adrenomedullin induces endothelium-dependent...

Adrenomedullin Induces Endothelium-DependentVasorelaxation via the Phosphatidylinositol

3-Kinase/Akt–Dependent Pathway in Rat AortaHiroaki Nishimatsu, Etsu Suzuki, Daisuke Nagata, Nobuo Moriyama, Hiroshi Satonaka,

Kenneth Walsh, Masataka Sata, Kenji Kangawa, Hisayuki Matsuo, Atsuo Goto,Tadaichi Kitamura, Yasunobu Hirata

Abstract—To study the mechanisms by which adrenomedullin (AM) induces endothelium-dependent vasorelaxation, weexamined whether AM-induced endothelium-dependent vasodilation was mediated by the phosphatidylinositol 3-kinase(PI3K)/Akt-dependent pathway in rat aorta, because it was recently reported that PI3K/Akt was implicated in theactivation of endothelial NO synthase. AM-induced vasorelaxation in thoracic aorta with intact endothelium wasinhibited by pretreatment with PI3K inhibitors to the same level as that in endothelium-denuded aorta. AM elicited Aktphosphorylation in a time- and dose-dependent manner. AM-induced Akt phosphorylation was inhibited by pretreatmentwith a calmodulin-dependent protein kinase inhibitor as well as with PI3K inhibitors. When an adenovirus constructexpressing a dominant-negative Akt mutant (Ad/dnAkt) was injected into abdominal aortas so that the mutant wasexpressed predominantly in the endothelium layer, AM-induced vasodilation was diminished to the same level as thatin endothelium-denuded aortas. Finally, AM-induced cGMP production, which was used as an indicator for NOproduction, was suppressed by PI3K inhibition or by Ad/dnAkt infection into the endothelium. These results suggestedthat AM induced Akt activation in the endothelium via the Ca21/calmodulin-dependent pathway and that this wasimplicated in the production of NO, which in turn induced endothelium-dependent vasodilation in rat aorta.(Circ Res.2001;89:63-70.)

Key Words: adrenomedullinn phosphatidylinositol 3-kinasen Akt n nitric oxide n gene transfer

V ascular endothelial cells (ECs) are critically implicatedin the modulation of vascular tone by producing a

variety of vasoactive substances. Among them, endothelinhas a vasoconstrictive activity, whereas NO and prostacyclinpossess a vasorelaxant activity. The modulation of vasculartone by ECs is known to be perturbed in pathophysiologicalstates such as hypertension, diabetes mellitus, andhyperlipidemia.1

A novel peptide, which increases the level of cAMP inplatelets, was recently isolated from human adrenal medullaand was dubbed adrenomedullin (AM).2 AM is a 52-aminoacid peptide with a potent vasorelaxant activity and natriuret-ic activity.3–5 Although AM was originally isolated fromhuman pheochromocytoma tissue, it is now well known thatAM is also produced by vascular ECs, vascular smoothmuscle cells, and macrophages,6–8 suggesting its role as alocal mediator in normal and/or diseased vessels. Althoughlittle is known of the intracellular signaling pathways that areactivated by AM, it is reported that AM increases intracellular

cAMP and [Ca21]i.9 AM also activates extracellular signal–regulated kinase (ERK), which appears to be involved inAM-induced cell proliferation,10–12 although AM seems toinhibit cell proliferation in some cells,13 depending on celltypes. It was originally postulated that AM-induced vasore-laxation might be mediated by its effect on intracellularcAMP. However, we have recently shown that AM inducedvasorelaxation, at least partly, in an endothelium-dependentmanner and that endothelium-dependent vasorelaxation wasmediated via the NO/cGMP-dependent pathway in rats.14

Furthermore, it has been reported that transgenic mice thatwere designed to express an excess amount of AM in vascularECs showed lower blood pressure than wild-type mice andthat the difference in blood pressure was abolished byadministration ofNG-monomethyl-L-arginine to those mice,15

suggesting the involvement of an NO/cGMP-dependent path-way for the AM-induced vasorelaxation, although the molec-ular mechanisms by which AM activated the NO/cGMP-dependent pathway remained to be elucidated.

Original received January 24, 2001; revision received April 18, 2001; accepted May 1, 2001.From the Department of Urology (H.N., N.M., T.K.) and the Second Department of Internal Medicine (E.S., D.N., H.S., M.S., A.G., Y.H.), Faculty

of Medicine, University of Tokyo, Japan; Division of Cardiovascular Research (K.W.), St. Elizabeth’s Medical Center of Boston, Boston, Mass; andNational Cardiovascular Center Research Institute (K.K., H.M.), Fujishirodai, Japan.

Correspondence to Etsu Suzuki, MD, PhD, The Second Department of Internal Medicine, Faculty of Medicine, University of Tokyo, 7-3-1 Hongo,Bunkyo-ku, Tokyo 113-8655, Japan. E-mail [email protected]

© 2001 American Heart Association, Inc.

Circulation Researchis available at http://www.circresaha.org

63

by guest on May 9, 2018

http://circres.ahajournals.org/D

ownloaded from

by guest on M

ay 9, 2018http://circres.ahajournals.org/

Dow

nloaded from



ownloaded from

by guest on M

ay 9, 2018http://circres.ahajournals.org/

Dow

nloaded from

http://circres.ahajournals.org/




It is well established that Ca21/calmodulin (CaM) is amajor activator of endothelial NO synthase (eNOS). When[Ca21]i is increased, CaM binds to the CaM binding domainof eNOS, which in turn changes the conformation of eNOSby an “allosteric” effect and activates eNOS.16–18 It istherefore possible that AM activates eNOS via the Ca21/CaM-dependent pathway. However, it has been reported that eNOSis activated by direct phosphorylation of serine 1179(Ser1179) by Akt, a downstream target of phosphatidylino-sitol 3-kinase (PI3K). When eNOS is phosphorylated onSer1179, eNOS is activated under a very low concentration ofCa21, which suggests that the affinity of eNOS to CaMincreases by Ser1179 phosphorylation.19,20Thus, it seemed tobe of interest to examine whether AM induced activation ofeNOS via the PI3K/Akt-dependent pathway.

In the present study, we examined the mechanisms bywhich AM induced endothelium-dependent vasorelaxation.We show that AM induces endothelium-dependent vasore-laxation and cGMP production via the PI3K/Akt-dependentpathway. We also show that AM induces phosphorylation ofAkt and that AM-induced Akt phosphorylation depends onthe Ca21/CaM-mediated pathway.

Materials and MethodsReagentsPhosphospecific anti-ERK1/2 antibody and phosphospecific anti-Akt antibody that recognize catalytically active ERK1/2 and Akt,respectively, were obtained from New England BioLabs. Anti-Aktand -ERK1 antibodies were obtained from Santa Cruz Biotechnol-ogy. NG-nitro-L-arginine methyl ester (L-NAME), forskolin, sodiumnitroprusside (SNP), KN-93, A23187, wortmannin, and LY294002(LY) were purchased from Sigma, and PD98059 (PD) was obtainedfrom New England Biolabs.

Ex Vivo ExperimentsEffects of AM on the tension of rat aortic rings were examined aspreviously described.14 In brief, thoracic aortas were excised from12-week-old male Wistar rats, and aortic rings with or withoutendothelium were mounted in organ chambers filled with Krebs-Ringer bicarbonate solution at 37°C with constant bubbling of 95%oxygen/5% carbon dioxide. Isometric tension was recorded with aforce transducer. The endothelium was denuded by gentle rubbingwith a twist of cotton, and this was confirmed by the lack ofacetylcholine-induced vasodilation. The aortic rings were precon-tracted withL-norepinephrine, and the effects of AM on vasodilationwere studied. In some experiments, several reagents such asL-NAME, wortmannin, LY, and PD were added to the chambersbefore precontraction withL-norepinephrine to examine their effectson AM-induced vasorelaxation. To prepare protein extracts, rataortas were placed in tubes containing oxygenated Krebs-Ringerbicarbonate solution at 37°C and incubated with AM with or withoutpretreatment with several reagents such as KN-93, wortmannin, LY,and PD.

Preparation of Protein ExtractsRat aortas were homogenized on ice in a Triton X-100 homogeni-zation buffer (in mmol/L, HEPES [pH 7.5] 50, NaCl 150, EDTA 1,and DTT 1, as well as 2% Triton X-100 and 10% glycerol)containing 1 mmol/L PMSF, 2mg/mL leupeptin, and 2mg/mLaprotinin. After centrifugation for 20 minutes at 4°C, the clearedsupernatant was used for Western blot analysis. Protein concentra-tion was measured according to Bradford’s method (Bio-Rad).

Western Blot AnalysisWestern blot analysis was performed as previously described.21

Antibodies that recognize catalytically active ERK1/2 and Akt wereused at a dilution of 1:500.

In Vivo Gene Transfer to Rat Abdominal AortaConstruction of an adenovirus expressing a dominant-negative Aktmutant (Ad/dnAkt) has been described previously.22 A recombinantadenovirus that expresses green fluorescence protein (Ad/GFP) wasobtained from Qbiogene. Twelve-week-old male Wistar rats fed ona normal diet were anesthetized by intraperitoneal injection ofpentobarbital (35 mg/kg) and placed on a thermoregulated surgicaltable. The abdominal aorta was exposed by dissecting the mesenter-ies from intestines, and the inferior mesenteric artery and lumbarvessels were clamped with plastic forceps. The left femoral arterywas then exposed by a lower-quadrant incision. A polyethylene tube(PE-10) attached to a microliter syringe was introduced into theartery and advanced up to the proximal side of the bifurcation of theaorta into the common iliac artery. The right common iliac arterywas then clamped, and the left one was ligated together with thePE-10 tube. The abdominal aorta was clamped distal to the diver-gence of the renal arteries. After saline was injected in the PE-10tube to confirm that there was no leakage, 150mL of Ad/dnAkt (1010

plaque-forming units [PFU]/mL) or Ad/GFP (1010 PFU/mL) wasinjected into the aorta and the vessel was incubated with the virussuspension for 30 minutes. The aortas were excised 48 hours afterinfection, and the tension of the aortic rings was examined. In someexperiments, the vessel was incubated with AM, and cGMP produc-tion was measured.

Immunohistological Analysis ofAdenovirus-Infected Abdominal AortaRats were anesthetized with pentobarbital 48 hours after the injectionof Ad/GFP into the abdominal aorta. Abdominal aortas were fixed byperfusion with 4% paraformaldehyde in PBS for 5 minutes. Theywere then excised and embedded in OCT (Sankyo-Miles Inc).Sections of 10mm each were prepared using a cryostat at228°C andthen air dried and immersed in PBS. The sections were treated with5% normal donkey serum (Jackson ImmunoResearch Laboratories),followed by overnight incubation at 4°C with anti-GFP antibody ata dilution of 1:200 (3E6 monoclonal antibody, Qbiogene). Afterwashing them with PBS, the sections were incubated for 1 hour with3.75 mg/mL of Cy-conjugated AffiniPure Donkey Anti-Mouse IgG(H1L) (Jackson ImmunoResearch Laboratories) and with To-PRO-3iodide for nuclear staining at a dilution of 1:500 (Molecular Probes).Thereafter, the sections were washed with PBS, mounted withantifade substance, and observed under a confocal laser scanningmicroscope (MRC-1024, Japan Bio-Rad Laboratories).

Measurement of cGMPRat aortas were homogenized in 4% trichloroacetic acid (pH 4.0) onice. After centrifugation, the supernatant was extracted 4 times withwater-saturated ether and then evaporated. The pellets were redis-solved in a buffer included in the ELISA kit. ELISA was performedaccording to the manufacturer’s recommendation (Amersham).

Statistical AnalysesValues are mean6SEM. The statistical analyses were performedusing ANOVA followed by the Student-Neumann-Keul test. Differ-ences with aP value of ,0.05 were considered statisticallysignificant.

ResultsAM-Induced Endothelium-DependentVasorelaxation Is Inhibited by Pretreatment WithPI3K InhibitorsWe first examined the effects of PI3K inhibition on AM-induced endothelium-dependent vasorelaxation using rat tho-

64 Circulation Research July 6, 2001



ownloaded from


racic aorta (Figure 1). AM ranging from 1029 to 1027 mol/Linduced vasorelaxation in aortic rings with endothelium (E/1aortic rings) precontracted with norepinephrine in a dose-dependent manner, whereas AM-induced vasorelaxation wasdiminished in aortic rings when endothelium was denuded(E/2 aortic rings). In fact, AM-induced vasorelaxation wasnegligible in E/2 aortic rings when 1029 mol/L AM wasadministered, whereas 1029 mol/L AM did induce vasodila-tion in E/1 aortic rings. Although AM, at concentrations of1028 and 1027 mol/L, elicited slight vasorelaxation in E/2aortic rings, the magnitude was 14.4% and 32.6%, respec-tively, compared with its effects on E/1 aortic rings at eachconcentration. Forskolin and SNP induced a potent vasodila-tion in E/2 aortic rings (data not shown). Thus, the reductionin AM-induced vasorelaxation observed in E/2 aortic ringsdid not seem to be due to potential injuries to smooth musclelayers when endothelium was denuded. Pretreatment of E/1aortic rings with 20mmol/L LY or 100 nmol/L wortmanninreduced AM-induced vasorelaxation almost to the samelevels as AM-induced vasorelaxation in E/2 aortic rings. Incontrast, pretreatment of E/1 aortic rings with 25mmol/LPD, an inhibitor of mitogen-activated protein kinase kinase(MEK) 1/2, did not inhibit AM-induced vasodilation. AM-induced vasodilation in E/1aortic rings was abolished bypretreatment with L-NAME. Furthermore, pretreatment ofE/1 aortic rings with LY or wortmannin did not suppressforskolin- or SNP-induced vasodilation (data not shown).These results indicated that AM elicited endothelium-dependent vasodilation, at least in part, via the NO-dependentpathway in rat aorta and that AM-induced endothelium-dependent vasodilation was suppressed by PI3K inhibitionbut not by MEK inhibition. The results also suggested that thesuppression of AM-induced endothelium-dependent vasodi-lation by PI3K inhibitors did not seem to be a result of thecytotoxic effects of LY and wortmannin on vascular smooth

muscle cells, because vasodilation induced by cAMP and NOdonor was intact.

AM Induces Akt Phosphorylation in Rat AortaWe next examined whether AM induced phosphorylation ofAkt in rat aorta. AM induced Akt phosphorylation in atime-dependent manner (Figure 2A). The phosphorylation ofAkt peaked'15 minutes after stimulation with 1027 mol/L

Figure 1. AM-induced endothelium-dependent vasorelaxation issuppressed by PI3K inhibition. Rat aortic rings with (v) or with-out (C) endothelium were precontracted with 1026 mol/L norepi-nephrine, and indicated doses of AM were administered toexamine its vasorelaxant effects. In some experiments, aorticrings were pretreated with 20 mmol/L LY (e), 100 nmol/L wort-mannin (▫), 20 mmol/L PD (n), or 1025 mol/L L-NAME (}) beforeprecontraction with norepinephrine, and effects of AM on vaso-dilation were examined. *P,0.01 and #P,0.05 vs aortic ringswith intact endothelium at 1027 mol/L AM stimulation (n54).

Figure 2. AM induces Akt phosphorylation. A, AM phosphory-lates Akt in a time-dependent manner. Rat aortas were incu-bated with 1027 mol/L AM for indicated periods. Fifty micro-grams of each protein extract was immunoblotted with aphosphospecific anti-Akt antibody (*pAkt), which recognizes cat-alytically active Akt, or anti-Akt antibody (Total Akt), which rec-ognizes total Akt1/2, regardless of whether Akt is phosphorylat-ed or not. B, AM phosphorylates Akt in a dose-dependentmanner. Rat aortas were incubated with AM at concentrationsranging from 10211 to 1027 mol/L for 15 minutes. Immunoblotanalyses were performed in the same way as in panel A. Histo-gram shows relative intensity of Akt phosphorylation by 1027

mol/L AM stimulation compared with nonstimulated control.*P,0.01 vs control. C, Effects of endothelium denudation onAM-induced Akt phosphorylation. Rat aortas with or without en-dothelium, E(1) or E(2), respectively, were incubated with 1027

mol/L AM for 15 minutes. Immunoblot analyses were performedin the same way as in panel A. Histogram shows relative inten-sity of Akt phosphorylation in rat aorta with or without endotheli-um. #P,0.05 vs E(2) (n54).

Nishimatsu et al Adrenomedullin-Induced Vasodilation and PI3K/Akt 65



ownloaded from


AM. AM also induced Akt phosphorylation in a dose-dependent fashion (Figure 2B). AM, even at a low concen-tration of 10211 mol/L, phosphorylated Akt and 1027 mol/LAM increased Akt phosphorylation by 5.3-fold (n54,P,0.01). We also examined the extent of Akt phosphoryla-tion in rat aorta in the presence and absence of endothelium(Figure 2C). The extent of Akt phosphorylation in denudedaorta was 69.5% (n53,P,0.05) compared with that in intactaorta. This finding indicated that'30% of total Akt phos-phorylation occurred in a monolayer of endothelium in rataorta, suggesting that enormous amounts of Akt were phos-phorylated by AM in vascular endothelium.

Previous reports showed that AM induced activation of ERK.We, therefore, examined whether AM induced ERK phosphor-ylation in rat aorta (Figure 3). AM induced ERK phosphoryla-tion in a dose-dependent manner in our ex vivo system. AM at1027 mol/L increased ERK phosphorylation by 2.4-fold (n53,P,0.01) 15 minutes after stimulation, although the ERK-dependent pathway did not seem to be involved in AM-inducedendothelium-dependent vasorelaxation.

We next examined whether LY and wortmannin indeedinhibited AM-induced Akt phosphorylation in our system. Asshown in Figure 4, AM-induced Akt phosphorylation wassignificantly inhibited by pretreatment of aortas with LY orwortmannin, whereas pretreatment with PD did not signifi-cantly inhibit AM-induced Akt phosphorylation. In fact,AM-induced Akt phosphorylation was suppressed to a level

lower than the control (nonstimulated) level by pretreatmentwith LY or wortmannin, suggesting a basal level of phos-phorylation of Akt in the vessels.

AM-Induced Akt Phosphorylation Is Mediated bythe Ca21/CaM-Dependent PathwayPrevious reports showed that PI3K activation was mediated bythe Ca21/CaM-dependent pathway in some cells.23 It was alsoreported that Akt was activated by the Ca21/CaM-dependentpathway in some cells.24 We, therefore, wondered whetherAM-induced Akt phosphorylation was mediated by the Ca21/CaM-dependent pathway. When the Ca21 concentration ofKrebs-Ringer bicarbonate solution was reduced, AM-inducedAkt phosphorylation decreased significantly in a dose-dependentmanner (Figure 5A). We also examined the effects of theCaM-dependent protein kinase inhibitor KN93 on AM-inducedAkt phosphorylation. AM-induced Akt phosphorylation wasinhibited by pretreatment with KN93 in a dose-dependentfashion (Figure 5B). Furthermore, A23187, a calcium iono-phore, increased Akt phosphorylation in a dose-dependent fash-ion, and A23187-induced Akt phosphorylation was abolished bypretreatment with PI3K inhibitors, suggesting that Akt phos-phorylation induced by Ca21/CaM was mediated by the PI3K-dependent pathway in rat aorta (see online Figure 1 in datasupplement available at http://www.circresaha.org).

AM-Induced Endothelium-Dependent VasodilationIs Diminished in Rat Aortas Expressing aDominant-Negative Akt Mutant in theEndotheliumTo examine more specifically the role of Akt expressed in theendothelium on AM-induced endothelium-dependent vasodi-

Figure 3. AM stimulates ERK phosphorylation. A, Rat aortaswere incubated with AM at concentrations ranging from 10211 to1027 mol/L for 15 minutes. Fifty micrograms of each proteinextract was immunoblotted with a phosphospecific anti-ERK1/2antibody (*pERK), which recognizes catalytically active ERK, oranti-ERK antibody (Total ERK), which recognizes total ERK1/2,regardless of whether ERK is phosphorylated or not. B, Histo-gram showing relative intensity of ERK phosphorylation by 1027

mol/L AM stimulation compared with nonstimulated control.#P,0.01 vs control (n53).

Figure 4. AM-induced Akt phosphorylation is inhibited by LYand wortmannin. A, Rat aortas were pretreated with 20 mmol/LLY, 100 nmol/L wortmannin (WT), or 20 mmol/L PD for 30 min-utes and stimulated with 1027 mol/L AM for 15 minutes. Immu-noblot analyses were performed as described in Figure 2 leg-end. B, Histogram showing relative intensity of phosphorylatedAkt. *P,0.001 vs 1027 mol/L AM stimulation (n54).




ownloaded from


lation, we infected Ad/dnAkt in rat abdominal aorta andexamined its effects. To confirm that adenovirus particlesinjected into the aortas had predominantly infected theendothelium, we injected Ad/GFP in abdominal aortas andexamined its localization by immunofluorescence. Becauserat aortas exhibited a marked green autofluorescence (fluo-rescence emission maxima [Em] 509 nm), we used an anti-GFP antibody to examine the localization of GFP expressionin them. As shown in Figure 6A, GFP was predominantlyexpressed in endothelium (red fluorescence [Em 570 nm],right panel), whereas its expression in the smooth musclelayer was negligible. Red fluorescence corresponding to GFPexpression was not observed in the endothelium of nonin-fected control aorta (left panel). Under the same conditions,AM induced vasodilation in a dose-dependent manner inaortic rings prepared from abdominal aortas infected withAd/GFP (Figure 6B). The extent of AM-induced vasodilationin Ad/GFP-infected aortic rings was almost the same as thatobserved in noninfected aortic rings with intact endothelium,suggesting that adenovirus infection was not cytotoxic toendothelium. In contrast, although AM elicited vasodilation

in aortic rings prepared from Ad/dnAkt-infected aortas in adose-dependent manner, AM-induced vasodilation was sig-nificantly reduced compared with that in Ad/GFP-infectedaortas. In fact, the AM-induced vasodilation observed inAd/dnAkt-infected aorta was reduced almost to the sameextent as that observed in endothelium-denuded aorta, sug-gesting that Akt expressed in endothelium played a major rolein AM-induced endothelium-dependent vasodilation.

AM-Induced cGMP Production Is Suppressed byPI3K/Akt InhibitionWe originally tried to examine whether AM induced eNOSphosphorylation and whether AM-induced eNOS phosphor-ylation was mediated by the PI3K/Akt-dependent pathway.We immunoprecipitated eNOS in protein extracts preparedfrom rat aortas and examined its phosphorylation by immu-noblotting with antiphosphoserine antibody. However, we didnot detect a significant amount of eNOS phosphorylation inaortas stimulated with AM (data not shown), probably be-cause the amount of phosphorylated eNOS was under detect-able levels. We therefore decided to use cGMP production inaortas as an indicator for NO production in endothelium. Weexamined whether AM induced cGMP production in rataortas and whether AM-induced cGMP production was me-diated by the PI3K/Akt-dependent pathway. AM significantlyincreased cGMP production in rat thoracic aortas in adose-dependent manner (Figure 7A). AM at 1027 mol/Lincreased the production of cGMP by 2.2-fold (n53,P,0.05). AM-induced increase of cGMP production wassignificantly inhibited to basal levels by pretreatment withLY. AM-induced increase of cGMP production was alsosignificantly inhibited to basal levels by chelating extracellu-lar Ca21 with EGTA. Therefore, the suppression of Aktphosphorylation by PI3K inhibition or by inhibition of theCa21/CaM-dependent pathway correlated with suppression ofcGMP production by inhibition of the PI3K- or Ca21/CaM-dependent pathway in rat aortas. Furthermore, pretreatmentwith L-NAME also significantly inhibited AM-inducedcGMP production to basal levels, suggesting that cGMPproduction could be an indicator of NO production in rataortas. Finally, we measured cGMP production in abdominalaorta infected with Ad/GFP or Ad/dnAkt (Figure 7B). Infec-tion of the abdominal aorta with those adenovirus constructsdid not remarkably change the basal production of cGMP. Incontrast, AM-induced increase of cGMP production wassignificantly inhibited in Ad/dnAkt-infected aortas comparedwith that in Ad/GFP-infected and noninfected aortas, indicat-ing that Akt expressed in the endothelium was involved inAM-induced NO/cGMP production.

Additional results appear in an online data supplementavailable at http://www.circresaha.org.

DiscussionAlthough it was reported that AM had a potent vasorelaxantactivity,2 the precise mechanisms were not clear. BecauseAM stimulates cAMP production, it was originally thoughtthat cAMP was a sole second messenger for AM-inducedvasodilation. On the other hand, we and others reported thatAM elicited endothelium-dependent vasodilation via, at least

Figure 5. AM-induced Akt phosphorylation depends on theCa21/CaM-mediated pathway. A, Rat aortas were incubated inKrebs-Ringer bicarbonate solution with decreasing concentra-tions of Ca21. Ca21-free solution was obtained by adding1 mmol/L EGTA to the solution. Rat aortas were then stimulatedwith 1027 mol/L AM for 15 minutes. Immunoblot analyses wereperformed as described in Figure 2 legend. Histogram showsrelative intensity of phosphorylated Akt. *P,0.01 and #P,0.05vs 2.5 mmol/L Ca21 (n54). B, Rat aortas were incubatedtogether with increasing doses of CaM-dependent proteinkinase inhibitor KN93. Rat aortas were then stimulated with 1027

mol/L AM for 15 minutes. Immunoblot analyses were performedas described in Figure 2 legend. Relative intensity of phosphory-lated Akt is also shown.




ownloaded from


partly, the NO/cGMP-mediated pathway.14,15 Thus, it waspossible that AM activated eNOS by stimulating the directbinding of CaM to eNOS and a subsequent conformationalchange of eNOS,16–18because AM reportedly has the capac-ity to increase [Ca21]i in endothelium.4,9 Here, we have shownthat AM-induced endothelium-dependent vasodilation andAM-induced increase of cGMP production, which we used asan indicator for NO production, depended on the PI3K/Akt-mediated pathway. To our knowledge, this is the first reportthat shows that AM induces Akt phosphorylation. Althoughwe detected Akt phosphorylation in whole vessels, severaldata presented in this paper indicated that Akt phosphoryla-tion occurring in vascular endothelium played a major role inAM-induced endothelium-dependent vasodilation. First, pre-treatment with LY or wortmannin of rat aortas with intactendothelium suppressed AM-induced vasodilation to thesame level as that in endothelium-denuded rat aortas. LY andwortmannin did not seem to cause remarkable cytotoxiceffects on vascular smooth muscle cell layers, because fors-kolin and SNP potently relaxed aortic rings pretreated withLY. Furthermore, pretreatment with PD did not inhibitAM-induced vasodilation in aortas with intact endothelium.Second, Akt phosphorylation occurring in a single layer ofvascular endothelium was estimated to be'30% of the totalphosphorylation of Akt occurring in the whole aorta, suggest-ing that a huge amount of Akt phosphorylation occurred invascular endothelium. Third, cGMP production in rat aortawas significantly inhibited by pretreatment with LY, and this

inhibition correlated with the suppression by LY of AM-induced endothelium-dependent vasorelaxation. Fourth, adnAkt mutant that expressed predominantly in endotheliumafter infection with Ad/dnAkt inhibited AM-induced vasodi-lation to the same level as that observed in endothelium-denuded aortas. Finally, AM-induced increase of cGMPproduction was significantly suppressed in aortas that ex-pressed dnAkt in the endothelium. Recently, two studies haveshown that Akt directly phosphorylates eNOS on Ser1179and that phosphorylated eNOS is active at a low Ca21

concentration.19,20 Thus, eNOS phosphorylation on Ser1179appears to be critical for facilitating CaM binding to eNOS.Our results indicated that AM-induced vasodilation wasalmost entirely endothelium dependent at a low concentrationof AM (1029 mol/L), at which concentration AM couldstimulate Akt phosphorylation and cGMP production. Incontrast, at high concentrations (1028 mol/L and 1027 mol/L),AM-induced vasodilation was mediated, partly, by an endo-thelium-independent mechanism; this was probably a directeffect of AM on vascular smooth muscle cells leading to anincrease of intracellular cAMP and vasodilation. Thus, al-though endothelium dependence of AM-induced vasodilationreportedly relies on the species and vessels used,5,25,26 itappears that, under physiological conditions, AM-inducedvasorelaxation is largely endothelium dependent and medi-ated by Akt activation, at least, in rat aorta.

Although it was reported that Ca21 was dispensable forshear stress-induced Akt phosphorylation,20 our results

Figure 6. AM-induced vasodilation is suppressedin rat aorta expressing a dominant-negative Aktmutant in endothelium. A, Adenovirus particlesinjected into rat abdominal aorta are predominantlyexpressed in the endothelium layer. Ad/GFP (109

PFU) was injected into rat abdominal aortas, andaortas were fixed 48 hours after infection. Nonin-fected rat abdominal aortas were used as negativecontrol. GFP-positive cells were stained with Cy(red fluorescence). Nuclei in aortas were stainedwith To-PRO-3 iodide (blue fluorescence). Areasshowing green fluorescence represent autofluores-cence in vessels. Note that nuclei in the endotheli-um that were double-stained with Cy andTo-PRO-3 iodide show pink fluorescence. Arrow-heads indicate GFP-positive nuclei in the endothe-lium. Bars510 mm. B, Ad/dnAkt (C) or Ad/GFP (v)was infected in the endothelium of rat abdominalaorta, and AM-induced vasodilation was examined48 hours after adenovirus infection. Noninfectedabdominal aortas with (▫) or without (n) endotheli-um were used as the control. #P,0.05 vs Ad/GFPinfection at each concentration of AM stimulation(n54).




ownloaded from


clearly showed that the Ca21/CaM-dependent pathway wasnecessary for AM-induced Akt phosphorylation. Our results alsoindicated that Ca21/CaM-induced Akt phosphorylation was me-diated by the PI3K-dependent pathway (see online Figure 1 indata supplement available at http://www.circresaha.org). Thus,Ca21/CaM appeared to be necessary not only for direct activationof eNOS by binding to the CaM binding domain in eNOS butalso for activation of eNOS via its phosphorylation by Akt. Theinvolvement of Ca21/CaM in the activation of PI3K/Akt hasbeen suggested in several reports. CaM-dependent protein ki-nase kinase directly phosphorylated Akt on threonine 308 andactivated Akt, which was required for antiapoptotic effectsinduced by the Ca21/CaM-dependent pathway.24 Insulin-inducedactivation of PI3K was inhibited in hepatocytes by pretreatmentwith EGTA or the Ca21 channel inhibitors gadolinium andnickel.23 Furthermore, the p110 subunits of PI3K contained ahighly conserved motif that bound to CaM in a Ca21-dependentway.27 On the other hand, it was reported that intracellular Ca21

was not required for epidermal growth factor– and platelet-

derived growth factor–induced Akt activation in BALB/c-3T3fibroblasts.28 It was also shown that the CaM-dependent proteinkinase inhibitor KN-62 did not inhibit insulin-induced PI3K orAkt activity in rat skeletal muscle.29 Thus, it appears that theinvolvement of Ca21/CaM in the activation of PI3K/Akt differsdepending on the mitogens and cell types used.

The precise pathways leading to activation of Ca21/CaMand PI3K/Akt by AM are not clear at present, partly becausethe cognate receptors for AM have not been sufficientlycharacterized.30,31 Although it was reported that AM in-creased [Ca21]i via phospholipase C activation and inositoltriphosphate formation,9 it is not clear how AM receptorsactivate phospholipase C. It also remains unclear whetherAM receptors have a docking site for the p85 subunit ofPI3K. Furthermore, there are no highly specific antagonistsfor AM receptors available at present. Thus, future studies arerequired to elucidate how AM receptors transmit signals tothe Ca21/CaM- and PI3K-dependent pathways.

Accumulated evidence has indicated that PI3K/Akt isimplicated in a variety of biological processes includingglycogen synthesis, antiapoptotic actions, modulation of cellproliferation and eNOS activation.19,20,32–34Furthermore, ithas been reported that PI3K/Akt is implicated in stimulationof angiogenesis.35 Thus, the PI3K/Akt-dependent pathwayseems to play pivotal roles in vascular endothelium such asmaintenance of vascular tone, protection against apoptoticcell death, repair of endothelium, and modulation of angio-genesis. Because AM has an activity to activate Akt andinduce NO production, modulation of AM activity in theendothelium may be a useful strategy to inhibit progression ofatherosclerosis.

AcknowledgmentsThis study was supported in part by Grants-in-Aid 09281206 and10218202 from the Ministry of Education, Culture and Science ofJapan (to Y.H.). We thank Reiko Sato, Etsuko Taira, and MarieMorita for technical assistance.

References1. Ross R, Glomset JA. The pathogenesis of atherosclerosis.N Engl J Med.

1976;295:369–377.2. Kitamura K, Kangawa K, Kawamoto M, Ichiki Y, Nakamura S, Matsuo

H, Eto T. Adrenomedullin: a novel hypotensive peptide isolated fromhuman pheochromocytoma.Biochem Biophys Res Commun. 1993;192:553–560.

3. Elhawary AM, Poon J, Pang CC. Effects of calcitonin gene-relatedpeptide receptor antagonists on renal actions of adrenomedullin.Br JPharmacol. 1995;115:1133–1140.

4. Hirata Y, Hayakawa H, Suzuki Y, Suzuki E, Ikenouchi H, Kohmoto O,Kimura K, Kitamura K, Eto T, Kangawa K, Matsuo H, Omata M.Mechanisms of adrenomedullin-induced vasodilation in the rat kidney.Hypertension. 1995;25:790–795.

5. Majid DS, Kadowitz PJ, Coy DH, Navar LG. Renal responses to intra-arterial administration of adrenomedullin in dogs.Am J Physiol. 1996;270:F200–F205.

6. Sugo S, Minamino N, Kangawa K, Miyamoto K, Kitamura K, Sakata J,Eto T, Matsuo H. Endothelial cells actively synthesize and secreteadrenomedullin.Biochem Biophys Res Commun. 1994;201:1160–1166.

7. Sugo S, Minamino N, Shoji H, Kangawa K, Kitamura K, Eto T, MatsuoH. Production and secretion of adrenomedullin from vascular smoothmuscle cells: augmented production by tumor necrosis factor-a. BiochemBiophys Res Commun. 1994;203:719–726.

8. Kubo A, Minamino N, Isumi Y, Katafuchi T, Kangawa K, Dohi K,Matsuo H. Production of adrenomedullin in macrophage cell line andperitoneal macrophage.J Biol Chem. 1998;273:16730–16738.

Figure 7. Effects of PI3K/Akt inhibition and low Ca21 onAM-induced cGMP production in rat aorta. A, Rat thoracic aor-tas were pretreated with 20 mmol/L LY or L-NAME or incubatedin Ca21-free Krebs-Ringer bicarbonate solution with 1 mmol/LEGTA. Rat aortas were then stimulated with 1027 mol/L AM for5 minutes, and cGMP extracted from aortas was measured byELISA. *P,0.05 vs 1027 mol/L AM stimulation (n55); #P,0.05vs nonstimulated control (n55). B, Ad/dnAkt (109 PFU) orAd/GFP (109 PFU) was injected into rat abdominal aortas, andexperiments were performed 48 hours after adenovirus infectionin the same way as in panel A. **P,0.001 and ##P,0.01, com-pared with cGMP level in Ad/dnAkt-infected aorta stimulatedwith AM (n53).




ownloaded from


9. Shimekake Y, Nagata K, Ohta S, Kambayashi Y, Teraoka H, Kitamura K,Eto T, Kangawa K, Matsuo H. Adrenomedullin stimulates two signaltransduction pathways, cAMP accumulation and Ca21 mobilization, inbovine aortic endothelial cells.J Biol Chem. 1995;270:4412–4417.

10. Iwasaki H, Eguchi S, Shichiri M, Marumo F, Hirata Y. Adrenomedullinas a novel growth-promoting factor for cultured vascular smooth musclecells: role of tyrosine kinase-mediated mitogen-activated protein kinaseactivation.Endocrinology. 1998;139:3432–3441.

11. Miller MJ, Martinez A, Unsworth EJ, Thiele CJ, Moody TW, Elsasser T,Cuttitta F. Adrenomedullin expression in human tumor cell lines: itspotential role as an autocrine growth factor.J Biol Chem. 1996;271:23345–23351.

12. Shimizu K, Tanaka H, Sunamori M, Marumo F, Shichiri M.Adrenomedullin receptor antagonism by calcitonin gene-related pep-tide(8–37) inhibits carotid artery neointimal hyperplasia after ballooninjury. Circ Res. 1999;85:1199–1205.

13. Michibata H, Mukoyama M, Tanaka I, Suga S, Nakagawa M, Ishibashi R,Goto M, Akaji K, Fujiwara Y, Kiso Y, Nakao K. Autocrine/paracrine roleof adrenomedullin in cultured endothelial and mesangial cells.Kidney Int.1998;53:979–985.

14. Hayakawa H, Hirata Y, Kakoki M, Suzuki Y, Nishimatsu H, Nagata D,Suzuki E, Kikuchi K, Nagano T, Kangawa K, Matsuo H, Sugimoto T,Omata M. Role of nitric oxide-cGMP pathway in adrenomedullin-induced vasodilation in the rat.Hypertension. 1999;33:689–693.

15. Shindo T, Kurihara H, Maemura K, Kurihara Y, Kuwaki T, Izumida T,Minamino N, Ju KH, Morita H, Oh-hashi Y, Kumada M, Kangawa K,Nagai R, Yazaki Y. Hypotension and resistance to lipopolysaccharide-induced shock in transgenic mice overexpressing adrenomedullin in theirvasculature.Circulation. 2000;101:2309–2316.

16. Venema RC, Sayegh HS, Kent JD, Harrison DG. Identification, charac-terization, and comparison of the calmodulin-binding domains of theendothelial and inducible nitric oxide synthases.J Biol Chem. 1996;271:6435–6440.

17. Salerno JC, Harris DE, Irizarry K, Patel B, Morales AJ, Smith SM,Martasek P, Roman LJ, Masters BS, Jones CL, Weissman BA, Lane P,Liu Q, Gross SS. An autoinhibitory control element defines calcium-regulated isoforms of nitric oxide synthase.J Biol Chem. 1997;272:29769–29777.

18. Nishida CR, Ortiz de Montellano PR. Autoinhibition of endothelialnitric-oxide synthase: identification of an electron transfer controlelement.J Biol Chem. 1999;274:14692–14698.

19. Fulton D, Gratton JP, McCabe TJ, Fontana J, Fujio Y, Walsh K, FrankeTF, Papapetropoulos A, Sessa WC. Regulation of endothelium-derivednitric oxide production by the protein kinase Akt.Nature. 1999;399:597–601.

20. Dimmeler S, Fleming I, Fisslthaler B, Hermann C, Busse R, Zeiher AM.Activation of nitric oxide synthase in endothelial cells by Akt-dependentphosphorylation.Nature. 1999;399:601–605.

21. Suzuki E, Nagata D, Yoshizumi M, Kakoki M, Goto A, Omata M, HirataY. Reentry into the cell cycle of contact-inhibited vascular endothelial

cells by a phosphatase inhibitor: possible involvement of extracellularsignal-regulated kinase and phosphatidylinositol 3-kinase.J Biol Chem.2000;275:3637–3644.

22. Fujio Y, Walsh K. Akt mediates cytoprotection of endothelial cells byvascular endothelial growth factor in an anchorage-dependent manner.J Biol Chem. 1999;274:16349–16354.

23. Benzeroual K, Pandey SK, Srivastava AK, van de Werve G, Haddad PS.Insulin-induced Ca21 entry in hepatocytes is important for PI 3-kinaseactivation, but not for insulin receptor and IRS-1 tyrosine phosphoryla-tion. Biochim Biophys Acta. 2000;1495:14–23.

24. Yano S, Tokumitsu H, Soderling TR. Calcium promotes cell survivalthrough CaM-K kinase activation of the protein-kinase-B pathway.Nature. 1998;396:584–587.

25. Feng CJ, Kang B, Kaye AD, Kadowitz PJ, Nossaman BD. L-NAMEmodulates responses to adrenomedullin in the hindquarters vascular bedof the rat.Life Sci. 1994;55:PL433–PL438.

26. Nossaman BD, Feng CJ, Kaye AD, DeWitt B, Coy DH, Murphy WA,Kadowitz PJ. Pulmonary vasodilator responses to adrenomedullin arereduced by NOS inhibitors in rats but not in cats.Am J Physiol. 1996;270:L782–L789.

27. Fischer R, Julsgart J, Berchtold MW. High affinity calmodulin targetsequence in the signalling molecule PI 3-kinase.FEBS Lett. 1998;425:175–177.

28. Conus NM, Hemmings BA, Pearson RB. Differential regulation bycalcium reveals distinct signaling requirements for the activation of Aktand p70S6k. J Biol Chem. 1998;273:4776–4782.

29. Brozinick JT Jr, Reynolds TH, Dean D, Cartee G, Cushman SW. 1-(N,O-bis-[5-isoquinolinesulphonyl]-N-methyl-L-tyrosyl)-4- phenylpiperazine(KN-62), an inhibitor of calcium-dependent camodulin protein kinase II,inhibits both insulin- and hypoxia-stimulated glucose transport in skeletalmuscle.Biochem J. 1999;339:533–540.

30. Kapas S, Catt KJ, Clark AJ. Cloning and expression of cDNA encodinga rat adrenomedullin receptor.J Biol Chem. 1995;270:25344–25347.

31. McLatchie LM, Fraser NJ, Main MJ, Wise A, Brown J, Thompson N,Solari R, Lee MG, Foord SM. RAMPs regulate the transport and ligandspecificity of the calcitonin-receptor-like receptor.Nature. 1998;393:333–339.

32. Chang HW, Aoki M, Fruman D, Auger KR, Bellacosa A, Tsichlis PN,Cantley LC, Roberts TM, Vogt PK. Transformation of chicken cells bythe gene encoding the catalytic subunit of PI 3-kinase.Science. 1997;276:1848–1850.

33. Cross DA, Alessi DR, Cohen P, Andjelkovich M, Hemmings BA. Inhi-bition of glycogen synthase kinase-3 by insulin mediated by proteinkinase B.Nature. 1995;378:785–789.

34. Yao R, Cooper GM. Requirement for phosphatidylinositol-3 kinase in theprevention of apoptosis by nerve growth factor.Science. 1995;267:2003–2006.

35. Kureishi Y, Luo Z, Shiojima I, Bialik A, Fulton D, Lefer DJ, Sessa WC,Walsh K. The HMG-CoA reductase inhibitor simvastatin activates theprotein kinase Akt and promotes angiogenesis in normocholesterolemicanimals.Nat Med. 2000;6:1004–1010.




ownloaded from


Kitamura and Yasunobu HirataKenneth Walsh, Masataka Sata, Kenji Kangawa, Hisayuki Matsuo, Atsuo Goto, Tadaichi Hiroaki Nishimatsu, Etsu Suzuki, Daisuke Nagata, Nobuo Moriyama, Hiroshi Satonaka,

Dependent Pathway in Rat Aorta−Phosphatidylinositol 3-Kinase/Akt Adrenomedullin Induces Endothelium-Dependent Vasorelaxation via the

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 2001 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/hh1301.0924982001;89:63-70; originally published online June 21, 2001;Circ Res.

http://circres.ahajournals.org/content/89/1/63World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circres.ahajournals.org/content/suppl/2001/06/13/hh1301.092498.DC1Data Supplement (unedited) at:

http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Researchin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:



ownloaded from

http://circres.ahajournals.org/content/89/1/63

http://circres.ahajournals.org/content/suppl/2001/06/13/hh1301.092498.DC1

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circres.ahajournals.org//subscriptions/


Online Supplementary Information

Corresponding author: Etsu Suzuki, M.D., Ph.D.

Title: Adrenomedullin Induces Endothelium-Dependent Vasorelaxation via the

Phosphatidylinositol 3-Kinase/Akt-Dependent Pathway in Rat Aorta

Submission type: online

RESULTS

To study more specifically the role of the Ca2+/CaM-dependent pathway in Akt

phosphorylation, we examined the effect of the Ca2+ ionophore A23187 on Akt

phosphorylation in rat aorta. A23187 stimulated Akt phosphorylation in a

dose-dependent manner (Online Figure 1). The effect of A23187 reached a maximal

level when 10-8 mol/L A23187 was used. A23187, at concentrations of 10-8 mol/L and

10-6 mol/L, increased Akt phosphorylation by 2.2-fold and 2.1-fold, respective ly,

compared with control aorta. Akt phosphorylation induced by 10-8 mol/L or 10-6 mol/L

A23187 was suppressed to a level lower than the control (nonstimulated) level by

pretreatment with 20 µmol/L LY or 100 nmol/L WT, suggesting that the

Ca2+/CaM-dependent Akt phosphorylation was mediated by the activation of PI3K.

FIGURE LEGEND

Online Figure 1 Effects of A23187 on Akt phosphorylation. Rat aortas were

incubated with increasing doses of A23187 for 15 min in the presence or absence of

20 µmol/L LY294002 (LY) or 100 nmol/L wortmannin (WT). One hundred µg of each

protein extract was immunoblotted with a phospho-specific anti-Akt antibody (*pAkt),

which recognizes catalytically active Akt, or anti-Akt antibody (Total Akt) which

recognizes total Akt1/2, regardless of whether Akt is phosphorylated or not. The

relative intensity of phosphorylated Akt is also indicated. Shown is a representative

result of two independent experiments in which the same results were obtained.

Matt Burns

2334/R1 online supplement

adrenomedullin induces endothelium-dependent...

Documents