells (%)CRFUTime (min)1000020000300004000050000Control2 µM4 µM6 µM8 µM10 µM1530Fig. 4. Ionomycin treatment up to 6 lM con-centration-dependently induces an increasein intracellular Ca2+concentrations in A172cells without inducing cytotoxicity. (A) Fluo3-mix dye-loaded A172 cells were treatedfor 30 min with increasing concentrations(2–10 lM) of ionomycin. The increase inintracellular Ca2+concentrations was ana-lyzed by scanning fluorometry. After thattime period, 5 mM Ca2+and ionomycin (to afinal concentration of 10 lM) were added(dotted arrow) to elicit a maximum Ca2+response to (a) confirm that the cells arestill able to respond to a Ca2+signal and (b)determine the maximal possible Ca2+signalintensity. Thereafter, EGTA (solid arrow)was added; this decreased the signal. Asshown in (B), the addition of ionomycin(straight arrow) and 5 mM Ca2+(dottedarrow) induced a rapid increase in intracellu-lar Ca2+concentrations (the mean valuesand standard deviations of three experi-ments are shown). (C) As measured by lac-tate dehydrogenase assays, treatment ofA172 cells with 6 lM ionomycin plus 5 mMCa2+for 15 min did not induce any cytotox-icity, whereas Triton X-100, which served asa positive control, induced cytotoxicity in aconcentration-dependent manner.Characterization of inhibitors of PDE1C T. R. Dunkern and A. Hatzelmann4816 FEBS Journal 274 (2007) 4812–4824 ª 2007 The Authors Journal compilation ª 2007 FEBSSCH51866) also enhanced the forskolin-inducedincrease in intracellular cAMP concentrations, whichmight not be due to the inhibition of PDE1C. At theseconcentrations of compounds 31 and 30, the cAMP-hydrolyzing PDE2, PDE3 and PDE4, which are alsoexpressed in A172 cells, became inhibited. Thus, theinhibition of these PDE family members might accountfor this effect. Figure 7B demonstrates that the concen-trations of compounds 31 and 30, which are effectivein modulating cAMP in the presence of ionomycin, donot modulate forskolin-induced cAMP concentrationsin the absence of ionomycin.In contrast to SCH51866, and compounds 31 and30, the older, well-described and widely used PDE1inhibitors 8MM-IBMX and vinpocetine from 10 nmup to 10 lm (which are concentrations reported inmany publications to be effective on organs or cells)did not influence the ionomycin-induced decrease incAMP concentrations (Fig. 8A). This is indicative oftheir weak cellular potency, which is of course relatedto a low enzymatic potency (Table 1). In the absenceof ionomycin, they also did not influence forskolin-induced cAMP concentrations (Fig. 8B).DiscussionThe heterogeneous expression of human PDEsthroughout the human body and their diverse func-tions make this enzyme superfamily an attractivetarget for pharmaceutical inhibitor development.Whereas inhibitors of PDE5, PDE4 and PDE3 haveundergone clinical trials in the past, are evenapproved as drugs, and are still under developmentfor various indications such as asthma, pulmonaryhypertension and chronic obstructive pulmonary dis-ease, inhibitors of other PDE subtypes are mostly inpreclinical development phases. This is due to the lim-ited information on the function of these enzymes, alack of target-specific cellular test systems for drugscreening purposes, and their drugability. The cellular1 5 200.000.250.500.751.001.251.50ControlIonomycinForskolinForskolin +IonomycinTime (min)cAMP (pmol/1x106 cells)Fig. 5. Pretreatment of A172 cells with ionomycin dampens a fors-kolin-induced increase in intracellular cAMP concentrations. A172cells pretreated with ionomycin (6 lM) or untreated were stimu-lated for adenylyl cyclase activation with forskolin (10 lM) or vehicle(control). cAMP was measured 1 min, 5 min and 20 min after fors-kolin stimulation in the corresponding cellular extracts by cAMP-ELISA. The result of a representative experiment is shown.Table 1. Inhibition of human PDE family members by vinpocetine, 8MM-IBMX, SCH51866 and compounds 31 and 30. The IC50(log M) val-ues of the different compounds obtained from at least three independent experiments using recombinant human PDEs (PDE1A3, PDE1B1,PDE1C1, PDE2A3, PDE3A1, PDE4D3, PDE4B2, PDE5A1, PDE7A1, PDE9A3, PDE9A5, PDE10A, PDE11A4) or PDEs from human platelets(PDE3, PDE5) are shown. Dashes indicate that the corresponding measurements have not been performed.Vinpocetine 8MM-IBMX SCH51866 Compound 31 Compound 30PDE1A3 ) 4.8 ) 5.6 ) 8.0 ) 9.2 ) 8.8PDE1B1 ) 4.7 ) 4.7 ) 7.0 ) 8.4 ) 8.0PDE1C1 ) 4 ) 5 ) 7.1 ) 9.3 ) 8.6PDE2A3 > ) 4 ) 4.6 ) 6.3 5.7 ) 5.3PDE3 ⁄ 3A1 – ⁄ > ) 4–⁄ > ) 4 ) 4.9 ⁄ ––⁄ ) 5.8 – ⁄ ) 5.8PDE4D3 ⁄ B2 – ⁄ > ) 5–⁄ ) 4.4 ) 5.3 ⁄ ––⁄ ) 6.2 – ⁄ ) 5.6PDE5 ⁄ 5A1 – ⁄ > ) 4–⁄ ) 4.5 ) 6.4 ⁄––⁄ ) 6.4 – ⁄ ) 6.2PDE7A1 > ) 4>) 4>) 4>) 4>) 4PDE9A3 ⁄ A5 – ⁄ > ) 4–⁄ > ) 4 ) 4.9 ⁄ ) 4.7 > ) 4>) 4PDE10A > ) 4 ) 4.6 ) 5.4 ) 4.8 ) 4.6PDE11A4 > ) 4 ) 4.6 ) 4.6 ) 5.5 ) 5.9T. R. Dunkern and A. Hatzelmann Characterization of inhibitors of PDE1CFEBS Journal 274 (2007) 4812–4824 ª 2007 The Authors Journal compilation ª 2007 FEBS 4817functions of the PDE1 gene family have often beenanalyzed by using the well-described PDE inhibitorsvinpocetine and 8MM-IBMX. Unfortunately, vinpoce-tine and 8MM-IBMX possess other pharmacologicactivities in addition to the inhibition of PDEs[18–21], which might interfere with the different find-ings described in the literature, such as lipopolysac-charide-induced IL-6 release from alveolar epithelialcells [12], or gall bladder relaxation [16], and mightthus give rise to false interpretations.In this article, we show that, in comparison to newercompounds such as SCH51866 [24], compound 31 andcompound 30 [23], the ‘classic’ inhibitors vinpocetineand 8MM-IBMX (Fig. 1) have low enzymatic potencyand low PDE subtype specificity. We asked: (a) whatthe effective cellular concentration of these inhibitorsis; and (b) whether the difference in enzymatic potencywould also transfer to differing cellular potency. Thisneed not necessarily be true, as cellular potency alsodepends on the physicochemical properties of inhibi-tors, such as lipid ⁄ liquid distribution and membranepermeability. Thus, we searched for a cellular testingsystem that would allow us to measure a functionaloutcome, which is explicitly due and directly linked tothe inhibition of PDE1.We showed that the human glioblastoma cell lineA172 expresses high levels of Ca2+⁄ calmodulin-stimu-lated PDE1C activity and protein. The identity ofPDE1C was confirmed by Kmvalue determination,demonstrating that the Ca2+⁄ calmodulin-stimulatedPDE1 activity can hydrolyze cAMP and cGMP withnearly equal efficacy, as previously described [4]. Fur-thermore, transfection of A172 cells with PDE1CsiRNA reduced the expression of the protein (asshown by western blotting). Interestingly, it took 72 hafter siRNA transfection for the PDE1C protein levelto decrease, whereas mRNA was reduced earlier, sug-gesting a long half-life of this protein in A172 cells. InControlForskolinForskolin + IonomycinSCH 51866 -8 log MSCH 51866 -7 log MSCH 51866 -6 log MSCH 51866 -5 log M012345Forskolin + Ionomycin515*********cAMP (fold of control)ControlForskolinSCH 51866 -8 log MSCH 51866 -7 log MSCH 51866 -6 log M051015BAForskolincAMP (fold of control)****Fig. 6. The PDE1 inhibitor SCH51866 concentration-dependentlyreverses the ionomycin effect on the forskolin-induced increase inintracellular cAMP concentrations. (A) A172 cells pretreated withionomycin (6 lM) and SCH51866 () 8to) 5 log M final concentra-tion) or untreated were stimulated for adenylyl cyclase activationwith forskolin (10 lM) or vehicle (control). cAMP was measured1 min after forskolin stimulation in the corresponding cellularextracts by cAMP-ELISA. The mean values and standard deviationsof three independent experiments are shown. A statistical analysis(Student’s paired t-test) was performed to compare the forskolinplus ionomycin-treated group with those treated in addition withSCH51866. (B) A172 cells were treated only with SCH51866 () 8to ) 6 logM final concentration) or vehicle (control), and stimulatedfor adenylyl cyclase activation with forskolin (10 lM); cAMP wasthen measured. A statistical analysis (Student’s paired t-test) wasused to compare the forskolin-treated group with control cells.*P<0.05, **P<0.01, ***P<0.005.Characterization of inhibitors of PDE1C T. R. Dunkern and A. Hatzelmann4818 FEBS Journal 274 (2007) 4812–4824 ª 2007 The Authors Journal compilation ª 2007 FEBSaddition to PDE1C activity in A172 cells, we identifiedPDE4 activity and a minor amount of PDE3 activityin A172 cells (both capable of hydrolyzing cAMP to5¢-AMP), as well as cGMP-hydrolyzing PDE5.In early experiments performed in our laboratories,A172 cells were incubated with PDE1 inhibitors withor without forskolin (to stimulate the adenylyl cyclase),and intracellular cAMP was subsequently measured.However, we did not observe any effects of the PDE1inhibitors on intracellular cyclic nucleotide concentra-tions within the PDE1-selective concentration range(data not shown). Thus, we speculated that under thegiven basal cell culture conditions, PDE1C might notbe active in these cells. To circumvent this, we treatedA172 cells with ionoymcin, which concentration-depen-dently increased intracellular Ca2+concentrations byCa2+store depletion and extracellular Ca2+influxwithin a few minutes (as measured by Ca2+imagingscanning fluorometry), which in theory should lead toan activation of PDE1C. We determined an optimalionomycin concentration for our experiments, whichinduced no cytotoxicity and increased Ca2+concentra-tions by about 40% of the maximal inducible Ca2+response in the time frame used for subsequent investi-gations.Cells pretreated with ionomycin in such a wayresponded to forskolin with an attenuated or evenabolished increase in intracellular cAMP concentra-tions. This demonstrates that the ionomycin-inducedincrease in intracellular Ca2+concentration leads toactivation of the PDE1C enzyme, which then hydro-lyzes cAMP generated by the forskolin-stimulated ade-nylyl cyclase. Activation of PDE1C was seen 1 minafter ionomycin treatment of the cells. This cellularsystem proved to be highly useful for the cellularpotency screening of PDE1 inhibitors. Pretreatment ofA172 cells with the three recently described PDE1inhibitors SCH51866 [24], compound 31 and com-pound 30 [23] reversed the attenuating effect ofionomycin on forskolin-induced cAMP concentrations.That means that, in comparison to forskolin plusionomycin-treated cells, the cAMP concentrationsincreased. Interestingly, these compounds becameeffective even at 1–10 nm, which is close to the IC50(measured at a competitive substrate concentration of0.5 lm cAMP ⁄ cGMP) of these compounds for theControlForskolin30 -10 log M30 -9 log M30 -8 log M 31 -10 log M31 -9 log M31 -8 log M01234BAForskolin*************cAMP (fold of control)******ControlForskolinForskolin + Ionomycin31 -10 log M31 -9 log M31 -8 log M31 -7 log M31 -6 log M31 -5 log M30 -10 log M30 -9 log M30 -8 log M30 -7 log M30 -6 log M30 -5 log M012345678Forskolin + Ionomycin**************cAMP (fold of control)Fig. 7. The PDE1 inhibitors compound 31 and compound 30 con-centration-dependently reverse the ionomycin effect on the forsko-lin-induced increase in intracellular cAMP concentrations, starting at1nM. (A) A172 cells pretreated with ionomycin (6 lM) and com-pounds 31 and 30 () 10 to ) 5 logM final concentration) oruntreated were stimulated for adenylyl cyclase activation with fors-kolin (10 lM) or vehicle (control). cAMP was measured 1 min afterforskolin stimulation in the corresponding cellular extracts bycAMP-ELISA. A statistical analysis (Student’s paired t-test) wasused to compare the forskolin plus ionomycin-treated group withthose treated in addition with compounds 31 and 30. (B) Cells weretreated only with compounds 31 and 30 () 10 to ) 8 logM finalconcentration) or vehicle (control), and stimulated for adenylylcyclase activation with forskolin (10 lM); cAMP was then mea-sured. A statistical analysis (Student’s paired t-test) was used tocompare the forskolin-treated group with control cells. *P<0.05,**P<0.01, ***P<0.005.T. R. Dunkern and A. Hatzelmann Characterization of inhibitors of PDE1CFEBS Journal 274 (2007) 4812–4824 ª 2007 The Authors Journal compilation ª 2007 FEBS 4819human PDE1C enzyme. Within this concentrationrange, or even higher, up to 100 nm, these three com-pounds do not inhibit any of the PDEs expressed inA172 cells except for PDE1C. Thus we can be sure toobserve effects of these inhibitors on cAMP concentra-tions controlled by PDE1C. At higher, nonselectiveconcentrations, SCH51866 (at 1–10 lm), compound 31or compound 30 (at 100 nm to 10 lm) caused a greaterincrease in cAMP concentrations (in the presence ofionomycin and forskolin) than that induced by forsko-lin. This increase must be attributed to the inhibitionof other PDEs in addition to PDE1C at these concen-trations by these compounds.As expected on the basis of their low inhibitorypotency against PDE1C, vinpocetine and 8MM-IBMXexhibited lower cellular potency. Even up to 10 lm,noeffects of these compounds were observed. This is quitesurprising, because there are several publications dem-onstrating effects of these drugs in this concentrationrange. For example, Haddad et al. [12] demonstratedon alveolar epithelial cells that 8MM-IBMX inhibitslipopolysaccharide-mediated IL-6 biosynthesis with anIC50of 7.08 lm. Vinpocetine has been shown tostrengthen the contraction of smooth muscle cells andshorten action potentials at 1 lm [13], and inhibit plate-let-derived growth factor-induced proliferation of pul-monary artery smooth muscle cells at 4 lm [14]. Even at30 nm, vinpocetine was shown to be effective in relaxingcarbachol-constricted gall bladders [16]. When consider-ing these contradictory data, we should keep in mindthat the nootropic drug vinpocetine exhibits additionalpharmacologic actions, which might have affected thefindings in the publications cited above. For example,vinpocetine has been shown to modulate different typesof K+current [19] and to block the NaV1.8 tetrodo-toxin-resistant Na+channels of rats [20].ConclusionsOur data demonstrate the suitability of the A172 cellsystem for determination of the PDE1C-specific effectof inhibitors on cAMP levels. They further suggestcaution in the interpretation of data arising from theexperimental use of 8MM-IBMX and vinpocetine,because of their low potency and selectivity. Thus,newer compounds, such as SCH51866, compound 31ControlForskolinForskolin + Ionomycin8MM-IBMX -8 log M8MM-IBMX -7 log M8MM-IBMX -6 log M8MM-IBMX -5 log MVinpocetin -8 log MVinpocetin -7 log MVinpocetin -6 log MVinpocetin -5 log M0123Forskolin + IonomycincAMP (fold of control)ControlForskolin8MM-IBMX -8 log M8MM-IBMX -7 log M8MM-IBMX -6 log M8MM-IBMX -5 log MVinpocetin -8 log MVinpocetin -7 log MVinpocetin -6 log MVinpocetin -5 log M0246810BAForskolincAMP (fold of control)Fig. 8. The low selective PDE1 inhibitors vinpocetine and 8MM-IBMX do not reverse the ionomycin effect on the forskolin-inducedincrease in intracellular cAMP up to concentrations of 10 lM. (A)A172 cells pretreated with ionomycin (6 lM) and vinpocetine ⁄ 8MM-IBMX () 8to) 5 logM final concentration) or untreated werestimulated for adenylyl cyclase activation with forskolin (10 lM)orvehicle (control). cAMP was measured 1 min after forskolin stimula-tion in the corresponding cellular extracts by cAMP-ELISA. (B) Forcomparison, A172 cells pretreated with vinpocetine ⁄ 8MM-IBMX() 8to) 5 logM) were stimulated with forskolin (10 lM) only, andcAMP was measured after 1 min. The mean values (as fold ofcontrol) and standard deviations of three independent experimentsare shown.Characterization of inhibitors of PDE1C T. R. Dunkern and A. Hatzelmann4820 FEBS Journal 274 (2007) 4812–4824 ª 2007 The Authors Journal compilation ª 2007 FEBSand compound 30, might be more useful as PDE1inhibitors, due to their enhanced potency and selec-tivity.Experimental proceduresMaterialsThe PDE inhibitors vinpocetine and 8MM-IBMX werepurchased from Sigma-Aldrich GmbH (Taufkirchen,Germany). The PDE1-selective inhibitor SCH51866, com-pound 31 and compound 30 (Schering-Plough, Kenilworth,NJ, USA) [23], the selective PDE4 inhibitor piclamilast(RP73401) [27] and the PDE5 inhibitor sildenafil were pre-pared at the chemical facilities of ALTANA Pharma AG(Konstanz, Germany). The selective PDE3 inhibitor mota-pizone [28] was a generous gift from Sanofi-Aventis(formerly Rhone-Poulenc Rorer, Ko¨ln, Germany).The polyclonal rabbit antibody to PDE1C is a product ofFabgennix (Taufkirchen, Germany), and has been reportedto be not cross-reactive with PDE1A, PDE1B or other PDEfamily members. The polyclonal rabbit antibody to ERK2was purchased from Santa Cruz Biotechnology Inc. (SantaCruz, CA, USA). Peroxidase-coupled secondary antibodiesused for western blotting are a product of Jackson Immuno-Research Laboratories Inc. (West Grove, PA, USA).Cell cultureThe human glioblastoma cell line A172 was cultured inDMEM plus 4500 mgÆL)1glucose (Gibco, Invitrogen LifeTechnologies, Grand Island, NY, USA), 2 mml-glutamine,1mm sodium pyruvate and 10% heat-inactivated fetalbovine serum at 37 °C and 5% CO2.Measurements of PDE isoenzyme activities andpreparation of cellular extractsWashed cells were sonicated in homogenization buffer(137 mm NaCl, 2.7 mm KCl, 8.1 mm Na2HPO4, 1.5 mmKH2PO4,10mm Hepes, 1 mm EGTA, 1 mm MgCl2,1mmmercaptoethanol, 5 lm pepstatin A, 10 lm leupeptin,50 lm phenylmethanesulfonyl fluoride, 10 lm soybeantrypsin inhibitor, 2 mm benzamidine, pH 8.2). The lysateswere used for PDE activity measurements as described byThompson & Appleman [29]. The assay mixture contained:30 mm Tris ⁄ HCl (pH 7.4), 5 m m MgCl2, 0.5 lmcAMP ⁄ cGMP, including [3H]cAMP or [3H]cGMP, 100 lmEGTA, PDE isoenzyme-specific activators and inhibitors,and lysates. Incubations were performed for 30 min at37 °C, and terminated by adding 0.2 m HCl. Assays wereincubated on ice for 10 min, and then 25 lgof5¢-nucleo-tidase (Crotalus atrox) was added for 10 min at 37 °C.Thererafter, mixtures were loaded onto QAE-Sephadexcolumns and eluted with 30 mm ammonium formate(pH 6.0). The radioactivity in the eluate was counted.Selective inhibitors and activators of PDE isoenzymes wereused to determine activities of PDE families as describedpreviously [30], with modifications. Briefly, PDE4 wascalculated as the difference in PDE activities at 0.5 lmcAMP in the presence and absence of 1 lm piclamilast(RP73401), and PDE3 as the difference between RP73401-inhibited cAMP hydrolysis in the presence and absence of10 lm motapizone. The fraction of cGMP (0.5 lm) hydro-lysis in the presence of 10 lm motapizone that was inhib-ited by 100 nm sildenafil represented PDE5. The increasein cAMP (0.5 lm) hydrolysis (in the presence of 1 lmRP73401 and 10 lm motapizone) induced by 5 lm cGMPrepresented PDE2. PDE1 was defined as the increment ofcAMP hydrolysis (in the presence of 1 lm RP73401 and10 lm motapizone) or cGMP hydrolysis induced by 1 m mCa2+and 100 nm calmodulin.The Kmvalues of the PDE1 activity were determined bymeasuring the specific activity of Ca2+⁄ calmodulin-stimu-lated cAMP or cGMP hydrolysis under different substrateconcentrations and by transferring the data into a linearLineweaver–Burk diagram.PDE assay⁄inhibitor testingInhibition of PDE activity by the inhibitors was measuredon human PDE enzymes in a modified scintillation proxim-ity assay (GE Healthcare UK Ltd, Little Chalfont, Buck-inghamshire, UK). The test volume was 100 lL, andcontained 20 mm Tris buffer (pH 7.4), 0.1 mg BSAÆmL)1,5mm Mg2+, 0.5 lm cAMP ⁄ cGMP (including [3H]cAMPand [3H]cGMP), 1 lL of the respective substance dilutionin dimethylsulfoxide and sufficient PDE to ensure that15–20% of the cAMP was converted. The reaction wasstarted by adding the substrate (cAMP ⁄ cGMP). After incu-bation for 15 min, assays were stopped by adding scintilla-tion proximity assay beads (50 lL, containing 3 mm 8MM-IBMX). After the beads had sedimented, the samples wereanalyzed in commercially available measuring appliances,and the corresponding IC50values of the compounds weredetermined from the concentration–effect curves.siRNA tranfection and real-time PCRFor siRNA tranfection with subsequent RNA preparation,3 · 105A172 cells were seeded into 24-well plates. For trans-fection with subsequent western blot analysis, 1 · 106cellswere seeded into 10 cm dishes. The day after seeding, cellswere transfected with 10 nm pan-PDE1C siRNA smart pool(Ambion Inc., Austin, TX, USA) [pool number M-007643–00, sequences CCAAGGAGATTGAAGAATT (1), GATCATGCACTGAAATTTA (2), GATGAAACCTCTCAAACTG (3), and CATCATCGCTGGACAATGT (4)], usingT. R. Dunkern and A. Hatzelmann Characterization of inhibitors of PDE1CFEBS Journal 274 (2007) 4812–4824 ª 2007 The Authors Journal compilation ª 2007 FEBS 4821[...]... supernatant was transferred to a 96-well plate, followed by 50 lL of a substrate solution (Promega, Madison, WI, USA; CytoTox96 assay), and after 15 min of incubation, an additional 50 lL of a stop solution Thereafter, extinction was measured at a wavelength of 490 nm The numbers of dead cells induced by ionomycin and the different concentrations of Triton X-100 were calculated on the basis of the standard... effects of prostacyclin analogs in hypoxic pulmonary vascular remodeling Am J Physiol Lung Cell Mol Physiol 288, 103–115 Bender AT, Ostenson CL, Wang EH & Beavo JA (2005) Selective up-regulation of PDE1B2 upon monocyte-to-macrophage differentiation Proc Natl Acad Sci USA 102, 497–502 Kaneda T, Watanabe A, Shimizu K, Urakawa N & Nakajyo S (2005) Effects of various selective phosphodiesterase inhibitors on. .. b-mercaptoethanol-supplemented RLT buffer provided with the RNeasyQuiagen RNA-preparation kit Thereafter, RNA isolation, quantification and cDNA synthesis from 1 lg of RNA was performed according to the manufacturer’s standard protocols Successful cDNA synthesis was controlled by gel electrophoresis of an aliquot of the cDNA samples cDNA was used for subsequent real-time Taq-Man PCR analysis for PDE1C (Applied... Sharma S, Jordan R, Souness JE & Webber SE et al (1994) Anti-inflammatory and bronchodilator properties of RP 73401, a novel and selective phosphodiesterase type IV inhibitor Br J Pharmacol 113, 1423–1431 Borbe HO, Hilboll G & Prop G (1986) Inhibition of human platelet aggregation by motapizone via an increase in intracellular cAMP Agents Actions Suppl 20, 249–257 Thompson WJ & Appleman MM (1979) Assay... Then, 6 lm ionomycin (or Ca2+-containing medium in controls) was added, and the samples were incubated for a further 5 min at 37 °C Then, the adenylyl cyclase activator forskolin (final concentration 10 lm) was added (or Ca2+-containing medium in controls) At this time, the final assay volume of 300 lL was reached One minute later, the incubations were stopped by adding 15 lL of 3.3 m HCl After a further... different concentrations of ionomycin were added (in some experiments, additional 5 mm Ca2+ was added at time point 20 s), and measurement was continued for a further 30 min If desired, at the end of some experiments the maximal possible Ca2+ signal was investigated by adding 10 lm ionomycin and 5 mm Ca2+ To reach a subsequent decrease in Ca2+ therafter at 35 min, 50 mm EGTA was added All measurements were... Pharmacol Therapeutics 109, 366–398 2 Polli JW & Kincaid RL (1992) Molecular cloning of DNA encoding a calmodulin-dependent phosphodiesterase enriched in striatum Proc Natl Acad Sci USA 89, 11079–11083 3 Yu J, Wolda SL, Frazier AL, Florio VA, Martins TJ & Snyder PB (1997) Identification and characterisation of a human calmodulin-stimulated phosphodiesterase PDE1B1 Cell Signal 9, 519–529 4 Yan C, Zhao AZ, Bentley... phosphodiesterase subtypes in human colonic epithelial cells J Membr Biol 185, 137–144 Kanda N & Watanabe S (2001) Regulatory roles of adenylate cyclase and cyclic nucleotide phosphodiesterases 1 and 4 in interleukin-13 production by activated human T cells Biochem Pharmacol 62, 495–507 Fisch JD, Behr B & Conti M (1998) Enhancement of motility and acrosome reaction in human spermatozoa: differential... differential activation by type-specific phosphodiesterase inhibitors Hum Reprod 13, 1248–1254 Haddad JJ, Land SC, Tarnow-Mordi WO, Zembala M, Kowalczyk D & Lauterbach R (2002) Immunopharmacological potential of selective phosphodiesterase inhibition I Differential regulation of lipopolysaccharidemediated proinflammatory cytokine (interleukin-6 and tumor necrosis factor-alpha) biosynthesis in alveolar epithelial... (Applied Biosystems, Foster City, CA, USA) The sense and antisense primer sequences were 5¢-TGTGAGT CCATTAATCGATGAAACC-3¢ and 5¢-ACCTGATCG CTTGGCATCTG-3¢, respectively The probe sequence was as follows: (FAM)-AGCTGATGCTATTCAAACTCGA ACGCCTCT-(TAMRA) A1 72 cells (1.5 · 105 cells in 5 mm Ca2+-containing medium) were preincubated for 5 min in tubes in the presence or absence of the various PDE1 inhibitors . Austin, TX, USA) [pool number M-007643–00, sequences CCAAGGAGATTGAAGAATT (1), GATCATGCACTGAAATTTA (2), GATGAAACCTCTCAAACTG (3), and CATCATCGCTGGACAATGT. Characterization of inhibitors of phosphodiesterase 1C on a human cellular system Torsten R. Dunkern and Armin HatzelmannBiochemistry 2 Inflammation, ALTANA
- Xem thêm -

Xem thêm: Báo cáo khoa học: Characterization of inhibitors of phosphodiesterase 1C on a human cellular system potx, Báo cáo khoa học: Characterization of inhibitors of phosphodiesterase 1C on a human cellular system potx, Báo cáo khoa học: Characterization of inhibitors of phosphodiesterase 1C on a human cellular system potx