Nitric oxide signaling in stretch-induced ... - The FASEB Journal

The FASEB Journal • FJ Express Full-Length Article

Nitric oxide signaling in stretch-induced apoptosis of neonatal rat cardiomyocytes Xudong Liao,*,†,1 Jun-Ming Liu,† Lei Du,* Aihui Tang,§ Yingli Shang,* Shi Qiang Wang,§ Lan-Ying Chen,†,2 and Quan Chen*,2 *State Key Laboratory of Biomembrane and Membrane Biotechnology, Institute of Zoology, Chinese Academy of Sciences; †Cardiovascular Institute and Fu Wai Hospital, Peking Union Medical College and Chinese Academy of Medical Sciences; and §State Key Laboratory of Biomembrane and Membrane Biotechnology, College of Life Sciences, Peking University, Beijing, China Pressure overload associated with hypertension is an important pathological factor leading to heart remodeling and ultimately heart failure partially due to cardiomyocyte apoptosis. Here we show that endogenous NO signaling plays a critical role in mechanical stretch-induced cardiomyocyte apoptosis. Mechanical stretch induced elevated expression of both eNOS and inducible NO synthase (iNOS) and increased synthesis of NO. A sustained increase in iNOS expression was also found in hearts of hypertensive rats in vivo. Blockade of NO signaling by inhibitors of NOS (L-NAME and AMT) or downstream guanylyl cyclase (ODQ) strongly inhibited stretch-induced apoptosis, mitochondria depolarization, and cytochrome c release, suggesting that NO is required in stretch-induced cardiomyocyte apoptosis. The expression of iNOS, but not eNOS, was blocked by L-NAME and ODQ, indicating that the iNOS induction is NO dependent. The initial elevation of NO is likely due to Ca2ⴙ-dependent activation of eNOS because elimination of intracellular calcium by EGTA-AM inhibited both iNOS induction and NO elevation. Other calcium signaling inhibitors (nifedipine, ryanodine, thapsigargin, and ionic gadolinium) also attenuated the initial NO elevation. These data indicate that mechanical signals initiate Ca2ⴙdependent NO synthesis, which is further amplified by activation of NO-induced iNOS expression, to regulate cardiomyocyte apoptosis.—Liao, X., Liu, J-M., Du, Lei, Tang, A., Shang, Y., Wang, S. Q., Chen, L-Y, Chen, Q. Nitric oxide signaling in stretch-induced apoptosis of neonatal rat cardiomyocytes. FASEB J. 20, E1196 –E1204 (2006) ABSTRACT

Key Words: transduction

hypertension 䡠 Ca2⫹ 䡠 eNOS 䡠 iNOS 䡠 signal

Abnormal mechanical load as seen in hypertension leads to heart remodeling and ultimately failure (1,2). Heart remodeling results from cardiomyocyte hypertrophy, cardiac fibroblast proliferation, and cardiomyocyte loss. Accumulating evidence suggests that the loss of functional cardiomyocytes due to apoptosis plays a critical role in the transition from hypertensive heart remodeling to heart failure (3, 4). Apoptosis has been E1196

recognized as a major mechanism in the development of a number of cardiovascular diseases, including hypertension (5–7). However, the mechanisms by which the mechanical signals lead to apoptosis are still obscure. Mechanical stretch was reported to activate the rennin-angiotensin system (RAS) in neonatal rat ventricular myocytes (NRVMs) to induce apoptosis in a p53dependent manner (8). However, there is evidence that argues against the requirement of RAS in regulating stretch-induced responses in cardiomyocytes (9). p53 may not be necessary for stretch-induced cardiomyocyte apoptosis since p53 is induced relatively late after stretch (10). The reactive oxygen species (ROS) were also reported to be critical for stretch-induced cardiac hypertrophic and apoptotic responses (11), but how ROS signaling was initiated by stretch and how it regulated apoptotic pathway are unknown. Recently, we and others reported that mechanical stretch activated Ca2⫹ signaling in cardiomyocytes (12, 13), and this Ca2⫹ signaling played an essential role in apoptosis initiation (12). However, the downstream signals that couple Ca2⫹ signaling to the apoptotic machinery remain to be elucidated. In this study, we demonstrate that nitric oxide (NO) is likely to be the link between stretch-induced Ca2⫹ signaling and apoptosis in cardiomyocytes. We found that mechanical signal initially activated a Ca2⫹-dependent NO synthesis, which was further amplified by activation of NO-induced iNOS expression, to regulate cardiomyocyte apoptosis. 1

Present address: Department of Molecular Genetics, Lerner Research Institute, Cleveland Clinic Foundation, 9500 Euclid Ave., Cleveland, OH 44195, USA. 2 Correspondence: Q.C., State Key Laboratory of Biomembrane and Membrane Biotechnology, Institute of Zoology, Chinese Academy of Sciences, 25 Beisihuanxi Rd., Beijing 100080, P.R. China. E-mail: [email protected]; L.-Y.C., Cardiovascular Institute & Fu Wai Hospital, Peking Union Medical College & Chinese Academy of Medical Sciences, 167 Beilishi Rd., Beijing 100037, P.R. China. E-mail: lanyingchen@hotmail. com doi: 10.1096/fj.06-5717fje 0892-6638/06/0020-1196 © FASEB

MATERIALS AND METHODS Materials Nitric oxide fluorescent probe DAF-FM DA (4-amino-5-methylamino-2⬘,7⬘-difluorofluorescein diacetate), mitochondrial membrane potential indicator DIOC6(3) (3,3⬘-dihexyloxacarbocyanine iodide), and EGTA-AM (EGTA Acetoxymethyl ester) were from Molecular Probes (Eugene, OR, USA). Nitrate reductase-based NO detection kit and immunohistochemistry avidin-biotin complex (ABC) kit were from Jingmei BioTech Co. Ltd (Beijing, China). DAPI (4,6-diamidino-2-phenyindole), EDTA, polyacrylamide gel reagents, bromodeoxyuridine (BrdU) (5-bromo-2⬘-deoxyuridine), arginine, and anti␤-actin monoclonal antibody (mAb) (Sigma-5316) were from Sigma (St. Louis, MO, USA). Trypsin, Dulbecco’s modified Eagle medium (DMEM)-F12 medium, and TRIzol reagent were products of Invitrogen (Carlsbad, CA, USA). FBS was from Hyclone (Logan, UT, USA). Anti-inducible nitric oxide synthase (NOS) (BD610328), anti-eNOS (BD610296), and anti-cyt c (BD556432) monoclonal antibodies were from BD (Franklin Lakes, NJ, USA). Horseradish peroxidase (HRP)labeled secondary antibodies were from KPL (Gaithersburg, MD, USA). The ECL kit was from Pierce (Rockford, IL, USA). SNAP (S-nitroso-N-acetylpenicillamine), AMT ((⫾)-2-amino5,6-dihydro-6-methyl-4H-1,3-thiazine, HCl), L-NAME (N␻-nitro-L-arginine-methyl ester), and ODQ (1H-[1,2,4]-oxadiazolo[4,3-A]-quinoxalin-1-one) were products of CalBiochem (Darmstadt, Germany). In Situ Cell Death Detection Kit was purchased from Roche Diagnostics (Indianapolis, IN, USA). Cell preparation and in vitro stretch model NRVMs were isolated from the hearts of Wistar rats (0 –3 days) and seeded in stretch chambers as described previously (12, 14). After 24 h serum starvation, the medium was changed, and the cells were subjected to 20% sustained stretch. Cells cultured under identical conditions without stretch were used as controls. Animal model of hypertension All animal studies were approved by Cardiovascular Institute Committee for the Care of Experimental Animals. Hypertensive rat model was produced in Wistar rats (male, 250 to 300 g, 8 to 10 wk old, supplied by the Experimental Animals Institute of Chinese Academy of Medical Sciences, Beijing, China) with abdominal aortic constriction (AAC) as described before (15). Rats in control group were sham-operated following a similar protocol, leaving the abdominal aortas intact. Carotid blood pressure was recorded in narcotic rats to monitor hypertension. Ventricles of AAC hearts as well as sham-operated controls were removed and used for immunohistochemistry and Western blot. Intracellular NO detection Intracellular NO level ([NO]i) was measured by confocal microscopy as described (12), using a NO-sensitive fluorescence probe DAF-FM (16). Briefly, cells were loaded with DAF-FM DA (10 ␮mol/l) at 37°C for 30 min in Tyrode buffer (in mmol/l: 137 NaCl, 5.4 KCl, 20 HEPES, 1.2 MgCl2, 1.0 CaCl2, 10 glucose (Glc), pH7.4), then gently washed twice and incubated for another 30 min to ensure complete cleavage of DAF-FM DA by the intracellular ester enzyme that releases the NO-sensitive probe (DAF-FM). Fluorescence was detected with a laser scanning confocal microscope (Zeiss ENDOGENOUS NO SIGNALING TOWARD APOPTOSIS

LSM 510) in Tyrode buffer containing arginine (100 ␮mol/l) at room temperature. [NO]i was calculated using the same approach as described for [Ca2⫹]i (12). Extracellular nitrate detection Extracellular total nitrate level ([NO]t) was determined by a nitrate reductase-based colorimetric kit. NO and NO-derived nitrate was converted to NO3⫺ by nitrate reductase, followed by the spectrophotometric quantification of nitrite levels using Griess Reagent. KNO3 (100 ␮mol/l) was used as nitrate standard. Immunohistochemistry Immunohistochemistry of iNOS and eNOS was performed following the regular protocol (17). Briefly, deparaffinized sections were blocked in 10% normal goat serum, then incubated with antibody iNOS or eNOS, 1:200 dilution) at 37°C for 1 h. Slides were washed and incubated with biotinylated secondary Ab. After treatment of slides with ABC kit, antigens were visualized with 3,3-diaminobenzidine (DAB) system. Apoptosis assays Apoptosis in cell culture was examined by annexin V-based flow cytometry and/or in situ DAPI staining as described (12). For flow cytometry, cells were stained with annexin V-FITC (25 ng/ml) simultaneously with PI (10 ng/ml) and the apoptotic population (annexin V⫹/PI⫺) was analyzed using a FACScan (BD) with CellQuest software. For DAPI staining, adhered cells in stretch chambers were fixed with 4% polyformaldehyde, permeabilized in PBS containing 0.1% Tween20, then stained with DAPI (10 ng/ml). DAPI fluorescence was visualized with fluorescent microscope under UV excitation. Apoptotic cells showed nuclear condensation and fragmentation. Apoptosis in rat heart in vivo was determined by in situ TUNEL assay following the manufacturer’s standard protocol. Briefly, 4% paraformaldehyde fixed cardiac tissues were paraffin embedded. After dewaxation, rehydration, and proteinase K treatment, samples were incubated with TUNEL reaction mixture for 60 min at 37°C in a humidified atmosphere in the dark. The slices were rinsed with PBS, then incubated with 50 ␮l converter POD in a humidified chamber for 30 min at 37°C. After DAB substrate treatment, the slices were mounted under glass coverslip and analyzed under a light microscope. Western blot Western blot was performed as described earlier (10). Cells were lysed in lysis buffer (in mmol/l: 25 HEPES, pH7.4, 5 EDTA, 8.0 EGTA, 1.0 Na3VO5, 0.25 NaF, 0.1 phenylmethylsulfonyl fluoride, 1.0 dithiothreitol, and 1% Nonidet P-40, 5 ␮g/ml aprotinin, 100 ␮g/ml leupeptin, 50 ␮g/ml trypsin inhibitor). Cellular protein (50 ␮g) was separated by SDSPAGE and transferred to a nitrocellulose membrane by standard electric transfer protocol. The membrane was probed with primary antibodies, then with HRP-labeled second Ab. Immunoreactive bands were visualized with ECL reagent. Cell fractionation Cell fractionation was performed as described before (10). Briefly, 1 ⫻ 106 cells were gently homogenized (5–10 strokes) E1197

with a Dounce homogenizer in buffer A (in mmol/l: 20 HEPES-KOH, pH7.2, 10 KCl, 1.5 MgCl2, 1.0 EDTA, 1.0 EGTA, 250 sucrose, 1.0 dithiothreitol, 0.1 phenylmethylsulfonyl fluoride, 5 ␮g/ml aprotinin, 100 ␮g/ml leupeptin, 50 ␮g/ml trypsin inhibitor) The homogenate was centrifuged at 750 g for 5 min at 4°C and the supernatant was subjected to further centrifugation at 10,000 g for 10 min at 4°C. The supernatant from this step was subjected to ultracentrifugation at 100,000 g for 45 min at 4°C, and the resulting pellet and supernatant were designated P100 and S100, respectively. Mitochondria membrane potential (⌬⌿m) assay The mitochondria membrane potential (⌬⌿m) was measured by DiOC6(3)-based flow cytometry as described (10). DiOC6(3) (20 nmol/l) was loaded into 4 ⫻ 105 cells suspended in 0.5 ␮l fresh DMEM (pH7.2) and fluorescence was examined at 530 ⫾ 30 nm (FL1 of BD FACScan). Data were obtained and analyzed using CellQuest software from a PI negative cell population. Reverse transcription-polymerase chain reaction (RT-polymerase chain reaction) Total RNA was isolated from cultured NRVMs using the TRIzol reagent and 5 ␮g RNA was reverse transcribed (RT) using oligo-dT. Two microliters of cDNA were amplified by polymerase chain reaction (PCR) using the following primers: iNOS, 5⬘-CCT AAG AGT CAC AAG CAT C-3⬘/5⬘-CTA TTT CCT TTG TTA CGG C-3⬘; eNOS, 5⬘-CGA GAT ATC TTC AGT CCC AAG C-3⬘/5⬘-GTG GAT TTG CTG CTC TGT AGG-3⬘; and glyceraldehyde-3-phosphate dehydrogenase (GAPDH), 5⬘-ATG CTG GTG CTG AGT ATG TC-3⬘/5⬘-TCC ACC ACC CTG TTG CTG TA-3⬘. Statistical analysis All results were expressed as the mean ⫾ sd of at least 3 independent experiments unless stated otherwise. Significance was determined with the Student’s t test. Difference between groups were considered significant at a value of P ⬍ 0.05.

RESULTS Stretch-activated NO synthesis in NRVMs Fluorescent microscopy data indicated that a 20% sustained stretch induced an increase in [NO]i within 5 min as demonstrated by an increase in DAF-FM fluorescence intensity (Fig 1A). The correlation between DAF-FM fluorescence and [NO]i was confirmed by the data from SNAP (a NO donor) -treated cells, which showed stronger fluorescence (Fig 1A). Stretch-induced [NO]i elevation was cardiomyocyte-specific since there was no detectable change of [NO]i in cardiac fibroblasts (data not shown). Quantification of [NO]i demonstrated that stretch induced rapid and significant [NO]i elevation (135⫾8% at 5 min, 121⫾7% at 10 min vs. 100% at 0 min, P ⬍ 0.05) in NRVMs (Fig 1B). After the initial transient elevation, [NO]i tended to recover but was maintained at a higher level than that E1198

Vol. 20

September 2006

Figure 1. Mechanical stretch induces Ca2⫹-dependent NO production in NRVMs. A) Intracellular NO concentration ([NO]i) was examined by DAF-based confocal analysis. Images were taken from cells before stretch (Control), 5 min after stretch (Stretch), and 10 min after SNAP (100 ␮mol/l) treatment (SNAP). NO donor SNAP treatment was used as a positive control. B) Confocal images were analyzed with IDL software to give quantitative NO levels. Digital images were taken from random fields (⬎50 pictures per time point) and pixel density was calculated, normalized with control data (from 10 min to 0 min before stretch), and changes of [NO]i (⌬[NO]i) were expressed as changes of fluorescence (⌬F/ F0). Error bars stand for se. *P ⬍ 0.05. C) Total nitrate level ([NO]t)was determined by nitrate reductase-based colorimetric method. Conditioned medium from control and 24 h stretched cells was examined. [NO]t was converted to [NO3⫺] using KNO3 (100 ␮mol/l) as standard. *P ⬍ 0.05. D) Cells were incubated with Ca2⫹ chelator EGTA-AM (2 ␮mol/l), NOS inhibitor L-NAME (100 ␮mol/l), or iNOS inhibitor AMT (100 ␮mol/l) for 30 min at 37°C before stretch. [NO]i was determined by DAF-based confocal analysis. Data from 10 min stretched cells were compared. *P ⬍ 0.05. E) Protein levels of iNOS and eNOS in NRVMs before and after 4 h stretch were detected by Western blot analysis. ␤-Actin was blotted to monitor equal loading.

of control cells (106⫾3% at 40 min vs. 100% at 0 min, P ⬍ 0.05) (Fig 1B). Microscopic methods are not useful in examining a long-term stretch effect. Therefore, we used another nitrate-reductase based method to detect NO after long-term stretch. Consistent with short-term stretch, there was an ⬃3-fold increase in total nitrate ([NO]t) detected in 24 h stretched cardiomyocyte compared to unstretched control (81.0⫾6.6 vs. 28.5⫾2.4, ␮mol/

The FASEB Journal

LIAO ET AL.

l.[NO⫺3], P⬍0.05) (Fig 1C). This result indicated that stretch-induced NO synthesis was sustained, although the initial [NO]i elevation was transient. L-NAME, a nonisoform-specific NOS inhibitor, completely abrogated stretch-induced [NO]i elevation, which suggested that stretch-induced elevation of [NO]i was due to de novo NO synthesis (Fig. 1D). However, AMT, an iNOS-specific inhibitor, failed to block the initial phase of stretch-induced NO elevation (Fig 1D), indicating that iNOS was not responsible for this portion of NO production. It is well known that Ca2⫹-dependent eNOS is constitutively expressed in cardiomyocyte (18) (also see Fig 1E). On the other hand, iNOS is Ca2⫹ independent and appears to be regulated at the transcriptional level (19). Therefore, we used EGTA-AM to eliminate the intracellular Ca2⫹ signal to distinguish roles of eNOS and iNOS in stretch-induced initial phase of NO synthesis. EGTA-AM treatment significantly inhibited stretch-induced [NO]i elevation (at 10 min), indicating the initial NO synthesis was Ca2⫹ dependent (Fig 1D). Together with the basal expression levels of eNOS and iNOS in cardiomyocyte (Fig 1E), these data demonstrated that stretch-induced initial [NO]i elevation was due to Ca2⫹dependent eNOS activation. Mechanical stretch also induced eNOS and iNOS expression in NRVMs as revealed by Western blot (Fig 1E). NOS expression and apoptosis in hearts of hypertension rats Since we found stretch induced significant induction of iNOS in vitro, we checked abdominal aorta constricted (AAC) hypertensive rats to confirm whether in vivo mechanical overload affected NOS expression in heart. We found that both eNOS and iNOS levels began to increase in hearts of AAC rats from 24 h after surgery compared to sham-operated controls, as revealed by immunohistochemistry (Fig 2A). Although eNOS expression returned to basal level within 7 days, iNOS induction remained significantly high after 2 wk from surgery (Fig 2A). Western blot analysis also showed that pressure overload induced expression of iNOS, but not eNOS, after 7 days of AAC operation (Fig 2B). The hypertension model was confirmed by monitoring rat mean arterial pressure (MAP). After surgery, MAP of AAC rats increased significantly within 24 h (blood pressure in mmHg: 121.0⫾7.5 vs. 104.5⫾6.3, P⬍0.05) and remained significantly higher (126.9⫾3.4 vs. 101.1⫾8.2 P⬍0.05) at 14 days after operation, despite a transient decrease at day 3 (98.8⫾8.7 vs. 109.2⫾1.7, P⬍0.05) (Fig 2C). In parallel with hypertension and NOS induction, there was increasing apoptosis in these AAC hypertensive hearts as revealed by TUNEL in situ apoptosis assay (Fig 2D). These observations demonstrated that in vivo mechanical overload indeed induced NOS expression in heart, which may have biological and pathological significance in regulating hypertension responses, including, but not exclusive to, apoptosis. ENDOGENOUS NO SIGNALING TOWARD APOPTOSIS

Figure 2. Pressure overload induces NOS expression in heart tissue from abdominal aortic constricted hypertensive rats. A) Immunohistochemistry of rat cardiac ventricle tissue. Left panel: Ventricle tissue sections from AAC (1 day) or sham-operated rats were stained with antibodies against iNOS and eNOS, respectively. Right panel: Quantitative analysis of iNOS and eNOS expression in vivo. The immunohistochemistry slices were scanned and digital pictures were analyzed with histological image software to obtain quantitative data. More than 50 pictures taken from different fields of each slice were calculated and the mean was taken as staining density. Three animals were included in each time point. All data were normalized with corresponding 1 day control (set density of 1 day sham-operated control as 1). Significance was determined by comparing means of staining density from 3 animals. *P ⬍ 0.05. Scale bar: 50 ␮m. AU: arbitrary unit. B) Two samples from AAC (7 days) and sham-operated rats were examined for iNOS and eNOS by Western blot. ␤-Actin was blotted to monitor equal loading. C) The mean arterial pressures (MAP) of the AAC and shamoperated groups at 1, 3, 7, and 14 days after surgery operation were compared. MAP was calculated as (2 ⫻ diastolic blood pressure ⫹ systolic blood pressure)/3. At least 4 animals were included at each time point. *P ⬍ 0.05. D) Hypertensionassociated apoptosis in heart tissue was determined by TUNEL in situ apoptosis assay. Dark brown staining indicated apoptotic nucleoli. Data presented were typical sections. Scale bar: 50 ␮m.

Requirement of NO signaling in stretch-induced cardiomyocyte apoptosis To directly examine the role of NO in mechanical stretch-induced apoptosis, we pretreated NRVMs with L-NAME 30 min before stretch to block NO synthesis and found that L-NAME blocked stretch-induced apoptosis (Fig 3A, B). After 4 h stretch, a significant portion E1199

inhibited stretch-induced apoptosis (Fig 3C), suggesting that the NO/cGMP pathway was involved in apoptosis regulation. Inconsistent with these NO blockers, NO donor SNAP enhanced stretch-induced apoptosis to a certain degree (Fig 3C). These data indicated that endogenous NO signaling was required for stretchinduced cardiomyocyte apoptosis. Although AMT inhibited apoptosis (Fig 3C), it failed to block iNOS induction (see Fig 5C), nor did it block the initial phase of NO production (Fig 1D), which suggested that the proapoptotic phase of NO elevation was produced by iNOS. NO signaling mediated mitochondria-dependent apoptotic pathway We next addressed the mechanism of NO-mediated apoptotic signaling in cardiomyocytes. We previously reported that mechanical stretch activated mitochondria-dependent apoptosis in NRVMs (10). So we checked whether NO signaling could mediate mitochondria-dependent apoptotic events. We found that L-NAME and AMT inhibited stretch-induced mitochondrial membrane potential (⌬⌿m) reduction and cytochrome c (cyt c) release (Fig. 4). ODQ also potently inhibited these events (Fig. 4). Collectively, these data

Figure 3. NO signaling participates in regulation of stretchinduced cardiomyocyte apoptosis. A) Apoptosis was examined with annexin V-based flow cytometry (upper panel) and DAPI-based fluorescent microscopy (lower panel). To block NO synthesis, cells were pretreated with L-NAME (100 ␮mol/l) at 37°C for 30 min before stretch. Cells were subjected to 20% sustained stretch for 4 h before apoptosis detection. Arrows indicated apoptotic nucleoli. B, C) Stretchinduced apoptosis was blocked by NO signaling inhibitors but enhanced by NO donor. Cells were incubated at 37°C for 30 min with L-NAME (100 ␮mol/l), AMT (100 ␮mol/l), ODQ (10 ␮mol/l), or SNAP (100 ␮mol/l), respectively, in each group before 4 h stretch. Control cells were incubated with inhibitors without stretch. Apoptotic index was shown as % of fragmented nucleoli (indicated by arrows in panel A). More than 200 nuclei from randomly selected fields (⬎10 fields per dish) were counted to calculate apoptosis index. *P ⬍ 0.05.

of stretched cells were annexin V⫹/PI⫺ as seen in flow cytometry analysis; a similar amount of cells had condensed and fragmented nuclei, morphological hallmarks of apoptosis as shown in DAPI staining. Apoptosis was significantly reduced in the presence of L-NAME by 2.4-fold (from 17⫾4% to 7⫾4%, Stretch only vs. Stretch plus L-NAME, P⬍0.05) (Fig 3B). In addition to L-NAME, a specific iNOS inhibitor, AMT, also inhibited apoptosis (Fig 3C), which indicated an essential role for iNOS. One of the key downstream targets of NO is soluble guanylyl cyclase (sGC), which produces the second message cGMP. Blockage of sGC by ODQ E1200

Vol. 20

September 2006

Figure 4. The roles of NO in mitochondrial apoptotic pathway and in regulating stretch-induced cardiomyocyte apoptosis. A) NO signaling inhibitors (L-NAME, AMT, or OQD) blocked stretch-induced mitochondria depolarization as shown by DiOC6(3)-based ⌬⌿m flow cytometry. Cells were incubated with inhibitors and stretched for 4 h as described in Fig. 3. Left shifting peak indicated a decrease of ⌬⌿m. Data were representative of 3 independent assays. B) L-NAME, AMT or OQD also blocked stretch-induced cyt c release from mitochondria. Cells were treated as described in panel A. Mitochondria free lysate fraction (S100) was examined by Western blot with anti-cyt c Ab. ␤-Actin was blotted to indicate equal loading.

The FASEB Journal

LIAO ET AL.

ical stretch induced iNOS gene at both the transcriptional and translational levels. We next asked what signals were responsible for stretch-induced NOSs (iNOS/eNOS) induction. We found that L-NAME blocked iNOS induction, suggesting that NO itself regulated iNOS expression (Fig. 5C). Since AMT failed to block the initial phase of NO production (Fig. 1D) and iNOS induction (Fig. 5C), it suggested that the initial phase of NO elevation was produced by eNOS and worked as the inducer of iNOS. OQD also blocked iNOS expression (Fig. 5C), which indicated that NO might induce iNOS expression through the cGMP pathway. Overall, these data argued that NO signaling was involved in the regulation of stretch-induced iNOS expression. These inhibitors did not affect eNOS expression, suggesting it was regulated through a different mechanism. The role of Ca2ⴙ signaling in NO cascade

Figure 5. Stretch induces eNOS and iNOS expression through different but interdependent pathways. A) The time course of iNOS and eNOS induction in NRVMs detected by Western blot. B) The time course of iNOS and eNOS induction in NRVMs detected by semiquantitative RT-PCR. GAPDH indicated equal loading in gel and equal cDNA input in RT-PCR. C) Roles of different NO signaling inhibitors in regulating stretch-induced NOS induction. Cells were treated with inhibitors, as described in Fig. 3, and were applied with or without 4 h stretch. The expression of iNOS and eNOS was detected by Western blot. ␤-Actin was used as loading control.

Since the initial [NO]i elevation was Ca2⫹ dependent (Fig. 1D), we further addressed how Ca2⫹ was involved in NO signaling. We previously demonstrated that mechanical stretch induces [Ca2⫹]i elevation, which depends on Ca2⫹ influx through L-type calcium channel (LCC), stretch-activated ion channel (SAC), and Ca2⫹-induced Ca2⫹ release (CICR) (12). We thus examined the effects of Ca2⫹ inhibitors (LCC blocker nifidipine, SAC blocker Gd3⫹, CICR inhibitors ryanodine, and thapsigargin) on NO production. These inhibitors have been shown to block stretch-induced [Ca2⫹]i elevation in cardiomyocytes (12), and here they also attenuated the initial [NO]i elevation (Fig 6A). EGTA-AM also attenuated stretch-induced NO production and iNOS expression (Fig. 1D, Fig. 6B). These data suggested that the stretch-induced Ca2⫹ signal was required for the initial NO burst and thus possibly acted as a trigger of the NO signaling cascade (by activating eNOS).

DISCUSSION showed that NO signaling mediated stretch-induced mitochondria-dependent cardiomyocyte apoptosis. Mechanisms of NOS activation We observed that stretch-induced NO signaling was required in apoptosis regulation in NRVMs and that the apoptotic NO was synthesized by NOS, perhaps iNOS. Therefore, we investigated the mechanisms by which iNOS and eNOS expressions were regulated. A time course study showed that iNOS induction was robust (increased with 30 min), and there was a 10-fold increase in iNOS protein expression 4 h after stretch compared to the basal level in unstretched (0 h) cells (Fig. 5A). We performed semiquantitative RT-polymerase chain reaction (RT-PCR) and found that iNOS mRNA also began to increase within 30 min after stretch (Fig. 5B). These results indicated that mechanENDOGENOUS NO SIGNALING TOWARD APOPTOSIS

In this study we addressed how mechanical stretch signals apoptosis in cardiomyocytes. Our data suggested a mechanical stretch-induced Ca2⫹-NO signaling cascade in NRVMs as summarized in Fig. 7. Stretch may initiate the cascade through [Ca2⫹]i elevation and Ca2⫹-dependent eNOS activation. eNOS produces the initial NO, which is further amplified through NOinduced iNOS expression. iNOS produces a high dose of NO, which is able to regulate stretch-induced cardiomyocyte apoptosis and/or other stretch-induced responses. We presented the following lines of evidence: 1) mechanical stretch activated NOS expression and Ca2⫹-dependent NO synthesis; 2) blockage of NO synthesis or downstream cGMP signaling inhibited stretchinduced cardiomyocyte apoptosis and critical apoptotic mitochondrial events (⌬⌿m depolarization, cyt c release); 3) blockage of initial NO synthesis or downE1201

(27). We have reported that both mitochondria and death receptor-mediated apoptosis were activated by mechanical stretch in NRVMs (10, 28). It is possible that NO participates in both types of apoptotic pathways (mitochondria and death receptor dependent) to regulate stretch-induced cardiomyocyte apoptosis (Fig. 7). NO can also activate soluble guanylate cyclase (sGC) to produce cGMP, and subsequently activates cGMPdependent PKG to induce apoptosis (29). This is consistent with our observation that sGC inhibitor ODQ potently inhibited apoptotic events. However, there is a report arguing that activation of sGC/cGMP pathway by low concentrations of NO rescues cells from apoptosis (30). The controversy about the pro- or antiapoptotic effects of NO may be due to different cellular models. Although a low concentration of NO has been found to protect macrophages from apoptosis (30), NO concentration in our settings may already be high enough to promote apoptosis in cardiomyocytes. We showed both in vitro and in vivo evidence that mechanical overload significantly increased iNOS expression in cardiac cells. Overall, once activated by

Figure 6. Stretch-induced calcium signaling regulates NO synthesis and iNOS induction. A) Ca2⫹ signaling inhibitors attenuated the initial phase of stretch-induced NO elevation. Cells were incubated at 37°C with or without nifedipine (2 ␮mol/l), ryanodine (10 ␮mol/l), thapsigargin (10 ␮mol/l) or GdCl3 (100 ␮mol/l) for 30 min before stretch, respectively in each group. [NO]i was determined by DAF-based confocal microscopy and calculated as described in Fig. 1. Changes of [NO]i (⌬[NO]i) after 10 min stretch were compared. *P ⬍ 0.05. B) Eliminating [Ca2⫹]i by EGTA-AM inhibited stretchinduced iNOS expression. Cells were incubated with EGTA-AM (2 ␮mol/l) at 37°C for 30 min before stretch to eliminate intracellular Ca2⫹. Cells were stretched for 4 h or left free as indicated. iNOS expression was examined by Western blot. ␤-Actin was blotted to indicate equal loading.

stream cGMP signaling blocked stretch-induced iNOS expression, and deactivating iNOS inhibited stretchinduced apoptosis; and 4) mechanical overload induced transient eNOS elevation and sustained iNOS up-regulation in cardiac ventricle of hypertension rats in vivo. NO has been reported to be a bidirectional regulator for apoptosis, as it can either promote or inhibit apoptosis in cardiomyocyte and other types of cells (20 –24). NO can directly induce the nitrosylation of some apoptosis regulatory molecules, such as caspase 3 and cyt c, to regulate apoptosis (25), or directly target mitochondria to induce apoptosis by inhibiting mitochondrial respiration (26). Besides the mitochondriadependent apoptotic pathway, NO also plays important roles in death receptor-dependent apoptotic pathway E1202

Vol. 20

September 2006

Figure 7. Summary of hypothetic NO signaling cascade activated by mechanical stretch in cardiomyocytes. Mechanical stretch initiates the NO cascade by activating constitutive expressed eNOS through [Ca2⫹]i elevation. The active eNOS produces a small amount NO, which further induces iNOS expression. The iNOS, once expressed, produces larger amounts of NO and thus amplifies NO signaling. Finally, NO concentration reaches a certain level that is high enough to induce apoptosis and/or other responses in cardiomyocytes. NO may affect mitochondria or other apoptotic pathways to induce apoptosis. Besides this NO cascade, stretch-induced Ca2⫹ may participate in other signaling pathways. Solid lines: signaling discovered in this study. Dashed lines: possible signaling.

The FASEB Journal

LIAO ET AL.

mechanical overload, NO signaling may have multiple targets to induce apoptosis in cardiomyocytes. Endogenous NO biosynthesis is tightly regulated by three distinct isoforms of NOS: the neuronal (nNOS, NOS I), endothelial (eNOS, NOS III), and inducible (iNOS, NOS II) forms. It has been reported that iNOS is proapoptotic, and increased iNOS expression has been found in a number of heart diseases (22–24, 31). Overexpression of iNOS in mouse heart results in heart block and sudden death (32), and lack of iNOS seems to be of benefit to the preservation of heart function in hypertensive mice (33). In this study we found that mechanical stretch significantly induced the expression of both eNOS and iNOS in vitro and in vivo. Notably, the in vivo increase of iNOS expression was sustained in hypertensive hearts. These results strongly suggest that iNOS plays important roles in regulating pathological outcomes of hypertension. Our data also demonstrated that iNOS-produced NO was proapoptotic, since AMT blocked apoptosis without altering iNOS expression or eNOS-produced initial NO. Although eNOS induction was transient in vivo, its activation was essential in initiating stretch-induced NO cascade, the inhibition of which resulted in blockage of iNOS gene expression. Our results are inconsistent with a previous report suggesting that stretch suppresses cytokine-induced iNOS expression in cardiomyocytes (34). The discrepancy may be due to the difference between the experimental models. The cyclic stretch in their study promoted hypertrophy rather than apoptosis (11), and we didn’t treat cells with any cytokines. Both L-NAME and ODQ strongly suppress stretchinduced iNOS expression, indicating that iNOS expression may be regulated by a NO-dependent mechanism, as reported (23). Moreover, many signaling pathways may participate in iNOS gene regulation (35, 36), therefore, it would be interesting to examine whether mechanical stretch-induced NO can activate some transcriptional factors, such as STATs, NF␬B, etc., to induce iNOS expression. The induction of eNOS is apparently regulated by a distinct pathway, since the NOS and Ca2⫹ inhibitors had little effect on its expression. Induction of eNOS (activity and/or expression) has been reported in genetically engineered hypertensive rats, and is related to heart failure, but the mechanisms remain unclear (37). The link between stretch-induced Ca2⫹ signaling and NO signaling has been demonstrated clearly in the present study, as we showed that Ca2⫹ inhibitors (EGTA-AM, nifidipine, Gd3⫹, ryanodine, and thapsigargin) significantly attenuated stretch-induced [NO]i elevation and iNOS expression (Fig 1D, Fig. 6). Therefore, the stretch-induced Ca2⫹ signal is very likely to be the trigger of the subsequent NO signaling in cardiomyocytes, and the NO signaling further amplifies itself by iNOS expression to regulate stretch-induced responses including, but maybe not exclusive to, apoptosis (Fig. 7). We previously discussed the possibility for Ca2⫹ signaling to be the initiator of stretch-induced apoptosis in cardiomyocytes (12), which may regulate ENDOGENOUS NO SIGNALING TOWARD APOPTOSIS

apoptosis through many known and/or unknown pathways (Fig. 7). Clearly, the NO signaling discovered in the present study is likely to be the most important one. Deregulation of eNOS and iNOS has been closely associated with hypertension, myocardial infarction, heart failure, and cardiac transplantation (24, 38). It is possible that activation of NO signaling could be an adaptive response to protect cardiac cells from mechanical overload-induced damage. However, maladaptive responses with a chronically high dose of NO and enhanced iNOS expression (and the potential feedback reinforcement) could be associated with the progression of heart diseases. Indeed, chronically increased NO levels and enhanced NOS expression have deteriorating effects on heart cells by directly targeting mitochondria or other cellular organelles (27, 36). Understanding the role of NO signaling in heart diseases would be useful for the development of drugs that can selectively promote the beneficial effect of NO. This study was supported by the “Major National Basic Research Program” (973 Program. No. 2004CB720007, 2002CB513100, and G2000056904 to Q.C, S.Q.W and L.Y.C.), National Natural Science Foundation of China (No. 30170467, 30421004 and 30425035 to Q.C., L.Y.C., and S.Q.W), the Ph.D. Programs Foundation (L.Y.C.), the Outstanding Young Investigator Award (Q.C.), and the “Knowledge Innovation Key Project (KSCX2-sw-2010)” (Q.C.). We thank David Shultz, Mark Jackson, and Meredith Crosby (Cleveland Clinic Foundation, Cleveland, OH, USA) for their critical reading of the manuscript.

REFERENCES 1.

2. 3. 4.

5. 6. 7. 8.

9.

Meghji, P., Nazir, S. A., Dick, D. J., Bailey, M. E., Johnson, K. J., and Lab, M. J. (1997) Regional workload induced changes in electrophysiology and immediate early gene expression in intact in situ porcine heart. J. Mol. Cell Cardiol. 29, 3147–3155 Swynghedauw, B., and Delcayre, C. (1982) Biology of cardiac overload. Pathobiol. Annu. 12, 137–183 Gill, C., Mestril, R., and Samali, A. (2002) Losing heart: the role of apoptosis in heart disease—a novel therapeutic target? FASEB J. 16, 135–146 Fortuno, M. A., Ravassa, S., Fortuno, A., Zalba, G., and Diez, J. (2001) Cardiomyocyte apoptotic cell death in arterial hypertension: mechanisms and potential management. Hypertension 38, 1406 –1412 Kang, P. M., and Izumo, S. (2003) Apoptosis in heart: basic mechanisms and implications in cardiovascular diseases. Trends Mol. Med. 9, 177–182 Bishopric, N. H., Andreka, P., Slepak, T., and Webster, K. A. (2001) Molecular mechanisms of apoptosis in the cardiac myocyte. Curr. Opin. Pharmacol. 1, 141–150 Narula, J., Kolodgie, F. D., and Virmani, R. (2000) Apoptosis and cardiomyopathy. Curr. Opin. Cardiol. 15, 183–188 Leri, A., Claudio, P. P., Li, Q., Wang, X., Reiss, K., Wang, S., Malhotra, A., Kajstura, J., and Anversa, P. (1998) Stretchmediated release of angiotensin II induces myocyte apoptosis by activating p53 that enhances the local renin-angiotensin system and decreases the Bcl-2-to-Bax protein ratio in the cell. J. Clin. Invest. 101, 1326 –1342 Van Kesteren, C. A., Saris, J. J., Dekkers, D. H., Lamers, J. M., Saxena, P. R., Schalekamp, M. A., and Danser, A. H. (1999) Cultured neonatal rat cardiac myocytes and fibroblasts do not synthesize renin or angiotensinogen: evidence for stretch-induced cardiomyocyte hypertrophy independent of angiotensin II. Cardiovasc. Res. 43, 148 –156

E1203

10.

11.

12.

13.

14. 15. 16.

17.

18. 19. 20.

21.

22. 23.

24.

Liao, X. D., Wang, X. H., Jin, H. J., Chen, L. Y., and Chen, Q. (2004) Mechanical stretch induces mitochondria-dependent apoptosis in neonatal rat cardiomyocytes and G2/M accumulation in cardiac fibroblasts. Cell Res. 14, 16 –26 Pimentel, D. R., Amin, J. K., Xiao, L., Miller, T., Viereck, J., Oliver-Krasinski, J., Baliga, R., Wang, J., Siwik, D. A., Singh, K., Pagano, P., Colucci, W. S., and Sawyer, D. B. (2001) Reactive oxygen species mediate amplitude-dependent hypertrophic and apoptotic responses to mechanical stretch in cardiac myocytes. Circ. Res. 89, 453– 460 Liao, X. D., Tang, A. H., Chen, Q., Jin, H. J., Wu, C. H., Chen, L. Y., and Wang, S. Q. (2003) Role of Ca2⫹ signaling in initiation of stretch-induced apoptosis in neonatal heart cells. Biochem. Biophys. Res. Commun. 310, 405– 411 Petroff, M. G., Kim, S. H., Pepe, S., Dessy, C., Marban, E., Balligand, J. L., and Sollott, S. J. (2001) Endogenous nitric oxide mechanisms mediate the stretch dependence of Ca2⫹ release in cardiomyocytes. Nat. Cell Biol. 3, 867– 873 Lee, A. A., Delhaas, T., Waldman, L. K., MacKenna, D. A., Villarreal, F. J., and McCulloch, A. D. (1996) An equibiaxial strain system for cultured cells. Am. J. Physiol. 271, C1400 –C1408 Xu, L., Liu, J. M., and Chen, L. Y. (2004) CREG, a new regulator of ERK1/2 in cardiac hypertrophy. J. Hypertens. 22, 1579 –1587 Kojima, H., Nakatsubo, N., Kikuchi, K., Kawahara, S., Kirino, Y., Nagoshi, H., Hirata, Y., and Nagano, T. (1998) Detection and imaging of nitric oxide with novel fluorescent indicators: diaminofluoresceins. Anal. Chem. 70, 2446 –2453 Aizawa, T., Ishizaka, N., Taguchi, J., Nagai, R., Mori, I., Tang, S. S., Ingelfinger, J. R., and Ohno, M. (2000) Heme oxygenase-1 is upregulated in the kidney of angiotensin II-induced hypertensive rats : possible role in renoprotection. Hypertension 35, 800 – 806 Hare, J. M. (2003) Nitric oxide and excitation-contraction coupling. J. Mol. Cell Cardiol. 35, 719 –729 Alderton, W. K., Cooper, C. E., and Knowles, R. G. (2001) Nitric oxide synthases: structure, function and inhibition. Biochem. J. 357, 593– 615 Pinsky, D. J., Aji, W., Szabolcs, M., Athan, E. S., Liu, Y., Yang, Y. M., Kline, R. P., Olson, K. E., and Cannon, P. J. (1999) Nitric oxide triggers programmed cell death (apoptosis) of adult rat ventricular myocytes in culture. Am. J. Physiol. 277, H1189 – H1199 Beltran, B., Mathur, A., Duchen, M. R., Erusalimsky, J. D., and Moncada, S. (2000) The effect of nitric oxide on cell respiration: A key to understanding its role in cell survival or death. Proc. Natl. Acad. Sci. U. S. A. 97, 14602–14607 Ing, D. J., Zang, J., Dzau, V. J., Webster, K. A., and Bishopric, N. H. (1999) Modulation of cytokine-induced cardiac myocyte apoptosis by nitric oxide, Bak, and Bcl-x. Circ. Res. 84, 21–33 Jones, W. K., Flaherty, M. P., Tang, X. L., Takano, H., Qiu, Y., Banerjee, S., Smith, T., and Bolli, R. (1999) Ischemic preconditioning increases iNOS transcript levels in conscious rabbits via a nitric oxide-dependent mechanism. J. Mol. Cell Cardiol. 31, 1469 –1481 Szabolcs, M., Michler, R. E., Yang, X., Aji, W., Roy, D., Athan, E., Sciacca, R. R., Minanov, O. P., and Cannon, P. J. (1996) Apoptosis of cardiac myocytes during cardiac allograft rejection.

E1204

Vol. 20

September 2006

25. 26. 27. 28. 29.

30.

31.

32.

33.

34. 35. 36. 37.

38.

Relation to induction of nitric oxide synthase. Circulation. 94, 1665–1673 Schonhoff, C. M., Gaston, B., and Mannick, J. B. (2003) Nitrosylation of cytochrome c during apoptosis. J. Biol. Chem. 278, 18265–18270 Moncada, S., and Erusalimsky, J. D. (2002) Does nitric oxide modulate mitochondrial energy generation and apoptosis? Nat. Rev. Mol. Cell. Biol. 3, 214 –220 Razavi, H. M., Hamilton, J. A., and Feng, Q. (2005) Modulation of apoptosis by nitric oxide: implications in myocardial ischemia and heart failure. Pharmacol. Ther. 106, 147–162 Liao, X., Wang, X., Gu, Y., Chen, Q., and Chen, L. Y. (2005) Involvement of death receptor signaling in mechanical stretchinduced cardiomyocyte apoptosis. Life Sci. 77, 160 –174 Valenti, L., Mathieu, J., Chancerelle, Y., Levacher, M., Chanaud, B., De Sousa, M., Strzalko, S., Dinh-Xuan, A. T., Giroud, J. P., and Florentin, I. (2003) Nitric oxide inhibits spleen cell proliferative response after burn injury by inducing cytostasis, apoptosis, and necrosis of activated T lymphocytes: role of the guanylate cyclase. Cell. Immunol. 221, 50 – 63 Yoshioka, Y., Yamamuro, A., and Maeda, S. (2003) Nitric oxide at a low concentration protects murine macrophage RAW264 cells against nitric oxide-induced death via cGMP signaling pathway. Br. J. Pharmacol. 139, 28 –34 Wu, B., Iwakiri, R., Tsunada, S., Utsumi, H., Kojima, M., Fujise, T., Ootani, A., and Fujimoto, K. (2002) iNOS enhances rat intestinal apoptosis after ischemia-reperfusion. Free Radic. Biol. Med. 33, 649 – 658 Mungrue, I. N., Gros, R., You, X., Pirani, A., Azad, A., Csont, T., Schulz, R., Butany, J., Stewart, D. J., and Husain, M. (2002) Cardiomyocyte overexpression of iNOS in mice results in peroxynitrite generation, heart block, and sudden death. J. Clin. Invest. 109, 735–743 Sun, Y., Carretero, O. A., Xu, J., Rhaleb, N. E., Wang, F., Lin, C., Yang, J. J., Pagano, P. J., and Yang, X. P. (2005) Lack of inducible NO synthase reduces oxidative stress and enhances cardiac response to isoproterenol in mice with deoxycorticosterone acetate-salt hypertension. Hypertension 46, 1355–1361 Yamamoto, K., Dang, Q. N., Kelly, R. A., and Lee, R. T. (1998) Mechanical strain suppresses inducible nitric-oxide synthase in cardiac myocytes. J. Biol. Chem. 273, 11862–11866 Song, W., Lu, X., and Feng, Q. (2000) Tumor necrosis factoralpha induces apoptosis via inducible nitric oxide synthase in neonatal mouse cardiomyocytes. Cardiovasc. Res. 45, 595– 602 Arstall, M. A., and Kelly, R. A. (1999) The role of nitric oxide in heart failure. Coron. Artery Dis. 10, 301–308 Khadour, F. H., Kao, R. H., Park, S., Armstrong, P. W., Holycross, B. J., and Schulz, R. (1997) Age-dependent augmentation of cardiac endothelial NOS in a genetic rat model of heart failure. Am. J. Physiol. 273, H1223–H1230 Linke, A., Recchia, F., Zhang, X., and Hintze, T. H. (2003) Acute and chronic endothelial dysfunction: implications for the development of heart failure. Heart Fail. Rev. 8, 87–97

The FASEB Journal

Received for publication January 24, 2006. Accepted for publication April 21, 2006.

LIAO ET AL.

The FASEB Journal • FJ Express Summary

Nitric oxide signaling in stretch-induced apoptosis of neonatal rat cardiomyocytes Xudong Liao,*,†,1 Jun-Ming Liu,† Lei Du,* Aihui Tang,§ Yingli Shang,* Shi Qiang Wang,§ Lan-Ying Chen,†,2 and Quan Chen*,2 *State Key Laboratory of Biomembrane and Membrane Biotechnology, Institute of Zoology, Chinese Academy of Sciences; †Cardiovascular Institute and Fu Wai Hospital, Peking Union Medical College and Chinese Academy of Medical Sciences; and §State Key Laboratory of Biomembrane and Membrane Biotechnology, College of Life Sciences, Peking University. Beijing, China To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.06-5717fje SPECIFIC AIMS Cardiomyocyte apoptosis has been recognized as an important pathological factor in the transition from hypertensive heart remodeling to heart failure, but the mechanisms by which the mechanical signals lead to apoptosis are not well understood. Building on our previous finding that stretch-activated Ca2⫹ signaling plays an essential role in initiating apoptosis, we studied the downstream signals and found that NO was the link between stretch-induced Ca2⫹ signaling and apoptosis in cardiomyocytes. PRINCIPAL FINDINGS 1. Activation of Ca2ⴙ-dependent NO signaling in cardiomyocytes by mechanical stretch Mechanical stretch induced rapid and significant intracellular NO ([NO]i) elevation in cardiomyocytes (Fig. 1). [NO]i elevation remained significantly higher for up to 24 h after stretch and was sustained, although the initial [NO]i elevation was transient (Fig. 1). We reasoned that stretch-induced [NO]i elevation to activated NO synthases (NOS) since L-NAME, a nonisoformspecific NOS inhibitor, completely abrogated stretchinduced [NO]i elevation (Fig 1D). The rapid initial NO synthesis was Ca2⫹-dependent and significantly attenuated by EGTA-AM (Fig 1D), indicating a potential role for Ca2⫹-activated eNOS. The inducible NOS (iNOS) -specific inhibitor AMT failed to block stretch-induced initial NO synthesis, suggesting that iNOS was not responsible for the initiation of NO production (Fig 1D). Together with the basal expression levels of eNOS and iNOS in cardiomyocytes (Fig 1E), these data demonstrated that stretch-induced initial [NO]i elevation was due to Ca2⫹-dependent eNOS activation. Mechanical stretch also induced eNOS and iNOS expression in cardiomyocytes as revealed by Western blot (Fig 1E). Transient eNOS but prolonged iNOS inductions were further confirmed in vivo in hypertensive rat hearts 0892-6638/06/0020-1883 © FASEB

using an abdominal aorta constricted (AAC) hypertensive model. The stretch-induced prolonged induction of iNOS may have biological and pathological significance in hypertension. 2. Requirement of NO signaling in stretch-induced cardiomyocyte apoptosis To directly examine the role of NO in mechanical stretch-induced apoptosis, we pretreated cardiomyocytes with L-NAME 30 min before stretch to block NO synthesis and found that L-NAME blocked stretchinduced apoptosis (Fig 2A, B). In addition to L-NAME, a specific iNOS inhibitor AMT also inhibited apoptosis (Fig 2C). However, AMT failed to block iNOS induction or the initial phase of NO production, which suggested that the proapoptotic phase of NO was produced by iNOS. Blockage of soluble guanylyl cyclase (sGC) by ODQ inhibited stretch-induced apoptosis (Fig 2C), suggesting that the NO/cGMP pathway was involved in apoptosis regulation. All these inhibitors also blocked stretch-induced mitochondrial membrane potential (⌬⌿m) reduction and cyt c release, critical events in mitochondria-dependent apoptosis. These data indicated that endogenous NO signaling was required for stretch-induced cardiomyocyte apoptosis and that iNOS played important proapoptotic roles. 1

Present address: Department of Molecular Genetics, Lerner Research Institute, Cleveland Clinic Foundation, 9500 Euclid Ave., Cleveland, OH 44195, USA. 2 Correspondence: Q.C., State Key Laboratory of Biomembrane and Membrane Biotechnology, Institute of Zoology, Chinese Academy of Sciences, 25 Beisihuanxi Rd., Beijing 100080, P.R. China. E-mail: [email protected]; L.-Y.C., Cardiovascular Institute & Fu Wai Hospital, Peking Union Medical College & Chinese Academy of Medical Sciences, 167 Beilishi Rd., Beijing 100037, P.R. China. E-mail: lanyingchen @hotmail.com doi: 10.1096/fj.06-5717fje 1883

regulator of iNOS induction. When inhibiting the initial NO production by L-NAME, it completely blocked iNOS expression. NO regulated iNOS expression mostly likely through the sGC/cGMP pathway since ODQ blocked stretch-induced iNOS induction. There was a positive feedback loop in the stretchinduced NO cascade, in which stretch activated eNOS to initiate NO production and the initial NO induced iNOS expression to produce more NO. The high level of NO produced by iNOS, which is a much more powerful NOS than eNOS, was capable of inducing apoptosis and/or other stretch-induced responses in cardiomyocytes.

Figure 1. Mechanical stretch induces Ca2⫹-dependent NO production in NRVMs. A) Intracellular NO concentration ([NO]i) was examined by DAF-based confocal analysis. Images shown were taken from cells before stretch (Control), 5 min after stretch (Stretch), and 10 min after SNAP (100 ␮mol/l) treatment (SNAP). NO donor SNAP treatment was used as a positive control. B) Confocal images were analyzed with IDL software to give quantitative NO levels. Digital images were taken from random fields (⬎50 pictures per time point were taken) and pixel density was calculated, normalized with control data (from 10 min to 0 min before stretch) and changes of [NO]i (⌬[NO]i) were expressed as changes of fluorescence (⌬F/F0). Error bars stand for se. *P ⬍ 0.05. C) Total nitrate level ([NO]t) was determined by nitrate reductase-based colorimetric method. Conditioned medium was examined. [NO]t was converted to [NO3⫺] using KNO3 (100 ␮mol/l) as a standard. *P ⬍ 0.05. D) Cells were incubated with Ca2⫹ chelator EGTA-AM (2 ␮mol/l), NOS inhibitor L-NAME (100 ␮mol/l), or iNOS inhibitor AMT (100 ␮mol/l) for 30 min at 37°C before stretch. [NO]i was determined by DAF-based confocal analysis. Data of 10 min stretched cells were compared. *P ⬍ 0.05. E) Protein levels of iNOS and eNOS in NRVMs before and after 4 h stretch were detected by Western blot analysis. ␤-Actin was blotted to monitor equal loading.

3. Roles of eNOS and iNOS in NO cascade Although stretch induced both eNOS and iNOS, the initiating NOS is most likely eNOS since it produced the initial phase of NO; clearly the later phase of apoptotic NO was produced by iNOS, as discussed above. The next question we asked was what signal links eNOS to iNOS. We found that NO itself was the key 1884

Vol. 20

September 2006

Figure 2. NO signaling participates in the regulation of stretch-induced cardiomyocyte apoptosis. A) Apoptosis was examined with annexin V-based flow cytometry (upper panel) and DAPI-based fluorescent microscopy (lower panel). To block NO synthesis, cells were pretreated with L-NAME (100 ␮mol/l) at 37°C for 30 min before stretch. Cells were subjected to 20% sustained stretch for 4 h before apoptosis detection. Arrows indicated apoptotic nucleoli. B, C) Cells were incubated at 37°C for 30 min with L-NAME (100 ␮mol/l), AMT (100 ␮mol/l), ODQ (10 ␮mol/l), or SNAP (100 ␮mol/l), respectively, in each group before a 4 h stretch. Control cells were incubated with inhibitors without stretch. Apoptotic index was shown as % of fragmented nucleoli (indicated by arrows in panel A). More than 200 nuclei from randomly selected fields (⬎10 fields per dish) were counted. *P ⬍ 0.05.

The FASEB Journal

LIAO ET AL.

4. The role of Ca2ⴙ signaling in NO cascade We demonstrated that stretch activated a NO cascade to induce cardiomyocyte apoptosis. The next question we asked was how this cascade was initiated. We showed that stretch-induced initial NO production was Ca2⫹dependent in this study, and have reported that mechanical stretch-induced [Ca2⫹]i elevation plays essential roles in apoptosis initiation. Therefore, Ca2⫹ signaling is likely to be the initiator of the NO cascade. Stretch-induced [Ca2⫹]i elevation was dependent on Ca2⫹ influx through L-type calcium channels (LCC) and stretch-activated ion channels (SAC), and Ca2⫹induced Ca2⫹ release (CICR). The Ca2⫹ inhibitors (EGTA-AM, LCC blocker nifidipine, SAC blocker Gd3⫹, CICR inhibitors ryanodine and thapsigargin), which have been shown to block stretch-induced [Ca2⫹]i elevation in cardiomyocytes, significantly attenuated the initial [NO]i elevation in this study. These data suggest that the stretch-induced Ca2⫹ signal is required for the initial NO burst and possibly acts as a trigger for the NO signaling cascade by activating eNOS.

CONCLUSIONS AND SIGNIFICANCE In this paper, we suggest that mechanical stretch induces a Ca2⫹-NO signaling cascade to induce apoptosis in cardiomyocytes, as summarized in Fig. 3. Stretch may initiate the cascade through [Ca2⫹]i elevation and Ca2⫹-dependent eNOS activation. eNOS produces the initial NO, which is further amplified through NOinduced iNOS expression. iNOS produces high doses of NO, which are capable of regulating stretch-induced cardiomyocyte apoptosis and/or other stretch-induced responses. This stretch-Ca2⫹-NO signaling cascade may be a common mechanism by which cardiomyocytes sense and transduce mechanical stretch signals into biological signals to regulate hypertension responses, including, but not exclusive to, apoptosis. Mechanical stretch significantly induced the expression of both eNOS and iNOS in vitro and in vivo. The in vivo increase of iNOS expression was sustained in hypertensive hearts. These results strongly suggest that iNOS plays an important role in regulating the pathological outcomes of hypertension. Our data demonstrated that iNOS-produced NO was proapoptotic, since AMT blocked apoptosis without altering iNOS expression or eNOS-induced initial NO. Although eNOS induction was transient in vivo, its activation was essential in initiating the stretch-induced NO cascade because its inhibition resulted in blockage of iNOS gene expression. NO has been reported to be a bidirectional regulator of apoptosis. NO can directly induce the nitrosylation of some apoptosis regulatory molecules, such as caspase 3 and cyt c, or directly inhibit mitochondrial respiration

ENDOGENOUS NO SIGNALING TOWARD APOPTOSIS

Figure 3. Schematic summary of a hypothetical NO signaling cascade activated by mechanical stretch in cardiomyocytes. Mechanical stretch initiates the NO cascade by activating constitutively expressed eNOS through [Ca2⫹]i elevation. The active eNOS produces a small amount of NO, which further induces iNOS expression. iNOS, once expressed, produces a larger amount of NO, thus amplifying NO signaling. Finally, NO concentration reaches a certain level that is high enough to induce apoptosis and/or other responses in cardiomyocytes. NO may affect mitochondria or other apoptotic pathways to induce apoptosis. Besides this NO cascade, stretch-induced Ca2⫹ may participate in other signaling pathways. Solid lines indicate signaling discovered in this study. Dashed lines indicate possible signaling.

to induce apoptosis. Besides its role in the mitochondria-dependent apoptotic pathway, NO is also important in death receptor-dependent apoptotic signaling. We have reported that mitochondria and death receptor-mediated apoptosis were both activated by mechanical stretch in cardiomyocytes. It is possible that NO participates in both types of apoptotic pathways (mitochondria and death receptor-dependent) to regulate stretch-induced cardiomyocyte apoptosis (Fig. 3). Deregulation of both eNOS and iNOS has been associated with a number of heart diseases. It is possible that the activation of NO signaling is an adaptive response to protect cardiac cells from mechanical overload-induced damage. However, maladaptive responses with chronically high doses of NO and enhanced iNOS expression (and the potential feedback reinforcement) could be associated with the progression of heart diseases. Understanding the role of NO signaling in heart disease would be useful for the development of drugs that can selectively promote the beneficial effects of NO.

1885