Inhibition of Phosphoinositide 3-Kinase Potentiates Relaxation Of

Home , NONOate

Advance Publication by J-STAGE Circulation Journal Official Journal of the Japanese Circulation Society http://www.j-circ.or.jp Inhibition of Phosphoinositide 3-Kinase Potentiates Relaxation of Porcine Coronary Arteries Induced by Nitroglycerin by Decreasing Phosphodiesterase Type 5 Activity Dou Dou; Yixuan Guo; Lei Ying; Juan Liu; Xiaojian Xu; Xiaoxing Yu; Yuansheng Gao

Background: Vessel tension can be modulated by phosphoinositide 3-kinase (PI3K) acting on l-type calcium channel, rho kinase and phosphodiesterase (PDE) type 3 in smooth muscle cells. Inhibition of PI3K could increase the relaxation of porcine coronary arteries to nitroglycerin independent of this pathway, and the aim of the present study was therefore to determine the underlying mechanisms.

Methods and Results: Isolated porcine coronary arteries were dissected from the heart and cut into rings in ice-cold modified Krebs-Ringer bicarbonate buffer. The response of these vessels was studied by using the organ chamber technique; the content of cyclic guanosine monophosphate (cGMP) was determined by using enzyme-linked immunosorbent assay kit; and PI3K and Akt activity were determined by measuring the phosphorylation level of their downstream signaling molecule on Western blot. Inhibition of PI3K with 2-(4-morpholinyl)-8-phenyl-1(4H)-benzopyran- 4-one hydrochloride (LY294002) potentiated the relaxation of porcine coronary arteries to nitroglycerin and nitric oxide (NO), but not to 8-bromo-guanosine 3’5’-cyclic monophosphate, isoproterenol or (R)-(+)-trans-4-(1-Aminoethyl)-N-(4- Pyridyl)cyclohexanecarboxamide dihydrochloride monohydrate (Y27632). Increased relaxation induced by LY294002 was eliminated by Akt1/2 kinase inhibitor (Akt-I: 1,3-dihydro-1-(1-((4-(6-phenyl-1H-imidazo(4,5-g)quinoxalin-7-yl)phenyl) methyl)-4-piperidinyl)-2H-benzimidazol-2-one trifluoroacetate salt hydrate) or zaprinast, but was not affected by 1H- (1,2,4)oxadiazolo(4,3-a)quinoxalin-1-one, nifedipine or milrinone. Inhibition of Akt caused similar effects as LY294002. Incubation with LY294002 or Akt-I decreased the activity of PI3K and Akt but augmented the elevation of cGMP caused by NO. Enhanced cGMP elevation induced by LY294002 or Akt-I was also eliminated by zaprinast.

Conclusions: PI3K–Akt signaling may affect vascular tone through a stimulatory effect on PDE type 5.

Key Words: Nitric oxide; Smooth muscle; Vasodilatation

hosphoinositide 3-kinase (PI3K) participates in various ApoE(–/–) mice by activation of the Akt/eNOS pathway.6 cellular functions such as cell survival, metabolism, Studies also suggest that PI3K affects the constriction of vas- P angiogenesis, adhesion and proliferation in vascular cular smooth muscle via actions on l-type calcium channel,7–11 system.1–3 Some studies have found that PI3K signaling is also rho kinase12–14 and phosphodiesterase type 3 (PDE3).15,16 a key regulator of vascular tone.4 In the endothelium, activa- Nitroglycerin is widely used in the treatment of angina pec- tion of the PI3K/Akt pathway activates endothelial nitric oxide toris, hypertension and acute heart failure. It causes vasodilata- synthase (eNOS) by phosphorylating eNOS at ser-1177, and tion after being converted to NO or NO-related intermediate, also enhances nitric oxide (NO) release.5 Olmesartan and azel- followed by elevation of cyclic guanosine monophosphate nidipine improve severely impaired vasodilatation in diabetic (cGMP) via soluble guanylyl cyclase (sGC) and activation of

Received July 25, 2011; revised manuscript received September 28, 2011; accepted September 29, 2011; released online November 25, 2011 Time for primary review: 7 days Department of Physiology and Pathophysiology, Peking University Health Science Center, Beijing (D.D., Y. Guo, L.Y., J.L., X.X., X.Y., Y. Gao); Key Laboratory of Molecular Cardiovascular Science (Peking University), Ministry of Education, Beijing (D.D., Y. Gao), China This work was supported by the National Natural Science Foundation of China (Grant No. 30870938 and 81001433), Doctoral Fund of Ministry of Education for New Teachers (Grant No. 20100001120037) and Peking University Health Science Center Foundation for New Teachers (Grant No. BMU20090510). Mailing address: Dou Dou, PhD, Department of Physiology and Pathophysiology, Peking University Health Science Center, 38 Xue Yuan Road, Beijing, China, 100191. E-mail: [email protected] ISSN-1346-9843 doi: 10.1253/circj.CJ-11-0802 All rights are reserved to the Japanese Circulation Society. For permissions, please e-mail: [email protected] Advance Publication by J-STAGE DOU D et al.

Figure 1. Effect of 2-(4-morpholinyl)-8-phenyl-1(4H)-benzopyran-4-one hydrochloride (LY294002) on the relaxation of porcine coronary arteries to (A) nitroglycerin, (B) 2,2’-(hydroxynitrosohydrazono) bis(ethanamine) (DETA NONOate) and (C) 8-bromo- guanosine 3’5’-cyclic monophosphate (8-Br-cGMP). (A) Inhibition of phosphoinositide 3-kinase (PI3K) by LY294002 at 10–5 mol/L but not at 10–6 mol/L or 3×10–6 mol/L enhanced the relaxation of porcine coronary arteries induced by nitroglycerin. (B) Inhibition of PI3K with LY294002 (10–5 mol/L) also increased the relaxation of porcine coronary arteries to 2,2’-(hydroxynitrosohydrazono) bis(ethanamine) (DETA NONOate) but not to (C) 8-Br-cGMP. Data given as mean ± SEM. Group size: (A) n=8–22; (B) n=6–7; (C) n=8. *Significant difference between control group and LY294002 group (P<0.05).

cGMP-dependent protein kinase (PKG).17,18 Phosphodiester- relaxation mediated by nitrovasodilators. ase type 5 (PDE5) is essential to cGMP hydrolysis in vascular smooth muscle cells and plays an important role in the vasodilatation caused by nitroglycerin and NO.19–21 Activity of PDE5 Methods is enhanced by phosphorylation of the enzyme at serine 92 by Porcine Coronary Artery Preparation PKG and cAMP-dependent protein kinase (PKA).22,23 Previ- Fresh porcine hearts were collected from a local slaughter- ous studies have shown that PDE3 activity is modulated by house. The porcine coronary left circumflex arteries and left PI3K,15,16 prompting us to test whether PI3K may also regulate anterior descending arteries were carefully dissected and cut the activity of PDE5. The present study demonstrates that in- into rings (length: 5 mm) in ice-cold modified Krebs-Ringer hibition of PI3K increases the relaxation of porcine coronary bicarbonate buffer (composition [in mmol/L]: NaCl, 118.3; arteries induced specifically by nitrate, but not by 8-bromo- KCl, 4.7; CaCl2, 2.5; MgSO4, 1.2; KH2PO4, 1.2; NaHCO3, 25.0; guanosine 3’5’-cyclic monophosphate (8-Br-cGMP), isopro- glucose, 11.1).27,28 terenol or Y27632, a specific inhibitor of rho kinase.24 Inhibi- tion of PI3K also enhanced the cGMP elevation of porcine Vessel Tension coronary arteries induced by NO. The effect was eliminated Rings of coronary arteries were repeatedly rinsed and sus- by zaprinast, a specific PDE5 inhibitor,25 but not by nifedip- pended in organ chambers filled with 10 ml of the modified ine, milrinone or 1H-(1,2,4)oxadiazolo(4,3-a)quinoxalin-1-one Krebs-Ringer bicarbonate solution maintained at 37±0.1°C (ODQ), the specific inhibitor of sGC.26 Hence, the present re- and aerated with 95% O2–5% CO2 (pH=7.4). Two stirrups sults suggest that PI3K may stimulate the activity of PDE5 and passed through the lumen suspended on each ring. One stirrup thereby increase the degradation of cGMP and attenuate the was anchored to the bottom of the organ chamber, the other Advance Publication by J-STAGE PI3K and PDE5

Figure 2. Effect of (A) 1H-(1,2,4)oxadiazolo(4,3-a)quinoxalin-1-one (ODQ) or (B) zaprinast on the increased relaxation of nitroglycerin induced by 2-(4-morpholinyl)-8-phenyl-1(4H)-benzopyran-4-one hydrochloride (LY294002). Relaxation of porcine coronary arteries induced by nitroglycerin under control conditions, compared to (A) in the presence of LY294002 (10–5 mol/L), ODQ (10–7 mol/L), and LY294002 (10–5 mol/L) plus ODQ (10–7 mol/L); or (B) in the presence of LY294002 (10–5 mol/L), zaprinast (10–4.5 mol/L or 10–5.5 mol/L), and LY294002 (10–5 mol/L) plus zaprinast (10–4.5 mol/L or 10–5.5 mol/L). Data given as mean ± SEM; n=5–8 for each group. *Significant difference between control group and LY294002 group (P<0.05); †(A) significant difference between ODQ group and LY294002 plus ODQ group (P<0.05); †(B) significant difference between zaprinast (10–5.5 mol/L) group and LY294002 plus zaprinast (10–5.5 mol/L) group (P<0.05).

one was connected to a strain gauge, and the isometric force arteries were incubated in Krebs-Ringer bicarbonate buffer was measured with a force transducer (MLT0202/D; ADIn- maintained at 37±0.5°C and aerated with 95% O2–5% CO2 struments, Castle Hill, NSW, Australia) and recorded with an (pH 7.4) in the presence of solvent, LY294002 (10–5 mol/L) or ML785 PowerLab/8sp Recording and Analysis System (ADIn- Akt-I (10–5 mol/L) for 30 min. The vessel rings were then rap- struments).27,28 idly removed and immediately snap-frozen with liquid nitro- At the beginning of the experiment, each vessel ring was gen. Tissue lysates were prepared from isolated porcine coro- stretched to its optimal resting tension. This was achieved by nary arteries after treatment as aforementioned. Lysates each step-wise stretching until the active contraction of the vessel containing 20μ g of protein were subjected to sodium dodecy- ring to 100 mmol/L KCl reached a plateau. The optimal rest- lsulfide – polyacrylamide gel electrophoresis, and electro- ing tension of porcine coronary arteries was approximately transferred to polyvinylidene fluoride. Non-specific binding 2.5 g. After the vessels were brought to optimal resting ten- of antibody was blocked by washing with TBS buffer contain- sion, 1 h of equilibration was allowed. Effects of nitroglyc- ing 10% milk for 1 h. The blot was then subjected to 2 brief erin, 2,2’-(hydroxynitrosohydrazono) bis(ethanamine) (DETA washes with TBS plus 0.5% Tween-20, incubated with the NONOate), 8-Br-cGMP, (R)-(+)-trans-4-(1-aminoethyl)-N-(4- first antibody of Akt (Cell Signaling Technology, MA, USA; pyridyl)cyclohexanecarboxamide dihydrochloride monohy- 1:1,000 dilution), Akt-p (S473; Cell Signaling Technology; drate (Y27632) and isoproterenol were determined in vessels 1:1,000 dilution), FoxO1 (Cell Signaling Technology; 1:1,000 pre-constricted with (9,11)-dideoxy-(11α,9α)-epoxymetha- dilution) or FoxO1-p (S256; Cell Signaling Technology; 1:1,000 noprostaglandin F2α (U46619). The concentration – response dilution), overnight at 4°C and the secondary antibody for 1 h curves for these dilators were constructed in a cumulative fash- at room temperature. The blot was developed using the chemi- ion. In some experiments 2-(4-morpholinyl)-8-phenyl-1(4H)- luminescent detection method (Amersham ECLTM). Proteins benzopyran-4-one hydrochloride (LY294002; 10–5 mol/L), of Akt, Akt-p, FoxO1 or FoxO1-p in blots were quantified on Akt1/2 kinase inhibitor (Akt-I: 1,3-dihydro-1-(1-((4-(6-phenyl- densitometry using a Gel Doc 2000 densitometer (Bio-Rad, 1H-imidazo(4,5-g)quinoxalin-7-yl)phenyl)methyl)-4-piper- CA, USA) and normalized to scanning signals of actin (Cal- idinyl)-2H-benzimidazol-2-one trifluoroacetate salt hydrate; biochem, CA, USA). 10–5 mol/L), nifedipine (10–5 mol/L), milrinone (10–4 mol/L), ODQ (10–7 mol/L) or zaprinast (3×10–6 mol/L, 3×10–5 mol/L) Enzyme-Linked Immunosorbent Assay (ELISA) of cGMP were added 30 min before vessels were constricted with U46619 Rings of porcine coronary arteries were incubated in con- and remained in contact with the tissue throughout the experi- trol solution containing indomethacin and nitro-l-arginine ment. All experiments were carried out in a parallel fashion (10–4 mol/L; 37°C, 95% O2–5% CO2, pH 7.4) in the presence under control conditions or with different treatments. To elim- of solvent, LY294002, or Akt-I. In some experiments zaprinast inate the possible involvement of endogenous prostanoids and (3×10–5 mol/L), a specific inhibitor of PDE5, was also present. endothelium-derived NO, indomethacin (10–5 mol/L) and nitro- After 30-min equilibration, DETA NONOate (10–4 mol/L) was l-arginine (10–4 mol/L) were included in the present study added. Ten min later vessel rings were snap-frozen with liquid throughout the experiments.29 nitrogen. Tissue samples were ground to a fine powder under liquid nitrogen in a mortar. After the liquid nitrogen had evap- Western Blot orated, the frozen tissue was weighed and homogenized in 10 After endothelium was denuded, isolated porcine coronary volumes of 0.1 mol/L HCl. Tissues were sonicated 3 times for Advance Publication by J-STAGE DOU D et al.

Figure 3. The increased relaxation to nitroglycerin induced by inhibition of phosphoinositide 3-kinase (PI3K) with 2-(4-morpholinyl)-8-phenyl-1(4H)-benzopyran-4-one hydrochloride (LY294002; 10–5 mol/L) was not affected by (A) nifedipine, (B) (R)-(+)-trans- 4-(1-aminoethyl)-N-(4-pyridyl)cyclohexanecarboxamide dihydrochloride monohydrate (Y27632), (C) isoproterenol or (D) milrinone. Data are given as mean ± SEM; n=5–10 for each group. *Significant difference between control group and LY294002 group (P<0.05).

3 s, and centrifuged (13,000 g for 10 min) at room temperature. used had no effect on contraction to U46619 and relaxation The supernatant was extracted and the content of cGMP was induced by the NO donor and cGMP analogs. The other drugs determined using a monoclonal anti-cGMP antibody-based were prepared using distilled water. direct cGMP ELISA kit (catalog: 80103; NewEast Biosci- ences, Malvern, PA, USA). The protein content was determined based on the Bradford dye-binding procedure. Results Vessel Tension Statistical Analysis Rings of porcine coronary arteries were constricted with U46619 Data are given as mean ± SEM. Means of 2 groups were com- (3×10–7 mol/L) prior to testing their relaxant responses. There pared with 1-way ANOVA with Student-Newman-Keuls test was no significant difference in tension of vessels evoked by for post-hoc testing of multiple comparisons. Statistical sig- U46619 between the control and LY294002 (10–5 mol/L) groups nificance was accepted for P<0.05 (2 tailed). In all experiments, (16.2±1.6 g vs. 16.1±2.4 g, n=6–7 for each group, P>0.05). n represents the number of animals. Inhibition of PI3K with LY294002 at 10–5 mol/L but not at 10–6 mol/L or 3×10–6 mol/L significantly enhanced the relax- Reagents ation of porcine coronary arteries to nitroglycerin (Figure 1A). The following drugs were used (unless otherwise specified, all Inhibition of PI3K with LY294002 (10–5 mol/L) also increased were obtained from Sigma, St Louis, MO, USA): 8-Br-cGMP, the relaxation of porcine coronary arteries to DETA NONOate DETA NONOate, Akt-I, LY294002, milrinone, nitroglycerin (Figure 1B) but not to 8-Br-cGMP (Figure 1C). (NTG; Beijing Yimin Pharmaceutical, Beijing, China), ODQ, Relaxation of the artery to nitroglycerin in the presence of nifedipine, isoproterenol, U46619, Y27632 and zaprinast. ODQ (10–7 mol/L, a specific inhibitor of sGC) was less than Akt-I, LY294002, milrinone, nifedipine, ODQ and zaprinast that of the control. The augmentation of relaxation to nitroglyc- were dissolved in DMSO (final concentration: <0.2%). Pre- erin caused by inhibition of PI3K was not affected by ODQ liminary experiments showed that DMSO at the concentration (Figure 2A). The relaxation to nitroglycerin in the presence of Advance Publication by J-STAGE PI3K and PDE5

Figure 4. (A) Increased relaxation of nitroglycerin induced by 2-(4-morpholinyl)-8-phenyl-1(4H)-benzopyran-4-one hydrochloride (LY294002; 10–5 mol/L) was prevented by 1,3-dihydro-1-(1-((4-(6-phenyl-1H-imidazo(4,5-g)quinoxalin-7-yl)phenyl)methyl)-4-piperidinyl)-2H-benzimidazol-2-one trifluoroacetate salt hydrate (Akt-I; 10–5 mol/L) in porcine coronary arteries; (B) inhibition of Akt enhanced relaxation induced by 2,2’-(hydroxynitrosohydrazono) bis(ethanamine) (DETA NONOate) but (C) not that caused by 8- bromo-guanosine 3’5’-cyclic monophosphate (8-Br-cGMP). (D) Increased relaxation of nitroglycerin induced by Akt-I (10–5 mol/L) was prevented by zaprinast (3×10–5 mol/L). Data are given as mean ± SEM; n=4–6 for each group. *Significant difference between control group and LY294002 group (P<0.05).

zaprinast (3×10–5 mol/L or 3×10–6 mol/L, a specific inhibitor of ated by zaprinast (3×10–5 mol/L; a specific inhibitor of PDE5). PDE5) was greater than that in the absence of zaprinast. The The augmentation of relaxation to nitroglycerin caused by augmentation of relaxation to nitroglycerin caused by inhibi- inhibition of Akt was eliminated by zaprinast (Figure 4D). tion of PI3K was fully eliminated by zaprinast at 3×10–5 mol/L (Figure 2B). Western Blot The augmentation of relaxation to nitroglycerin caused Incubation of porcine coronary arteries with LY294002 for by LY294002 (10–5 mol/L) was not affected by nifedipine 30 min reduced the protein level of phosphorylated Akt sig- (10–5 mol/L, a specific inhibitor of l-type calcium channel; nificantly. The protein level of total Akt in the different groups Figure 3A). Inhibition of PI3K with LY294002 (10–5 mol/L) was not significantly changed Figure( 5A). Incubation with had no effect on relaxation of the arteries induced by Y27632 LY294002 or Akt-I for 30 min decreased the protein level of (10–5 mol/L, a specific inhibitor of rho kinase; Figure 3B) or phosphorylated FoxO1 significantly. The protein level of total isoproterenol (Figure 3C). The enhanced relaxation to nitro- FoxO1 in the different groups was not significantly affected glycerin caused by inhibition of PI3K was not affected by mil- (Figure 5B). rinone (10–4 mol/L, a specific inhibitor of PDE3;Figure 3D). Inhibition of Akt with Akt-I (10–5 mol/L, a specific inhibitor cGMP Assay of Akt1/2 kinase) significantly potentiated the relaxation of the DETA NONOate (10–4 mol/L; 10 min) significantly increased artery induced by nitroglycerin. Akt-I plus LY294002 caused the cGMP content of porcine coronary artery. Inhibition of a similar effect on the relaxation of porcine coronary arteries to PI3K with LY294002 potentiated the elevation of cGMP of the nitroglycerin as compared with LY294002 alone (Figure 4A). artery caused by DETA NONOate. The elevation of cGMP Inhibition of Akt also enhanced relaxation induced by DETA induced by DETA NONOate in the presence of zaprinast NONOate (Figure 4B), but not that induced by 8-Br-cGMP (3×10–5 mol/L; a specific inhibitor of PDE5) was more than (Figure 4C). Relaxation induced by nitroglycerin was potenti- that in those not treated with zaprinast. The augmentation of Advance Publication by J-STAGE DOU D et al.

Figure 5. (A) Akt-p (S473) and total Akt protein of porcine coronary arteries following 0.5-h incubation with solvent or 2-(4-morpholinyl)-8-phenyl-1(4H)-benzopyran-4-one hydrochloride (LY294002; 10–5 mol/L). (B) FoxO1-p (S256) and total FoxO1 protein of porcine coronary arteries following 0.5-h incubation with solvent, LY294002 (10–5 mol/L) or Akt-I (10–5 mol/L). (Lower panels) Densitometric scanning of (A) Akt-p (S473) protein normalized to total Akt or (B) FoxO1-p (S256) protein normalized to total FoxO1. Data given as mean ± SEM; n=6 for each group. *Significantly different from the solvent group (P<0.05).

Figure 6. Elevation of the intracellular content of cyclic guanosine monophosphate (cGMP) of porcine coronary arteries caused by 2,2’-(hydroxynitrosohydrazono) bis(ethanamine) (DETA NONOate) (10–4 mol/L) in the (A) absence of solvent, 2-(4-morpholinyl)- 8-phenyl-1(4H)-benzopyran-4-one hydrochloride (LY294002; 10–5 mol/L), zaprinast (10–4.5 mol/L) and LY294002 plus zaprinast or (B) in the presence of solvent, 1,3-dihydro-1-(1-((4-(6-phenyl-1H-imidazo(4,5-g)quinoxalin-7-yl)phenyl)methyl)-4-piperidinyl)-2H- benzimidazol-2-one trifluoroacetate salt hydrate (Akt-I; 10–5 mol/L), zaprinast (10–4.5 mol/L) and Akt-I plus zaprinast. Data are given as mean ± SEM; n=6 for each group. *Significantly different from the respective controls (P<0.05); †significantly different from the solvent group (P<0.05).

elevation of cGMP caused by inhibition of PI3K, however, of Akt was eliminated by zaprinast (Figure 6B). was eliminated by zaprinast (Figure 6A). Inhibition of Akt also potentiated the elevation of cGMP of the arteries caused by DETA NONOate. Elevation of cGMP Discussion induced by DETA NONOate was augmented by zaprinast but Previous studies have demonstrated that nitroglycerin is mainly the augmentation of elevation of cGMP caused by inhibition converted into NO or NO-related intermediates by mitochon- Advance Publication by J-STAGE PI3K and PDE5 drial aldehyde dehydrogenase (mtALDH) and cytochrome tive effect on nitroglycerin (Figure 4A). Inhibition of Akt P450 in vascular smooth muscle cells.17,18,30–32 It then leads to potentiated the relaxation of porcine coronary arteries caused elevation of cGMP by activation of sGC followed by an by NO (Figure 4B), but not to 8-Br-cGMP (Figure 4C). Fur- increased activity of PKG and vasodilatation. The present study thermore, the increased relaxation of the arteries caused by has demonstrated that inhibition of PI3K significantly potenti- inhibition of Akt could be prevented by zaprinast, the specific ated the relaxation of porcine coronary arteries caused by nitro- inhibitor of PDE5 (Figure 4D). Inhibition of Akt potentiated glycerin and NO, but not by 8-Br-cGMP (Figure 1). Hence, it the elevation of cGMP induced by DETA NONOate and such is likely that inhibition of PI3K augments the elevation of an effect was also blocked by zaprinast (Figure 6B). Hence, cGMP but not the activation of PKG by cGMP. it is likely that the relaxation of nitroglycerin modulated by It is well known that cGMP is a second messenger and plays PI3K is mediated by Akt. a key role in regulation of vessel tension.33,34 cGMP is synthe- Activation of PI3K increases phosphorylation of Akt at sized by guanylyl cyclase (GC), which converts guanosine-5’- serine 473,42 and active Akt can phosphorylate FoxO1 on triphosphate (GTP) to cGMP.35 GC exists in both membrane- serine 256.43 Thus, the levels of phosphorylation of Akt (S473) bound (pGC; particulate guanylyl cyclase) and cytosolic forms and FoxO1 (S256) represent the activity of PI3K and Akt. The (sGC). The pGC is activated by peptide hormones such as the present study showed that incubation of vessel with LY294002 atrial natriuretic factor,36 while sGC is typically activated by or Akt-I significantly decreased the activity of PI3K and Akt NO to stimulate cGMP synthesis.37 Cyclic nucleotide phospho- (Figure 5), which provides evidence for the effectiveness of diesterase (PDE) degrades cGMP by hydrolyzing cGMP into these inhibitors at the concentrations used. 5’-GMP.19–21 Although there are 11 different PDE gene fami- PDE5 has 2 homologous allosteric cGMP-binding sites, and lies expressed in mammalian tissues, the major PDE present cGMP binding to those sites does not affect the catalysis but in arterial smooth muscle are PDE1A, 1B and 1C, PDE3A and regulates the phosphorylation of PDE5.44 Phosphorylation of 3B, and PDE5.19 It is generally recognized that PDE5 is the PDE5 by PKG and PKA can alter its catalytic and allosteric most active cGMP-hydrolyzing PDE, while PDE1 and PDE3 cGMP-binding activities.22,23 Akt can phosphorylate PDE3 can hydrolyze both cAMP and cGMP.19 In the present study, and increase its activity15,16 while inhibition of PI3K/Akt leads we found that the increased relaxation of porcine coronary to accumulation of cAMP by attenuating PDE3 activity and arteries caused by inhibition of PI3K was not affected by ODQ, activation of PKA. In the present study, inhibition of PI3K had a specific inhibitor of sGC (Figure 2A), but was prevented no effect on the relaxation to isoproterenol (Figure 3C), and by zaprinast (3×10–5 mol/L), a specific inhibitor of PDE5 the increased relaxation induced by LY294002 could not be (Figure 2B). Inhibition of PI3K with LY294002 potentiated prevented by the specific inhibitor of PDE3 (Figure 3D). the elevation of cGMP induced by DETA NONOate signifi- Therefore it seems unlikely that PI3K/Akt affects PDE5 activ- cantly in porcine coronary arteries, and this was also blocked ity via the PDE3/PKA pathway. Rather, PI3K/Akt phosphory- by zaprinast (Figure 6A). These results suggest that inhibition lates and activates PDE5 instead of PDE3. of PI3K may affect vasodilatation by inhibiting the degradation In conclusion, the present study has demonstrated that inhibi- of cGMP through PDE5. PI3K inhibitor had no effect on the tion of PI3K specifically potentiates the vasodilatation induced relaxation induced by 8-Br-cGMP (Figure 1C). This may be by nitroglycerin or NO by decreasing the activity of PDE5. It because 8-Br-cGMP is more resistant to PDE than cGMP.38 may represent a novel mechanism by which PI3K modulates PI3K may modulate vasoconstriction by acting on l-type vascular tone. The present results also suggest that the modula- calcium channels.7–11 The present study showed that the tion of PDE5 by PI3K is mediated by Akt. Recent studies have increased relaxation caused by LY294002 could not be pre- shown that both expression and activity of PI3K are upregu- vented by nifedipine (Figure 3A). It has been reported that lated in hypertension and diabetes.45–49 This new mechanism PI3K is involved in the activation of rho kinase induced by may be involved in the development of hypertension or diabetic endothelin-1 in rabbit basilar artery.12,13 The present study arterial disease. showed that inhibition of PI3K had no effect on the relaxation of porcine coronary arteries to Y27632, the specific inhibitor of Acknowledgments rho kinase (Figure 3B). Recent studies also showed that activa- This work was supported by the National Natural Science Foundation of tion of PI3K could activate Akt and then increase the activity China (Grant No. 30870938 and 81001433), Doctoral Fund of Ministry of PDE3 by phosphorylating its serine at 290–292. PI3K inhib- of Education for New Teachers (Grant No. 20100001120037) and Peking itor such as Wortmannin or LY294002 inhibited insulin- University Health Science Center Foundation for New Teachers (Grant induced activation of Akt and PDE3, and their co-immunopre- No. BMU20090510). cipitation.15,16 In the present study we found that inhibition of PI3K had no effect on the relaxation of porcine coronary arter- References ies to isoproterenol (Figure 3C), and the increased relaxation 1. Kok K, Geering B, Vanhaesebroeck B. Regulation of phosphoinositide caused by LY294002 could not be prevented by milrinone, the 3-kinase expression in health and disease. Trends Biochem Sci 2009; specific inhibitor of PDE3 Figure( 3D). These results suggest 34: 115 – 127. 2. Morello F, Perino A, Hirsch E. Phosphoinositide 3-kinase signalling that, in porcine coronary artery, inhibition of PI3K regulates the in the vascular system. Cardiovasc Res 2009; 82: 261 – 271. nitroglycerin/NO-induced relaxation specifically by PDE5, 3. Oudit GY, Sun H, Kerfant BG, Crackower MA, Penninger JM, Backx while l-type calcium channel, rho kinase and PDE3 may not be PH. The role of phosphoinositide-3 kinase and PTEN in cardiovascu- involved. lar physiology and disease. J Mol Cell Cardiol 2004; 37: 449 – 471. 4. Sata M, Nagai R. Phosphatidylinositol 3-kinase: A key regulator of Akt, a serine/threonine protein kinase, is a key downstream vascular tone? Circ Res 2002; 91: 273 – 275. signaling molecule of PI3K and plays a critical role in multiple 5. Dimmeler S, Fleming I, Fisslthaler B, Hermann C, Busse R, Zeiher cellular processes caused by PI3K such as glucose metabo- AM. Activation of nitric oxide synthase in endothelial cells by Akt- lism, cell proliferation, apoptosis, transcription and cell migra- dependent phosphorylation. Nature 1999; 399: 601 – 605. 39–41 6. Hosoya M, Ohashi J, Sawada A, Takaki A, Shimokawa H. Combina- tion. The present study showed that inhibition of Akt tion therapy with olmesartan and azelnidipine improves EDHF-medi- enhanced the relaxation of porcine coronary arteries to nitro- ated responses in diabetic apolipoprotein E-deficient mice. Circ J glycerin. LY294002 plus Akt inhibitor, however, had no addi- 2010; 74: 798 – 806. Advance Publication by J-STAGE DOU D et al.

7. Viard P, Exner T, Maier U, Mironneau J, Nürnberg B, Macrez N. leucine zipper-positive isoform of MYPT1 may contribute to devel- Gbetagamma dimers stimulate vascular L-type Ca2+ channels via phos- opment of nitrate tolerance. Cardiovasc Res 2010; 86: 151 – 159. phoinositide 3-kinase. FASEB J 1999; 13: 685 – 694. 29. Gao Y, Dhanakoti S, Trevino EM, Sander FC, Portugal AM, Raj JU. 8. Viard P, Macrez N, Mironneau C, Mironneau J. Involvement of both Effect of oxygen on cyclic GMP-dependent protein kinase-mediated G protein alphas and beta gamma subunits in beta-adrenergic stimu- relaxation in ovine fetal pulmonary arteries and veins. Am J Physiol lation of vascular L-type Ca2+ channels. Br J Pharmacol 2001; 132: 2003; 285: L611 – L618. 669 – 676. 30. Kollau A, Hofer A, Russwurm M, Koesling D, Keung WM, Schmidt 9. Quignard JF, Mironneau J, Carricaburu V, Fournier B, Babich A, K, et al. Contribution of aldehyde dehydrogenase to mitochondrial Nurnberg B, et al. Phosphoinositide 3-kinase gamma mediates angio- bioactivation of nitroglycerin: Evidence for the activation of purified tensin II-induced stimulation of L-type calcium channels in vascular soluble guanylate cyclase through direct formation of nitric oxide. myocytes. J Biol Chem 2001; 276: 32545 – 32551. Biochem J 2005; 385(Pt 3): 769 – 777. 10. Le Blanc C, Mironneau C, Barbot C, Henaff M, Bondeva T, Wetzker 31. Yuan R, Sumi M, Benet LZ. Investigation of aortic CYP3A bioactiva- R, et al. Regulation of vascular L-type Ca2+ channels by phosphati- tion of nitroglycerin in vivo. J Pharmacol Exp Ther 1997; 281: 1499 – dylinositol 3,4,5-trisphosphate. Circ Res 2004; 95: 300 – 307. 1505. 11. Macrez N, Mironneau C, Carricaburu V, Quignard JF, Babich A, 32. Nakahira A, Minamiyama Y, Takemura S, Hirai H, Sasaki Y, Okada Czupalla C, et al. Phosphoinositide 3-kinase isoforms selectively S, et al. Coadministration of carvedilol attenuates nitrate tolerance couple receptors to vascular L-type Ca2+ channels. Circ Res 2001; 89: by preventing cytochrome p450 depletion. Circ J 2010; 74: 1711 – 692 – 699. 1717. 12. Miao L, Dai Y, Zhang J. Mechanism of RhoA/Rho kinase activation 33. Ignarro LJ, Kadowitz PJ. The pharmacological and physiological in endothelin-1-induced contraction in rabbit basilar artery. Am J role of cyclic GMP in vascular smooth muscle relaxation. Annu Rev Physiol Heart Circ Physiol 2002; 283: H983 – H989. Pharmacol Toxicol 1985; 25: 171 – 191. 13. Wehrwein EA, Northcott CA, Loberg RD, Watts SW. Rho/Rho kinase 34. Warner TD, Mitchell JA, Sheng H, Murad F. Effects of cyclic GMP and phosphoinositide 3-kinase are parallel pathways in the develop- on smooth muscle relaxation. Adv Pharmacol 1994; 26: 171 – 194. ment of spontaneous arterial tone in deoxycorticosterone acetate-salt 35. Ignarro LJ. Haem-dependent activation of guanylate cyclase and cyclic hypertension. J Pharmacol Exp Ther 2004; 309: 1011 – 1019. GMP formation by endogenous nitric oxide: A unique transduction 14. Seok YM, Azam MA, Okamoto Y, Sato A, Yoshioka K, Maeda M, et mechanism for transcellular signaling. Pharmacol Toxicol 1990; 67: al. Enhanced Ca2+-dependent activation of phosphoinositide 3-kinase 1 – 7. class IIα isoform-Rho axis in blood vessels of spontaneously hyper- 36. Chinkers M, Garbers DL, Chang MS, Lowe DG, Chin HM, Goeddel tensive rats. Hypertension 2010; 56: 934 – 941. DV, et al. A membrane form of guanylate cyclase is an atrial natri- 15. Han SJ, Vaccari S, Nedachi T, Andersen CB, Kovacina KS, Roth RA, uretic peptide receptor. Nature 1989; 338: 78 – 83. et al. Protein kinase B/Akt phosphorylation of PDE3A and its role in 37. Friebe A, Koesling D. Regulation of nitric oxide-sensitive guanylyl mammalian oocyte maturation. EMBO J 2006; 25: 5716 – 5725. cyclase. Circ Res 2003; 93: 96 – 105. 16. Ahmad F, Lindh R, Tang Y, Weston M, Degerman E, Manganiello 38. Lincoln TM, Cornwell TL. Intracelluar cyclic GMP receptor proteins. VC. Insulin-induced formation of macromolecular complexes involved FASEB J 1993; 7: 328 – 338. in activation of cyclic nucleotide phosphodiesterase 3B (PDE3B) and 39. Fayard E, Xue G, Parcellier A, Bozulic L, Hemmings BA. Protein its interaction with PKB. Biochem J 2007; 404: 257 – 268. kinase B (PKB/Akt), a key mediator of the PI3K signaling pathway. 17. Qin X, Zheng X, Qi H, Dou D, Raj JU, Gao Y. cGMP-dependent Curr Top Microbiol Immunol 2010; 346: 31 – 56. protein kinase in regulation of basal tone and in nitroglycerin- and nitric 40. Scheid MP, Woodgett JR. PKB/AKT: Functional insights from oxide-induced relaxation in porcine coronary artery. Eur J Physiol genetic models. Nat Rev Mol Cell Biol 2001; 2: 760 – 768. (Pflugers Arch)2007; 454: 913 – 923. 41. Eisenreich A, Malz R, Pepke W, Ayral Y, Poller W, Schultheiss HP, 18. Qi H, Zheng X, Qin X, Dou D, Xu H, Raj JU, et al. PKG regulates the et al. Role of the phosphatidylinositol 3-kinase/protein kinase B path- basal tension and plays a major role in nitrovasodilator-induced relax- way in regulating alternative splicing of tissue factor mRNA in human ation of porcine coronary veins. Br J Pharmocol 2007; 152: 1060 – endothelial cells. Circ J 2009; 73: 1746 – 1752. 1069. 42. Marte BM, Downward J. PKB/Akt: Connecting phosphoinositide 19. Rybalkin SD, Yan C, Bornfeldt KE, Beavo JA. Cyclic GMP phospho- 3-kinase to cell survival and beyond. Trends Biochem Sci 1997; 22: diesterases and regulation of smooth muscle function. Circ Res 2003; 355 – 358. 93: 280 – 291. 43. Zhang X, Gan L, Pan H, Guo S, He X, Olson ST, et al. Phosphoryla- 20. Omori K, Kotera J. Overview of PDEs and their regulation. Circ Res tion of serine 256 suppresses transactivation by FKHR (FOXO1) by 2007; 100: 309 – 327. multiple mechanisms. Direct and indirect effects on nuclear/cyto- 21. Kass DA, Takimoto E, Nagayama T, Champion HC. Phosphodies- plasmic shuttling and DNA binding. J Biol Chem 2002; 277: 45276 – terase regulation of nitric oxide signaling. Cardiovasc Res 2007; 75: 45284. 303 – 314. 44. Turko IV, Francis SH, Corbin JD. Binding of cGMP to both alloste- 22. Corbin JD, Turko IV, Beasley A, Francis SH. Phosphorylation of ric sites of cGMP-binding cGMP-specific phosphodiesterase (PDE5) phosphodiesterase-5 by cyclic nucleotide-dependent protein kinase is required for its phosphorylation. Clin Exp Biochem J 1998; 329 alters its catalytic and allosteric cGMP-binding activities. Eur J Bio- (Pt 3): 505 – 510. chem 2000; 267: 2760 – 2767. 45. Northcott CA, Hayflick JS, Watts SW. PI3-kinase upregulation and 23. Rybalkin SD, Rybalkina IG, Feil R, Hofmann F, Beavo JA. Regula- involvement in spontaneous tone in arteries from DOCA-salt rats: Is tion of cGMP-specific phosphodiesterase (PDE5) phosphorylation in p110delta the culprit? Hypertension 2004; 43: 885 – 890. smooth muscle cells. J Biol Chem 2002; 277: 3310 – 3317. 46. Northcott CA, Hayflick J, Watts SW. Upregulated function of phos- 24. Ishizaki T, Uehata M, Tamechika I, Keel J, Nonomura K, Maekawa phatidylinositol-3-kinase in genetically hypertensive rats: A modera- M, et al. Pharmacological properties of Y-27632, a specific inhibitor tor of arterial hypercontractility. Clin Exp Pharmacol Physiol 2005; of rho-associated kinases. Mol Pharmacol 2000; 57: 976 – 983. 32: 851 – 858. 25. Turko IV, Francis SH, Corbin JD. Potential roles of conserved amino 47. Northcott CA, Poy MN, Najjar SM, Watts SW. Phosphoinositide acids in the catalytic domain of the cGMP-binding cGMP-specific 3-kinase mediates enhanced spontaneous and agonist-induced con- phosphodiesterase. J Biol Chem 1998; 273: 6460 – 6466. traction in aorta of deoxycorticosterone acetate-salt hypertensive 26. Schrammel A, Behrends S, Schmidt K, Koesling D, Mayer B. Char- rats. Circ Res 2002; 91: 360 – 369. acterization of 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one as a 48. Vecchione C, Patrucco E, Marino G, Barberis L, Poulet R, Aretini A, heme-site inhibitor of nitric oxide-sensitive guanylyl cyclase. Mol et al. Protection from angiotensin II-mediated vasculotoxic and hyper- Pharmacol 1996; 50: 1 – 5. tensive response in mice lacking PI3Kgamma. J Exp Med 2005; 201: 27. Dou D, Zheng X, Qin X, Qi H, Liu L, Raj JU, et al. Role of cGMP- 1217 – 1228. dependent protein kinase in development of tolerance to nitroglycer- 49. Kobayashi T, Matsumoto T, Kamata K. The PI3-K/Akt pathway: Roles ine in porcine coronary arteries. Br J Pharmacol 2008; 153: 497 – related to alterations in vasomotor responses in diabetic models. J 507. Smooth Muscle Res 2005; 41: 283 – 302. 28. Dou D, Ma H, Zheng X, Ying L, Guo Y, Yu X, et al. Degradation of