Ginsenoside Rb1 Reverses H2O2-Induced Senescence in Human Umbilical Endothelial Cells: Involvement of Enos Pathway

ORIGINAL ARTICLE

Ginsenoside Rb1 Reverses H2O2-induced Senescence in Human Umbilical Endothelial Cells: Involvement of eNOS Pathway

Ding-Hui Liu, MD, PhD,* Yan-Ming Chen, MD, PhD,† Yong Liu, MD,* Bao-Shun Hao, MD,* Bin Zhou, MD,* Lin Wu, MD,* Min Wang, MD,* Lin Chen, MD,* Wei-Kang Wu, MD,‡ and Xiao-Xian Qian, MD, PhD*‡

INTRODUCTION Objective: Senescence of endothelial cells has been implicated in The process of cellular senescence was first described by endothelial dysfunction and atherogenesis. This study investigated Hayflick and Moorhead at early 1960s, in which they observed the effects of Rb1, a major ginsenoside in ginseng, on H2O2-induced that normal human fibroblasts were able to enter a state of senescence in primary human umbilical vein endothelial cells irreversible growth arrest, know as replicative senescence, after (HUVECs). serial cultivation in vitro. Replicative senescence is accompanied fi Methods and Results: Real-time PCR and Western blot were by a speci c set of changes in cell function, gene expression such as decreased endothelial-derived relaxing factors, increased used to detect the mRNA and protein expression, respectively. H2O2 (40;100 mmol/L) effectively increased SA-b-gal activity and PAI-1 endothelial-derived constricting factors, and increased production of molecules that are associated with inflammation and mRNA levels, two important senescence related biomarkers, in 1 HUVECs, which were dramatically inhibited by Rb1 pre-incubation. morphology. These cellular morphologic changes may contribute to age-associated diseases, including atherosclerosis. Over Furthermore, Rb1 administration reversed the H2O2-decreased pro- fi tein and mRNA levels of eNOS and its phosphorylation at Ser-1177, the past few decades, signi cant progress has been made in and the increased eNOS phosphorylation at Thr-495. As a result, our understanding of the mechanisms underlying cellular senes- Rb1 pretreatment restored the NO generation, as assayed by nitrate cence. One widely discussed hypothesis of senescence is reductase method. However, pretreatment with L-NAME, a NOS the telomere hypothesis that shortening of telomeres caused by attrition or altered protection may contribute to cellular inhibitor, abolished all the inhibitory effects of Rb1 on senescence. 2,3 Importantly, Rb1 modulated the H O -altered caveolin-1 and pAkt, senescence. However, cellular senescence can also be induced 2 2 by various kind of stress independently of the replicative age of two most important regulators of eNOS expression and activity, in 2 HUVECs. the cell. Therefore, based on this phenomenon, stress-induced premature senescence was developed to investigate the underly- Conclusions: We showed that Rb1 effectively protects HUVECs ing mechanisms involved in age-associated diseases. from senescence through eNOS modulation. With the development of modern technology, more and more active substances have been isolated and purified from Key Words: senescence, endothelial nitric oxide synthase, HUVEC, Ginseng, a tonic used in Chinese traditional medicine for over ginsenoside Rb1 2000 years. Among over 40 ginsenosides isolated from several (J Cardiovasc Pharmacol 2012;59:222–230) species of ginseng, Rb1 has gained most attention by research groups around the world. As a representative constituent from Panax ginseng and American ginseng,4 Rb1havebeenshowed to have a reversal effect on homocysteine-induced antiprolifer- Received for publication January 30, 2011; accepted October 7, 2011. ation of endothelial cell line by blocking superoxide anion pro- From the Departments of *Cardiology; †Endocrinology, The Third Affiliated duction5 and on homocysteine-associated vascular injuries in Hospital of Sun Yat-sen University, Guangzhou, Guangdong, China; and ‡Institute Integrated Traditional Chinese and Western Medicine, Sun Yat- porcine coronary arteries by decreasing superoxide anion gen- sen University, Guangzhou, Guangdong, China. eration and increasing endothelial nitric oxide synthase (eNOS) D.-H. Liu and Y.-M. Chen have contributed equally. abundance,6 suggesting that Rb1 protect vascular injury from Supported by the grants from Major State Basic Research Development oxidative stress through modulation of eNOS production. Program (grant number: 2005CB523305); Natural Science Foundation of Nitric oxide (NO) is recognized as a major antiathero- Guangdong Province (grant number: 8151008901000209); Science and Technology Planning Project of Guangdong Province (grant numbers: genic factor in the vasculature due to a number of vaso- 2007B060401024 and 2011B031800024); Science and Technology protective actions. Its adequate output is an important feature Planning Project of Guangzhou City (grant number: 2011B031800155). of a healthy endothelium. NO is stoichiometrically produced fl The authors report no con icts of interest. by eNOS through oxidating the guanidine group of L-arginine Reprints: Dr. Xiao-Xian Qian MD, PhD, Department of Cardiology, The Third Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong, in the presence of molecular oxygen and various cofactors. 510630, China (e-mail: [email protected]). Owing to the short half-life of NO, tight temporal and spatial Copyright © 2012 by Lippincott Williams & Wilkins regulation of eNOS including gene expression and reversible

222 | www.jcvp.org J Cardiovasc Pharmacol Volume 59, Number 3, March 2012 J Cardiovasc Pharmacol Volume 59, Number 3, March 2012 Involvement of eNOS Pathway

posttranslational modification is essential. The regulation of After H2O2 treatment, the culture media were collected to eNOS gene expression usually exhibits as epigenetic mecha- measure NO concentration; cells were washed with normal nisms, such as methylation of the eNOS promoter and deace- media once, harvested, and subjected to Western blot analy- tylation of associated histone proteins. And posttranslational sis, real-time polymerase chain reaction (RT-PCR), or SA- modification is mainly performed by eNOS phosphorylation b-gal activity assay as described in the following sections. at the site of Ser1177 and Thr495.7 Although as mentioned above, Rb1 has been suggested Senescence-associated to protect endothelial cell from vascular injury during b oxidative stimulus treatment, it is not known what kind of -galactosidase Assay b b vascular injury is caused by oxidative stimuli. Because Senescence-associated -galactosidase (SA- -gal) assay senescence is supposed to be involved in the pathogenesis was used to determine the number of senescent cells as 9 fl of atherosclerosis, we hypothesize that Rb1 can counteract described previously. Brie y, cells were washed twice in fi oxidative stress-induced senescence of endothelial cell phosphate-buffered saline, xed in 2% formaldehyde and 0.2% through modulating eNOS expression and activity. Here, glutaraldehyde for 5 minutes at room temperature, washed twice, b we established stress-induced senescent model in vitro by and stained with fresh SA- -gal staining solution (1 mg/mL of using primary human umbilical vein endothelial cells 5-bromo-4-chloro-3-indolyl-D-galactoside, 40 mmol/L of citric (HUVECs) to investigate the protective role of Rb1 in acid/sodium phosphate dibasic, pH 6.0, 150 mmol/L NaCl, cellular senescence. 2mmol/LofMgCl2, 5 mmol/L of K3Fe[CN]6, and 5 mmol/L of K4Fe[CN]6) for 18 hours at 37°C without CO2.Imagesof samples were taken under an inverted microscope with a ·100 fi fi MATERIALS AND METHODS magni cation. Senescent cells were identi ed as blue-stained cells, and a total of 400 contiguous cells were counted to Cell Culture determine the percentage of SA-b-gal–positive cells. HUVECs were isolated with 0.1% collagenase (Gibco) perfusion of term umbilical cord vein8 and maintained in Medium 199 (M199) (Gibco) [supplemented with 20% fetal Apoptosis bovine serum (FBS) (Gibco, Grand Island, NY), 20% serum- Morphologic changes in the nuclear chromatin in cells free medium (SFM) (Gibco), 2 mmol/l glutamine, 60 mg/ml undergoing apoptosis were detected by staining with Hoechst 33342 (Beyotime, Jiangsu, China) and 4,6-diamino-2-phenyl endothelial cell growth supplement (ECGS) (BD, San Diego, fi CA), and 5 U/mL heparin]. The cell type was characterized by indole (DAPI) (Aldrich Sigma). The cells were xed with 4% “cobblestone” cell morphology under phase-contrast micro- paraformaldehyde for 15 minutes at room temperature, washed fl twice with phosphate-buffered saline, and stained with Hoechst scope and by ow cytometric analysis with a monoclonal m anti-human CD31 antibody (BD). More than 99% of the cells at room temperature for 5 minutes or with DAPI (1 g/mL) at 30°C for 15 minutes. Images from Hoechst and DAPI-stained were positive for CD31. The protocol was approved by the · fl Research Committee of the Third Affiliated Hospital of Sun samples were acquired under a 10 objective lens using a uo- Yat-sen University. The parturients enrolled in the current pro- rescence microscope. tocol were negative for human immunodeficiency virus, hepatitis B virus, and hepatitis C virus and gave written informed Real-time Polymerase Chain Reaction consent to providing their umbilical cords. Total RNA in HUVECs was isolated using Trizol methods (Invitrogen, Carlsbad, CA) following the manufacturer’s instruction. The complementary DNA was generated by Induction of Premature Senescence of reverse transcript from messenger RNA (mRNA) using the HUVECs by H2O2 and Cell Treatment PrimeScript RT reagent Kit (TaKaRa, Dalian, China) as per For induction of premature senescence, the exponen- the manufacturer’s instruction. The SYBR Premix Ex Taq Kit tially growing HUVECs were inoculated at a cell density (TaKaRa) was then used for RT-PCR reaction. b-actin, of 1 · 105 per well in 6-well plate and cultured overnight at a housekeeping gene, was used as internal control for PAI-1 37°C. Cells were starved in M199 containing 2% FBS for at and eNOS mRNA expression. The sequences of PAI-1 (Gen- least 8 hours and then exposed to various concentrations of Bank No. NM_001165413) primers were 5’-TGC TGG TGA H2O2 for 60 minutes to induce premature senescence. For ATG CCC TCT ACT-3’ (forward) and 5’-CGG TCA TTC effects of Rb1 on H2O2-induced premature senescence, cells CCA GGT TCT CTA-3’ (reverse). The sequences of eNOS were pretreated with 10, 20, and 40 mmol/L Rb1 (purity $ (GenBank No. NM_001160110.1) primers were 5’-TGG TAC 98%, purchased from the Division of Chinese Materia Medica ATG AGC ACT GAG ATC G-3’ (forward) and 5’-CCA CGT and Natural products, National Institute for the Control of TGA TTT CCA CTG CTG-3’ (reverse). The sequences of Pharmaceutical and Biological Products, Ministry of Public b-actin (GenBank No. NM_001101.3) primers were 5’-AGC Health, China), respectively, for 30 minutes before H2O2 GGG AAA TCG TGC GTG AC-3’ (forward) and 5’-TCC treatment. For measuring the effect of eNOS inhibitor ATG CCC AGG AAG GAA GG-3’ (reverse). The relative L-NAME on Rb1’s treatment, cells were incubated with expression of eNOS or PAI-1 in each sample was normalized − − − 100 mmol/L L-NAME (Aldrich Sigma, St. Louis, MO) for against b-actin and shown as 2 [Ct(eNOS) Ct(beta actin)] or −[Ct(PAI−1)−ct(beta−actin)] 60 minutes and Rb1 for 30 minutes before H2O2 treatment. 2 .

2012 Lippincott Williams & Wilkins www.jcvp.org | 223 Liu et al J Cardiovasc Pharmacol Volume 59, Number 3, March 2012

Western Blot indicated by increased numbers of positively stained cells (as The protein was isolated from HUVECs using whole shown in Fig. 1C). The experiments were performed several times cell lysis buffer (KeyGen BioTechonology, Nanjing, China) and obtained similar results. according to the manufacturer’s instruction. We used another HUVECs senescence was also examined by SA-b-gal extraction recipe as follows: 0.125 M Tris-HCl (pH 6.8), staining assay. The proportion of positively stained cells was 5% (wt/vol) sodium dodecyl sulfate (SDS), 2.5% (vol/vol) calculated by a counting total of 400 contiguous cells. As b-mercaptoethanol, and 5% glycerol in double distilled water shown in Figure 1B, the senescent rates were 13.8%, to isolate protein for analyzing caveolin-1. Protein concentra- 17.8%, 26.8%, 49.2%, 59.8%, and 63.8% in cells treated with m tion in cell lysates was quantified by bicinchoninic acid 0, 20, 40, 60, 80, and 100 mol/L H2O2, respectively, show- (BCA) methods (Sangon BioTechonology, Shanghai, China). ing a dose-dependent effect. Compared with that of non- The same amount of protein in each sample was resolved treated cells, the senescent rates in all H2O2-treated cells electrophoretically by sodium dodecyl sulfate-polyacrylamide except for 20 mmol/L were significantly increased (P , gel electrophoresis (SDS-PAGE) and transferred onto a poly- 0.01). Meanwhile, total cell numbers in 60 mmol/L or less fi vinylidene difluoride membrane (Millipore, Billerica, MA). H2O2-treated cells were not signi cantly changed. Hence, m Membranes were blocked using 5% nonfat dry milk in tris- 60 mol/L H2O2 were used in subsequent experiments. m buffered saline tween (TBST) for 1 hour at room temperature To exclude the possibility that 60 mol/L H2O2 may with gentle shaking. eNOS was detected using a rabbit anti- lead to cell apoptosis, which would have an effect on our human eNOS polyclonal antibody (Cell Signaling Technol- hypothesis, we detected morphologic changes in the nuclear ogy, Danvers, MA) at the dilution of 1:200; the antibody of chromatin in the HUVECs treated with or without H2O2 by phosphorylated eNOS at T495 and S1177 (BD) and mono- Hoechst and DAPI. As shown in Figure 1D, treatment with m clonal antibody of PAI-1 (Santa Cruz, CA) were used at the 60 mol/L H2O2 did not cause apoptotic features such as dilution of 1:1000. The antibody of caveolin-1 (Cell Signal- morphologic changes and chromosomal condensation and ing Technology) was used at the dilution of 1:50,000. Signals fragmentation compared with control group although slightly were developed by the secondary HRP-conjugated antibody decreased cell numbers. and visualized by incubation with a chemiluminescent horse- fi radish peroxidase (HRP) substrate (Millipore) and quanti ed Rb1 Blocks H2O2-induced using Quantity One Software (Bio-Rad, Hercules, CA). HUVEC Senescence Pretreatment of HUVEC with Rb1 attenuated the increased number of senescent HUVECs induced by H2O2, Measurement of NO Production in with a maximum of approximately 2-fold reduction of Culture Medium senescent positive cell numbers at the concentration of The production of NO was measured as the generation of 20 mmol/L Rb1, compared with non–Rb1-pretreated group, nitrite, the stable metabolite of NO, in culture supernatants by as shown in Figure 1F. In parallel to the decreased positive nitrate reductase method using a NO assay kit (Nanjing staining of SA-b-gal, the cell numbers increased in Rb1- Jiancheng Institute of Biological Engineering, Nanjing, China). pretreated groups (data not shown).

Rb1 Promotes the Expression of eNOS Statistical Analysis in HUVECs Differences between groups were analyzed using 1-way To evaluate whether the reversal effect of Rb1 on analysis of variance. A P value ,0.05 was considered statis- HUVECs senescence is mediated through modulating eNOS, tically significant. Data are reported as mean ± SEM. we measured the protein and mRNA levels of eNOS and its phosphorylation. Compared with the nontreated cells, m fi RESULTS 60 mol/L H2O2 treatment of HUVECs signi cantly decreased eNOS phosphorylation at Ser1177 and increased Induction of Senescence the phosphorylation at Thr495, with a slight change in total To ascertain appropriate concentration of H2O2-inducing eNOS protein expression that exhibited a trend of decrease HUVEC senescence, we treated the cells with different concen- after the treatment of 24, 48, 72 hours (Figs. 2A–C). However, trations of H2O2 (as indicated in Fig. 1) for 60 minutes. Compared Rb1 administration reversed the H2O2-decreased protein levels with nontreated group, the numbers of attached cells decreased to of eNOS and its phosphorylation at Ser1177 and the increased 73.9%, 56.9%, 45.1%, and 37.3% when treated with 40, 60, 80, eNOS phosphorylation at Thr495. Of note, the highest reversal and 100 mmol/L H2O2, respectively, and slightly increased when rate of eNOS expression was achieved by treatment with treated with 20 mmol/L H2O2. Correspondingly, the cell numbers 20 mmol/L Rb1 despite for 24, 48, and 72 hours. in supernatants increased gradually compared with that of Because eNOS proteins changed from 24 to 72 hours nontreated cells. The percentages of unattached cells to attached after H2O2 treatment, we quantified eNOS mRNA by RT-PCR cells were 17% and 28% when treated with 80 and 100 mmol/L at the time point of 24 hours after different treatment. As H2O2, respectively. Moreover, cell viabilities measured by using shown in Figure 2D, 24-hour treatment with 60 mmol/L trypan blue staining decreased when treated with increased H2O2 H2O2 decreased eNOS mRNA by 3.43-fold compared with concentrations in both attached and floating cells, which was nontreated cells (P , 0.01). On the contrary, eNOS mRNA

224 | www.jcvp.org 2012 Lippincott Williams & Wilkins J Cardiovasc Pharmacol Volume 59, Number 3, March 2012 Involvement of eNOS Pathway

FIGURE 1. Cell morphological changes and SA-b-gal staining after H2O2 treatment and Rb1 pretreatment. A, Representative SA-b-gal staining pictures in HUVECs 24 hours after treatment with H2O2 for 1 hour. B, A total of 400 contiguous cells were counted to determine the percentage of SA-beta-gal positive cells were counted as the formula: the number of blue cells/the total counted cells. The results were represented as means ± SEM. **P , 0.01 versus control, n = 5. C, The numbers of attached and su- pernatant cells were detected as the indicated time point. D, Nuclear morphology was observed by Hoechst and DAPI staining. Left, control group; m right, 60 mol/L H2O2-treated group. E, Reversal effect of Rb1 on H2O2- induced senescence in HUVECs. SA-b-gal staining pictures in HUVECs in different concentrations of Rb1 pretreatment groups. F, Percentage of SA-b-gal–positive staining cell. Bars represent means ± SEM. ***P , 0.001 versus 60 mMH2O2 alone, n = 5. level increased by 2.79, 3.88, and 1.75-fold in 10, 20, and L-NAME Blocked the Preventive Effects of 40 mmol/L Rb1-pretreated cells, respectively, compared with Rb1 on H2O2-Induced Senescence in HUVECs cells treated with H2O2 alone (P , 0.01). As the control, Rb1 We then treated the cells with nitric oxide synthase alone did not significantly change the expression of eNOS at inhibitor, L-NAME, to test if blocking NO production could either mRNA or protein level. block the effect of Rb1 on HUVECs senescence. PAI-1 mRNA levels were measured by RT-PCR in Rb1-pretreated HUVECs with or without L-NAME presence. As shown in Changed PAI-1 Expression in Different Figure 4A, PAI-1 mRNA level was increased by 1.57-fold in Treated Cells cells treated with L-NAME compared with that in cells trea- To confirm the relationship between the senescence ted with Rb1 alone (P , 0.05). Similarly, PAI-1 protein and eNOS expression changes, we measured the senescent expression was also increased in cells treated with L-NAME biomarker, PAI-1 expression in different treated groups compared to that in cells treated with Rb1 alone. Taken (as shown in Fig. 3). Western blotting results demonstrated together, it is suggested that Rb1 prevent HUVEC senes- that the level of PAI-1 protein increased in H2O2-treated cence through increasing eNOS activity. group compared with untreated group after 24, 48, and 72-hour incubation. However, Rb1 reversed the increased expression of PAI-1 protein in HUVECs treated with NO Production H2O2. As the control, Rb1 alone did not significantly change As shown in Figure 4C, NO production was signifi- the expression of PAI-1. cantly decreased in cells treated with H2O2 after treatment

2012 Lippincott Williams & Wilkins www.jcvp.org | 225 Liu et al J Cardiovasc Pharmacol Volume 59, Number 3, March 2012

FIGURE 2. Effects of Rb1 on modulating of total eNOS, pS1177 eNOS, and pT495 eNOS protein levels after the treatment. A–C, Western blot results of total eNOS, pS1177 eNOS, and pT495 eNOS protein levels in indicated time point of 24, 48, 72 hours, respectively. The densito- metric data collected from 3 inde- pendent experiments at each time point. The ratio of pS1177 eNOS · total eNOS to pT495 was used to evaluate eNOS activity. **P , 0.01, *P , 0.05 versus 60 mMH2O2 alone. D, After pretreatment by different concentrations of Rb1 for 30 minutes, HUVECs were incubated with 60 mM H2O2 for 1 hour and then the medium was replaced. eNOS mRNA expression was measured using RT-PCR. The relative expression of eNOS in each sample was normalized against b- actin, shown as 2−[Ct(eNOS)−Ct(b−actin)]. Values were means ± SEM. **P , 0.01 versus 60 mMH2O2 alone, n = 3. of 24, 48, and 72 hours, respectively, compared with that in DISCUSSION the nontreated cells (P , 0.05). Pretreatment with 20 mmol/L Findings in experimental animals10,11 and observations Rb1 can effectively reverse H2O2-decreased NO production in humans support the conception that systemic and vascular (P , 0.05). However, pretreatment with L-NAME blocked oxidative stress develops with aging, even in healthy sub- NO production in cells pretreated with Rb1. jects.12,13 Aging is associated with a marked increase in risk of chronic vascular disease and partly caused by oxidative stress.14 Oxidative stress is believed to be an important pro- eNOS-related Regulation Protein: cess contributing to the development of chronic vascular dis- 15 Caveolin-1, pAkt ease. As an inducer of oxidative stress, H2O2 has been Akt and caveolin-1 have been considered as positive shown to induce oxidative stress–related cellular senescence and negative regulators, respectively, of eNOS activity and in vitro in a number of studies.16–18 Different cell types such expression. To elucidate their roles in the changes of eNOS in as fibroblasts and endothelial cells respond to H2O2 in a dose- our experimental system, we detected total Akt, total dependent manner. The fibroblasts are more resistant to 17 caveolin-1, and pAkt abundance in cells pretreated with or H2O2-induced senescence than endothelial cells. In endo- without Rb1 and treated with H2O2 by Western blot using thelial cells, the ideal H2O2 concentration for inducing senes- specific antibodies directed against caveolin-1, pAkt, and Akt, cence is from 30 to 100 mmol/L.18 Similarly, we found that 20 respectively. As shown in Figure 5A, pAkt was increased mmol/L H2O2 could not effectively induce endothelial cell after H2O2 treatment and this enhanced pAkt level was further senescence, even if it slightly increased cell number compared amplified by pretreatment with Rb1. Despite the changes in with that in the nontreated cells after 24 hours of culture. pAkt expression, the total Akt was not changed in different However, 40 ; 100 mmol/L H2O2 could effectively induce groups. On the other hand, 24 hours after treatment with HUVECs senescence in a dose-dependent manner as indi- H2O2 for 60 minutes, caveolin-1 expression was elevated, cated by SA-b-gal staining assay. Accompanied with the although this increased caveolin-1 was blocked by pretreat- increased senescent rate in groups treated with a high con- ment with Rb1, an effect that could be abolished by L-NAME centration of H2O2, the attached cell and the viable cell num- preincubation (Fig. 5B). bers decreased, suggesting that sublethal dose of H2O2

226 | www.jcvp.org 2012 Lippincott Williams & Wilkins J Cardiovasc Pharmacol Volume 59, Number 3, March 2012 Involvement of eNOS Pathway

to the reduction in NO production, which is produced by 3 nitric oxide synthase isoenzymes: inducible nitric oxide synthase (iNOS), neuronal nitric oxide synthase, and eNOS.19 Differences in biologic action among these isoenzymes reﬂect their differences in location and distribution. Under normal condition, iNOS is not expressed in endothelial cells, whereas eNOS is expressed in a modest way to produce adequate NO, maintaining the function of vessel. It is found that senescent endothelial cells produce less NO caused by the reduced expression of eNOS than normal cells.20,21 In agreement with the previous studies that reported the protective roles of Rb1 in vascular injury,5,6 we found that pretreatment with ginsenoside Rb1 protects the HUVECs from H2O2-induced senescence. Based on the previous study,6 we focus on whether Rb1 protect HUVECs from H2O2-induced cellular senescence through modulating eNOS expression. Interestingly, we found that the mRNA expression of eNOS obviously reduced at 24 hours after H2O2 treatment, whereas the total eNOS protein expression exhibited a trend of decrease from 24 to 72 hours after H2O2 treatment, suggesting that H2O2 modulate eNOS expression at both mRNA level and protein level. According to the study of Ota et al,22 the total eNOS expression should be sharply decreased in the immortalized HUVECs after 10 days of culture. We did not FIGURE 3. Rb1 act on the expression of senescent marker of perform the experiment at such time point because the cell PAI-1. HUVECs were pretreated with indicated doses of Rb1 for used in our study was primary cultured cell. These cells were 30 minutes before administration of 60 mMH2O2. After different time points, cells were then collected and lysated. The not quite resistant to environmental changes from medium expression of PAI-1 was evaluated by immunoblotting analysis. compared with immortalized cell. H2O2 is recognized to act Untreated cells were used as a control. Immunoblotting with as a bidirectional role in modulating eNOS expression for antibody speciﬁc for glyceraldehyde-3-phosphate de- a long time. Studies have shown that the administration hydrogenase (GAPDH) was done to show equal loading. A, 24 of H2O2 with endothelial cells increased eNOS activity and hours; B, 48 hours; C, 72 hours. expression by phosphorylating Ser1177 through PI3-kinase- Akt, inducing gene transcription, or elongating the half-life of treatment increased the number of growth-arrested cells and eNOS mRNA.23–26 Paradoxically, studies have also shown subsequently led to reduction in cell numbers. that H2O2 decreased eNOS activity and eNOS transcription The most of adult-onset cardiovascular diseases are in cultured endothelial cells.27–29 The peak for eNOS expres- associated with age-related vascular dysfunction and related sion was at 30 minutes after H2O2 stimulation. In our

FIGURE 4. L-NAME blocks NO production in Rb1-pretreated cells and elevates PAI-1 expression. The eNOS inhibitor L-NAME was added 30 minutes before Rb1 treatment and then Rb1 was added without replacing the medium. After that, the medium was replaced when cells m were incubated with 60 MH2O2 for 1 hour. A, mRNA expression was measured using RT-PCR analysis, and (B) protein expression was measured using Western blot. Values were means ± SEM. #P , 0.05 versus 60 mMH2O2 +20mM Rb1, **P , 0.01 versus H2O2 alone, n = 3. C, NO concentration was measured using NO assay kit according to the manufacturer’s instruction. Values were means ± SEM (n = 3). #P , 0.05 versus 60 mMH2O2 +20mM Rb1, *P , 0.05 versus H2O2 alone.

2012 Lippincott Williams & Wilkins www.jcvp.org | 227 Liu et al J Cardiovasc Pharmacol Volume 59, Number 3, March 2012

To determine whether the changes of eNOS level in endothelial cells were related to senescence phenotype, we measured PAI-1, a senescence-related biomarker reported recently.22,32 As expected, PAI-1 expression is increased in cells treated with H2O2, whereas decreased in cells pretreated with Rb1. To further confirm eNOS expression indeed affects senescence phenotype, we treated the cell with eNOS inhibitor, L-NAME, and found that L-NAME effectively blocked NO production in cells pretreated with or without Rb1. Meanwhile, the mRNA and protein levels of PAI-1 elevated after L-NAME treatment. These data strongly suggest that eNOS involves in the cellular senescence and Rb1 may obstacle the senescence progress through modulating eNOS pathway. Oxidative and antioxidative balance is a delicate procedure.33 It has been shown that the altered function of eNOS produced superoxide instead of NO in several conditions when the expression of eNOS monomers increased.34–37 To figure out whether the changed expression of eNOS eventually leads to the increase of NO production, we determined the NO production in the super- natant of cultured cells after different treatments. Indeed, FIGURE 5. Expression of Akt and caveolin-1. The eNOS inhibitor NO production was elevated in Rb1-pretreated cells com- L-NAME was added 30 minutes before Rb1 treatment and then pared with cells treated with H2O2 alone. Rb1 was added without replacing the medium. After that, the m Recently, both iNOS and eNOS are demonstrated to medium was replaced when cells were incubated with 60 M involve in the cellular senescence of human endothelial cells. H2O2 for 1 hour. After treatment for 24 hours, caveolin-1, Akt, 31 and pAkt were analyzed by Western blot as described in Yoon et al have shown in their experiments that NO produc- the “Methods” section. A, Expression of Akt. B, Expression of tion was produced largely by eNOS in young cells although was caveolin-1. produced by iNOS in replicative senescent endothelial cells. It is known that the increased iNOS expression is a compensatory mechanism in the condition of inadequate production of endo- 38 experiments, sublethal dose of H2O2 was used to incubate thelial NO. Moreover, in senescence-accelerated prone mice with HUVEC for 1 hour and then removed. It is assumed that (SAMP8), iNOS activity is elevated.39 Despite the elevated ac- H2O2 may increase eNOS expression at the early stage, tivity of iNOS in senescent cells or senescence-accelerated mice, but the feedback mechanism might be involved in and the overall NO production decreased in those cells and mice.31 limit the increased expression in eNOS protein. Furthermore, Consistently with the above reports, decreased NO production the research of Ohashi et al5 and Zhou et al6 confirmed and total eNOS protein and changed eNOS phosphorylation at that homocysteine causes vascular injury through elevating both Ser1177 and Thr495 were observed in H2O2-induced sen- H2O2 level and decreasing eNOS expression, suggesting endog- escent HUVECs in the present study. Taken together, it is enous H2O2 generation negatively correlates to eNOS expression clearly indicates that Rb1 prevents HUVECs from senescence during vascular injury. As a stimulus in our study, H2O2 is by elevating eNOS expression and activity. widely accepted as a mediator for causing homocysteine-induced Caveolin-1 is a membrane protein identified in the plasma vascular injury.30 Although this decreased eNOS expression after membrane caveolae in many cell types, including endothelial H2O2 treatment is reasonable, a further functional experiment cells and vascular smooth muscle cells. Evidence suggests investigating the relationship between endogenous H2O2 gener- caveolin-1 anchors eNOS to caveolae, thus limits the trans- ation and/or oxidative status andeNOSexpressionincellularor location and activation of eNOS. Thus, caveolin-1 regulates animal level should be taken in the future. vascular function through modulating eNOS activity.40,41 Sec- Despite the complicated mechanisms regarding eNOS ondary to loss of caveolin-1, the chronic activation of eNOS expression modulated by oxidation status, Rb1 antagonized leads to pulmonary hypertension.42 In response to oxidative H2O2-induced cellular senescence through modulating stress, caveolin-1 is phosphorylated at Tyr-14 and its expression mRNA expression and activity of eNOS. Activity of eNOS is upregulated.43,44 Our results suggest that the presence of Rb1 is regulated by posttranslational modifications. Phosphoryla- reverses the negative effect of caveolin-1 on modulating eNOS tion at Ser1177 significantly increases its activity, whereas activity. This in vitro result may provide supportive evidence phosphorylation at Thr495 decreases its activity.31 In our for the clinical investigation reported previously.45 studies, alterations of phosphorylation at Thr495 and Akt activation is triggered by various signals including Ser1177 induced by H2O2 treatment were reversed by Rb1, oxidative stress that activates upstream phosphatidylinositol which was shown the highest effect existed in the group of 3’-kinase, resulting in the activation of multiple downstream 20 mmol/L Rb1 pretreatment in the entire 3 time points, sug- effectors. Akt exerts its effects on senescence by elevating 46 gesting that Rb1 reversed the H2O2-induced senescence is not intracellular reactive oxygen species. However, Akt is also a dose-dependent manner in our study. a critical kinase for modulating eNOS activity and expression.47

228 | www.jcvp.org 2012 Lippincott Williams & Wilkins J Cardiovasc Pharmacol Volume 59, Number 3, March 2012 Involvement of eNOS Pathway

It is likely that Rb1 pretreatment further enhances pAkt induced 14. Taddei S, Virdis A, Ghiadoni L, et al. Age-related reduction of NO avail- by H O in senescent HUVECs, which leads to reversal of ability and oxidative stress in human. Hypertension. 2001;38:274–279. 2 2 15. Donato AJ, Eskurza I, Silver AE, et al. Direct evidence of endothelial senescence in this system, suggesting that pAkt plays a dual oxidative stress with aging in humans: relation to impaired endothelium- role through different pathways to accelerate or blockade dependent dilation and upregulation of nuclear factor-kB. Circ Res. cellular senescence. More interestingly, the phosphorylation at 2007;100:1659–1666. Ser1177 of eNOS showed an opposite change with pAkt in 16. Furukawa A, Tada-Oikawa S, Kawanishi S, et al. H2O2 accelerate cel- H O -treated cells. These unparalleled findings in our study lular senescence by accumulation of acetylated p53 via decrease in the 2 2 function of SIRT1 by NAD+ depletion. Cell Physiol Biochem. 2007;20: may result from the increased caveolin-1 expression. Decreased 45–54. phosphorylation at Ser1177 of eNOS may be generated as 17. Lee YA, Cho EJ, Yokozawa T. Protective effect of persimmon (Dio- 48 a result of tight coupling of eNOS and caveolae. Furthermore, spyros kaki) peel proanthocyanidin against oxidative damage under H2O2 phosphorylation of Ser1177 is only partially dependent on induced cellular senescence. Biol Pharm Bull. 2008;31:1265–1269. pAkt; adenosine monophosphate–activated protein kinase alfa 18. Oeseburg H, Insuf D, van der Harst P, et al. Bradykinin protects against 49,50 oxidative stress induced endothelial cell senescence. Hypertension. 2009; can phosphorylate eNOS at Ser1177 in some conditions. As 53(pt 2):1–6. shown in our study, these inconsistent changes of phosphory- 19. Aoyagi M, Arvai AS, Tainer JA, et al. Structural basis for endothelial lation (Ser1177) and pAkt may be weighed by the caveolin-1 nitric oxide synthase binding to calmodulin. EMBO J. 2003;22: – production and pAkt. The imbalance changes ultimately led to 766 775. 20. Matsushita H, Chang E, Glassford AJ, et al. eNOS activity is reduced in the decreased Ser1177 phosphorylation in H2O2-treated cells. senescent human endothelial cells. Preservation by hTERT immortaliza- tion. Circ Res. 2001;89:793–798. 21. Sato I, Morita I, Kaji K, et al. Reduction of nitric oxide producing CONCLUSIONS activity associated with in vitro aging in cultured human umbilical In conclusion, the eNOS is associated with the vein endothelial cell. Biochem Biophys Res Commun. 1993;195: senescent mechanism in HUVECs. Ginsenoside Rb1 protects 1070–1076. HUVECs from senescence by modulating eNOS activity and 22. Ota H, Akishita M, Eto M, et al. SIRT1 modulates premature senescence- expression. This modulation is correlated to caveolin-1 and like phenotype in human endothelial cells. J Mol Cell Cardiol. 2007;43: 571–579. pAkt, the 2 key regulators of eNOS. 23. Cai H, Davis ME, Drummond GR, et al. Induction of endothelial NO synthase by hydrogen peroxide via a Ca2+/calmodulin-denpendent protein kinase II/janus kinase 2-dependent pathway. Arterioscler Thromb ACKNOWLEDGMENTS Vasc Biol. 2001;21:1571–1576. The authors thank the members of our laboratories for 24. Cai H, Li Z, Davis ME, et al. Akt-dependent phosphorylation of serine their insight and technical support. 1179 and mitogen-activated protein kinase kinase/extracellular signal- regulated kinase 1/2 cooperatively mediate activation of the endothelial nitric-oxide synthase by hydrogen peroxide. Mol Pharmacol. 2003;63: REFERENCES 325–331. 1. Collado M, Blasco MA, Serrano M. Cell senescence in cancer and aging. 25. Drummond GR, Cai H, Davis ME, et al. Transcriptional and posttran- Cell. 2007;130:223–233. scriptional regulation of endothelial nitric oxide synthase expression by 2. Minamino T, Miyauci H, Yoshida T, et al. Endothelial cell senescence hydrogen peroxide. Circ Res. 2000;86:347–354. in human atherosclerosis: role of telomere in endothelial dysfunction. 26. Thomas SR, Chen K, Keaney JF, Jr. Hydrogen peroxide activates endo- Circulation. 2002;105:1541–1544. thelial nitric-oxide synthase through coordinated phosphorylation and 3. Karlseder J, Smogorzewska A, de Lange T. Senescence induced by dephosphorylation via a phosphoinositide 3-kinase-dependent signaling altered telomere state, not telomere loss. Science. 2002;295:2446–2449. pathway. J Biol Chem. 2002;277:6017–6024. 4. Cheng Y, Shen L, Zhang J. Anti-amnestic and anti-aging effects of 27. Wedgwood S, Black SM. Endothelin-1 decrease endothelial NOS ginsenoside Rg1 and Rb1 and its mechanism of action. Acta Pharmacol expression and activity through ETA receptor-mediated generation of Sin. 2005;26:143–149. hydrogen peroxide. Am J Physiol Lung Cell Mol Physiol. 2005;288: 5. Ohashi R, Yan S, Mu H, et al. Effects of homocysteine and ginsenoside 480–487. Rb1 on endothelial cell line proliferation and superoxide anion produc- 28. Kumar S, Sun X, Wedgwood S, et al. Hydrogen peroxide decreases tion. J Surg Res. 2006;133:89–94. endothelial nitric oxide synthase promoter activity through the inhibition 6. Zhou W, Chai H, Lin PH, et al. Ginsenoside Rb1 blocks homocysteine- of AP-1 activity. Am J Physiol Lung Cell Mol Physiol. 2008;295: induced endothelial dysfunction in porcine coronary arteries. J Vasc 370–377. Surg. 2005;41:861–868. 29. Kumar S, Sun X, Wiseman DA, et al. Hydrogen peroxide decreases 7. Schmitt CA, Dirsch VM. Modulation of endothelial nitric oxide by plant- endothelial nitric oxide synthase promoter activity through the inhibition derived products. Nitric Oxide. 2009;21:77–91. of Sp1 activity. DNA Cell Biol. 2009;28:119–129. 8. Bandin B, Bruneel A, Bosselut N, et al. A protocol for isolation and 30. Álvarez-Maqueda M, Beka RE, Monteseirín J, et al. Homocysteine culture of human umbilical vein endothelial cells. Nat Protoc. 2007;2: enhances superoxide anion release and NADPH oxidase assembly by 481–485. human neutrophils. Effects on MAPK activation and neutrophil migra- 9. Dimri GP, Lee X, Basile G, et al. A biomarker that identifies senescent tion. Atherosclerosis. 2004;172:229–238. human cells in culture and in aging skin in vivo. Proc Natl Acad Sci U S A. 31. Yoon HJ, Cho SW, Ahn BW, et al. Alterations in the activity and 1995;92:9363–9367. expression of endothelial NO synthase in aged human endothelial cells. 10. Madamanchi NR, Vendrov A, Runge MS. Oxidative stress and vascular Mech Ageing Dev. 2010;131:119–123. disease. Arteroiscler Thromb Vasc Biol. 2005;25:29–38. 32. Yokoi T, Fukuo K, Yasuda O, et al. Apoptosis signal-regulating kinase 11. Seals DR, Moreau KL, Gates PE, et al. Modulatory influences on ageing 1 mediate cellular senescence induced by high glucose in endothelial of the vasculature in healthy humans. Exp Gerontol. 2006;41:501–507. cells. Diabetes. 2006;55:1660–1665. 12. Hamilton CA, Brosnan MJ, Mclntyre M, et al. Superoxide excess in 33. Bernhard D, Wang XL. Smoking, oxidative stress and cardiovascular hypertension and aging: a common cause of endothelial dysfunction. diseases—do anti-oxidative therapies fail? Curr Med Chem. 2007;14: Hypertension. 2001;37:529–534. 1703–1712. 13. Francia P, delli Gatti C, Bachschmid M, et al. Deletion of p66shc gene 34. Durante W, Johnson FK, Johnson RA. Arginase: a critical regulator of protects against age-related endothelial dysfunction. Circulation. 2004; nitric oxide synthesis and vascular function. Clin Exp Pharmacol Phys- 110:2889–2895. iol. 2007;34:906–911.

2012 Lippincott Williams & Wilkins www.jcvp.org | 229 Liu et al J Cardiovasc Pharmacol Volume 59, Number 3, March 2012

35. Yang YM, Huang A, Kaley G, et al. eNOS uncoupling and endothelial 1-dependent perinuclear trafficking. JBiolChem. 2006;281: dysfunction in aged vessels. Am J Physiol Heart Circ Physiol. 2009;297: 14486–14493. 1829–1836. 44. Dai SM, Shan ZZ, Nakamura H, et al. Catabolic stress induces features 36. Zou MH, Cohen R, Ullrich V. Peroxynitrite and vascular endothelial of chondrocyte senescence through overexpression of caveolin-1. Possi- dysfunction in diabetes mellitus. Endothelium. 2004;11:89–97. ble involvement of caveolin-1-induced down-regulation of articular 37. Zou MH, Shi C, Cohen RA. Oxidation of the zinc-thiolate complex and chondrocytes in the pathogenesis of osteoarthritis. Arthritis Rheum. uncoupling of endothelial nitric oxide synthase by peroxynitrite. J Clin 2006;54:818–831. Invest. 2002;109:817–826. 45. Voghel G, Thorin-Trescases N, Farhat N, et al. Cellular senescence in 38. Jankowski J, van der Giet M, Jankowski V, et al. Increased plasma endothelial cells from atherosclerotic patients is accelerated by oxidative phenylacetic acid in patients with end-stage renal failure inhibits iNOS stress associated with cardiovascular risk factors. Mech Ageing Dev. expression. J Clin Invest. 2003;112:256–264. 2007;128:662–671. 39. Ali AK, Banks WA, Kumar VB, et al. Nitric oxide activity and isoenzyme 46. Nogueira V, Park Y, Chen CC, et al. Akt determines replicative senes- expression in the senescence-accelerated mouse P8 model of Alzheimer’s cence and oxidative or oncogenic premature senescence and sensitizes disease—effects of anti-amyloid antibody and antisense treatments. JGer- cells to oxidative apoptosis. Cancer Cell. 2008;14:458–470. ontol A Biol Sci Med Sci. 2009;64:1025–1030. 47. Levine YC, Li GK, Michel T. Agonist-modulated regulation of AMP- 40. Pojoga LH, Yao TM, Sinha S, et al. Effect of dietary sodium on vasocon- activated protein kinase (AMPK) in endothelial cells. Evidence for an striction and eNOS-mediated vascular relaxation in caveolin-1-deficient AMPK-Rac1-Akt-endothelial nitric oxide synthase pathway. J Biol mice. Am J Physiol Heart Circ Physiol. 2008;294:1258–1265. Chem. 2007;283:20351–20364. 41. Maniatis NA, Brovkovych V, Allen SE, et al. Novel mechanism of 48. Murata T, Kinoshita K, Hori M, et al. Statin protects endothelial nitric endothelial nitric oxide synthase activation mediated by caveolae inter- oxide synthase activity in hypoxia-induced pulmonary hypertension. nalization in endothelial cells. Circ Res. 2006;99:870–877. Arterioscler Thromb Vasc Biol. 2005;25:2335–2342. 42. Zhao YY, Zhao YD, Mirza MK, et al. Persistent eNOS activation 49. Cokic VP, Beleslin-Cokic BB, Tomic M, et al. Hydroxyurea induces secondary to caveolin-1 deficiency induces pulmonary hypertension the eNOS-cGMP pathway in endothelial cells. Blood. 2006;108: in mice and humans through PKG nitration. J Clin Invest. 2009;119: 184–191. 2009–2018. 50. Osuka K, Watanabe Y, Usuda N, et al. Modification of endothelial nitric 43. Khan EM, Heidinger JM, Levy M, et al. Epidermal growth factor oxide synthase through AMPK after experimental subarachnoid hemor- receptor exposed to oxidative stress undergoes Src- and caveolin- rhage. J Neurotrauma. 2009;26:1157–1165.

230 | www.jcvp.org 2012 Lippincott Williams & Wilkins