Introduction

Aging is associated with a number of diseases, including dementia, atherosclerosis, osteoarthritis, infection, and cancer [1]. These diseases are associated with inflammation and oxidative stress, likely to promote premature senescent cells formation [2, 3].

Premature senescence is described as a shortened inherent replicative lifespan in stress-exposed cells [2, 4]. Accumulating evidence supports the relationship between oxidative stress and aging, by showing that hydrogen peroxide (H2O2) induces cellular premature senescence [5,6,7]. Senescent cells normally exhibit permanent arrest, large flat morphology, and up-regulate senescence-associated β-galactosidase (SA-β-gal) activity [6]. Senescence is also associated with various altered protein expressions. The tumor suppressor p53 is an important transcription activator, and directly up-regulates the cyclin-dependent kinase (CDK) 2 inhibitor p21, a critical negative cell growth regulator, to induce cellular senescence [8, 9]. Caveolin-1 is a major caveolae membrane component and mediates cell cycle arrest through the p53/p21-dependent pathway [10]. Retinoblastoma protein (Rb) is another tumor suppressor shown to play a crucial role in senescence induction, and it is inactivated by cell cycle-dependent phosphorylation regulators such as cyclin D1 [11, 12]. In senescent human dermal fibroblasts (HDFs), p21 and p16 overexpression results in Rb hypophosphorylation [13]. Sirtuin 1 (SIRT1) is a nicotinamide adenine dinucleotide (NAD+)-dependent protein deacetylase involved in stress resistance and anti-senescence, through negative regulation of p53-mediated pathways [14]. Previous studies have shown SIRT1 inhibition promotes p53 acetylation at K373/382 by down-regulating p21 protein expression [14,15,16]. Akt is a well-known oxidative stress mediator [17]. The Akt pathway has been suggested to enhance p53 translation and protein stability, and promote p53 accumulation and downstream senescence [18].

Dalbergia odorifera T. Chen. (Leguminosa) possesses a wide range of biological activities, including anti-osteoporotic, anti-inflammatory, anti-oxidant, and neuroprotective effects, in diverse cell types [19,20,21]. In a previous study, 6,4′-Dihydroxy-7-methoxyflavanone (DMF), a naturally occurring flavonoid, isolates from the heartwood of Dalbergia odorifera (D. odorifera), and exhibits protective effects against glutamate-induced oxidative injury in HT22 cells [22]. However, the anti-aging effect of DMF remains unknown. This study therefore aimed at examining the anti-senescence effect of DMF and the associated mechanism, and whether this effect could protect H2O2- exposed HDFs against senescence.

Materials and methods

Preparation of DMF

DMF (no. NNMBP012) was obtained from the Standardized Material Bank, Wonkwang University (Republic of Korea) (Fig. 1a). DMF (> 98%) was isolated from D. odorifera as previously described [22]. DMF was dissolved in DMSO before used and the final vehicle content was less than 0.01% in each experiment.

Fig. 1
figure 1

The chemical structure of DMF (a) and its effect on viability (b). b HDFs were incubated for 12 h with different concentrations of DMF (20–320 μM). Cell viability was determined by MTT assay. All the data are presented as the means ± S. D. of 3 three independent experiments. *P < 0.05 versus control

Full size image

Reagents

H2O2 and 5-Bromo-4-chloro-3-indolyl-β-D-galactosidase (X-Gal) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Anti-acetylated-p53 (-ac-p53), -p16INk4α, -sirtuin 1 (-SIRT1), -phospho-Rb (-pRb), -Akt, -phospho-Akt (-p-Akt), -extracellular signal-regulated kinase (-ERK), -phospho-ERK (-p-ERK), -p38, -phospho-p38 (-p-p38), -c-Jun N-terminal kinase (JNK), and -phospho-JNK (-p-JNK) antibodies were purchased from Cell Signaling Technology (Beverly, MA, USA). Anti-p21Cip1/WAF1, -cyclin D1, and -β-actin antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). LY294002 and PD98059 were purchased from Caliches (San Diego, CA, USA). All other chemicals were obtained from Sigma-Aldrich (St. Louis, MO, USA).

Cell culture

Primary Human dermal fibroblasts (HDFs) were purchased from ATCC (ATCC PCS-201-010). Cells were maintained in low-glucose Dulbecco’s Modified Eagle’s medium (DMEM; GIBCO BRL, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS; GIBCO BRL) and 1% penicillin/streptomycin (GIBCO BRL) at 37 °C in a humidified incubator with 5% CO2. Cells at early passages (between 9 and 18 passages) were used in all experiments.

Drug treatment

The HDFs were seeded in 60 mm dishes or 6-well plates and cultured for 24 h. Then the cells were treated with different concentrations of DMF (20, 40, 80 μM) for 12 h. The DMF containing medium was then replaced with a new medium containing 200 μM H2O2 for 2 h. After 2 h, the cells were cultured in fresh medium for 72 h after removal of H2O2. The HDFs were maintained in complete medium (DMEM, 10% FBS, 1% penicillin/streptomycin) in all the experimental procedure.

Senescence-associated β-galactosidase (SA-β-gal) staining

SA-β-gal activity assay was performed at pH 6.0 as previously described [23]. Briefly, all the cells were washed twice with PBS, fixed in 3% formaldehyde for 5 min, and then stained at 37 °C with SA-β-gal staining solution (1 mg/ml X-gal, 40 mM citric acid/ sodium phosphate, pH 6.0, 5 mM potassium ferrocyanide, 5 mM potassium ferricyanide, 150 mM NaCl, 2 mM MgCl2). After 24 h, the stained cells were photographed and counted.

Western blot analysis

Western blot analysis was performed as Cheng et al. [24]. Briefly, all the cells were harvested, and lysed with RIPA buffer (50 mM Tris–HCl, pH 7.4, 150 mM NaCl, 2.5% deoxycholic acid, 1% NP-40, 10 mM EDTA) containing a protease inhibitor mixture (0.1 mM phenylmethylsulfonyl fluoride, 1 μg/ml chymostatin, 5 μg/ml aprotinin, and 5 μg/ml pepstatin A). Protein concentration was determined using a Bio-Rad protein assay kit. Lysates (50 μg/well) were separated by 12% or 14% SDS-PAGE and electro-transferred to PVDF membranes. Each membrane was incubated with primary antibody at 4 °C overnight, followed by 2 h in secondary antibody. Protein bands were visualized using the ECL system (Amersham Pharmacia Biotech, Piscataway, NJ).

Cell viability and cell proliferation by MTT assay

Cell viability and cell proliferation were performed as Lee et al. and Yu et al. described, respectively [25, 26]. Briefly, Cells (4 × 104/well) were seeded in 96-well plates and treated with DMF and H2O2 as described in figure legends (Figs. 1b and 2d). Then 20 μl of MTT (3-[4, 5-dimethythiazol-2-yl]-2,5diphenyltetrazolium bromide, a tetrazolium salt) test solution was added and incubated for 4 h at 37 °C in a humidified incubator with 5% CO2. The absorbance was measured at a wavelength of 490 nm using a SpectraMax M3 instrument (Molecular Devices, California, USA).

Fig. 2
figure 2

DMF protected cells against H2O2-induced premature senescence. a–c HDFs were pretreated with DMF (20–80 µM) for 12 h, incubated in 200 µM H2O2 for 2 h, and then incubated in fresh medium for 72 h. The expression of ac-p53, p53, p21Cip1/WAF1, p16Ink4α, pRb, Rb, cyclin D1, and β-actin were measured by Western blotting and quantified by densitometry based on immunoblot images (a). SA-β-gal staining was performed (b). The percentages of SA-β-gal positive cells were calculated from 5 fields of view (c). d HDFs were pretreated with different concentrations of DMF, exposed to 200 µM H2O2 for 2 h, and then cultured in fresh medium at different timepoints (6–72 h). The cell proliferation was assessed by MTT assay. All the data are presented as the means ± S. D. of 3 three independent experiments. *P < 0.05 versus control. #P < 0.05 versus H2O2-treated cells

Full size image

Reverse transcription-quantitative polymerase chain reaction (RT-qPCR)

HDFs were washed twice with PBS and the total RNA was extracted by easy-Blue™ kit, according to the manufacturer’s protocol. RNA concentration was read using a GeneQuant pro RNA/DNA calculator (Amersham Biosciences, Uppsala, Sweden). According to the manufacturer’s protocol, cDNA was synthesized with total RNA using a High-capacity RNA-to-cDNA kit (Applied Biosystems, Foster city, CA). The cDNA was mixed with SIRT1 primer (4331182) (Applied Biosystems; Thermo Fisher Scientific, Inc) and then performed using CFX96™ Real-Time System (Bio-Rad). Cycling conditions were performed as follows: preparation at 50 °C for 2 min, denaturation at 95 °C for 10 min, and followed by 40 cycles at 95 °C for 10 s and at 60 °C for 30 s. The data were analyzed using StepOne™ software.

SIRT1 activity assay

SIRT1 activity was performed using the SIRT1 fluorometric drug discovery kit (Enzo Life Sciences, New York, NY, USA), according to the manufacturer’s instruction. Briefly, HDFs were harvested and lysed with RIPA buffer. Lysates (50 μg/well) were incubated in SIRT1 assay buffer (100 μM Fluor de Lys-SIRT1 substrate, 5 μM TSA and 200 μM NAD+) for 45 min at 37 °C, the reaction was stopped by 2 × SIRT1 Developer (contains 2 M Nicotinamide) for 15 min at room temperature. Fluorescence of the samples were read using a SpectraMax M3 instrument with an excitation set to 355 nm and emission set to 460 nm.

Transfection with siRNA

Instructions for using SIRT1 siRNA and scrambled siRNA were provided by Santa Cruz Biotechnology (CA, USA). Cells were transfected with Sirt1 siRNA or scrambled siRNA (80 nM) for 18 h in Opti-MEM®I Reduced Serum Medium (GIBCO BRL) using Lipofectamine 2000 reagent according to the manufacturer’s protocol (Invitrogen, Carlsbad, CA, USA). Following transfection, the medium was changed, and the cells were used in other experiment.

Statistical analysis

All experiments were operated three times. The results were expressed as mean ± standard deviation (S. D.). One-way ANOVA analysis (R software) was used for data comparisons within multiple groups, with P < 0.05 considered as statistically significant.

Results

DMF pretreatment inhibits H2O2-induced cellular senescence

In this study, we evaluated the effect of DMF on viability of HDFs to determine the cytotoxic potential of it. Up to a concentration of 80 μM, no cytotoxic effects could be detected using MTT assay (Fig. 1b). To establish a senescent model, HDFs were exposed to 200 µM H2O2 for 2 h and subsequently cultured for another 72 h (Fig. S1). We found that HDFs presented senescent cellular morphology, meanwhile, ac-p53 and p21 Cip1/WAF1, senescence markers, significantly increased and reached maximal. Pretreatment with DMF (20–80 µM, 12 h) resulted in decreased ac-p53, p21Cip1/WAF1, and p16Ink4α expression (Fig. 2a). In contrast, pRb and cyclin D1 levels increased (Fig. 2a). Next, we examined the activity of SA-β-gal, a specific cytoplasmic senescent cells marker [23]. We found that DMF pretreatment of cells reduced cytosolic positive stain for senescence (blue) (Fig. 2b) and SA-β-gal positive cells percentage, concentration-dependently (Fig. 2c). Cell proliferation result showed that DMF pretreatment increased HDF growth rate, compared with H2O2 treatment only (Fig. 2d).

DMF treatment induces SIRT1 expression in HDFs

Since SIRT1, a longevity gene, plays a vital role in cellular senescence [16], and can be activated by many flavonoids [27], we examined DMF effect on SIRT1 expression and activity. HDFs were treated with different DMF concentrations for 12 h. DMF up-regulated SIRT1 expression and activity, concentration-dependently (Fig. 3a–c). Optimum SIRT1 induction was achieved at 80 µM DMF. Treating cells with DMF (80 µM) increased SIRT1 expression and activity, time-dependently (Fig. 3d–f). SIRT1 induction by DMF occurred within 3 h, reached maximum at 12 h, and then decreased at 24 h post treatment (Fig. 3d–f).

Fig. 3
figure 3

SIRT1 expression was up-regulated by DMF treatment in HDFs. a-c SIRT1 expression was measured 12 h after treatment with indicated concentrations of DMF. d-f Cells were treated with 80 µM DMF and harvested at indicated time points. The expression of SIRT1 and β-actin were measured by Western blotting and quantified by densitometry based on immunoblot images (a, d). SIRT1 mRNA was determined using RT-qRCR (b, e). The SIRT1 activity was performed using a SIRT1 fluorometric kit (c, f). All the data are presented as the means ± S. D. of 3 three independent experiments. *P < 0.05 versus control

Full size image

SIRT1 inhibition ceases the protective effect of DMF against H2O2-induced premature senescence

From the above-mentioned results, DMF reduced H2O2-induced HDF senescence and up-regulated SIRT1 expression. To investigate the role of SIRT1 in DMF protection against premature senescence, we suppressed SIRT1 in HDFs by transfecting SIRT1 siRNA and used scrambled siRNA as control. Efficient silencing of SIRT1 in HDFs was confirmed by Western blotting, and SIRT1 activity (Fig. 4a, b). We found that SIRT1 inhibition prevented DMF-mediated ac-p53, p21Cip1/WAF1, and p16Ink4α reduction and abolished its protective role on SA-β-gal positive cells percentage, while pRb and cyclin D1 expression levels decreased (Fig. 4a, c, d). These results demonstrated that DMF-induced SIRT1 expression played an important role in preventing H2O2-induced premature senescence.

Fig. 4
figure 4

SIRT1 inhibition abrogated DMF effect against H2O2-induced premature senescence. HDFs were pre-transfected with two siRNAs targeting SIRT1 or a scrambled siRNA (80 nM, 18 h) in the presence or absence of DMF (80 µM, 12 h) before H2O2 (200 μM, 2 h) exposure and then incubated in fresh medium for 72 h. a The expression of SIRT1, ac-p53, p21Cip1/WAF1, p16Ink4α, pRb, cyclin D1, and β-actin were measured by Western blotting and quantified by densitometry based on immunoblot images. b The SIRT1 activity was performed using a SIRT1 fluorometric kit. c SA-β-gal staining was performed. d The percentages of SA-β-gal positive cells were calculated from 5 fields of view. All the data are presented as the means ± S. D. of 3 three independent experiments. *P < 0.05 vs. control. #P < 0.05 versus H2O2-treated cells

Full size image

DMF treatment attenuates H2O2-induced premature senescence via its inhibitory effect on the activation of Akt signaling

Previous studies have shown that H2O2 activates various kinases, which are associated with cellular senescence [28, 29]. Next, we examined the role of H2O2 exposure on Akt, ERK, p38, and JNK activation by checking their phosphorylation levels in HDFs. Cells exposed to 200 µM H2O2 time-dependently increased Akt phosphorylation, which peaked around 3–6 h and slowly declined thereafter (Fig. 5a). ERK activation by H2O2 immediately increased by 0.25–0.5 h, and reduction thereafter (Fig. 5a). Conversely, p38 and JNK activation were not detected after H2O2 exposure (Fig. 5a). We speculated that activation of either Akt or ERK, or both Akt and ERK, might lead to H2O2-induced premature senescence in HDFs. Pretreatment of the cells with LY294002 and PD98059 inhibited Akt and ERK phosphorylation, respectively, we found that selective PI3 K inhibition by LY294002 dramatically reduced SA-β-gal positive cells percentage after H2O2 exposure, whereas no reduction was detected in PD98059-pretreated cells (Fig. 5b, c). Moreover, 80 µM DMF (15%) showed more effective protection than 20 µM LY294002 (53%) (Fig. 5c). To further investigate DMF inhibition effect on Akt activation, HDFs were pretreated with LY294002 (as a positive control) and DMF prior to 3 h of incubation with H2O2 (Fig. 5d). We found that DMF pretreatment of cells significantly inhibited Akt phosphorylation (Fig. 5d). These results suggested that DMF could attenuate H2O2-induced premature senescence by blocking the Akt pathway.

Fig. 5
figure 5

DMF reduced H2O2-induced premature senescence via its inhibitory effect on the activation of PI3 K/Akt pathway. a HDFs were exposed in 200 µM H2O2 all the time until harvested at indicated time points. The phosphorylation of Akt, ERK, p38, and JNK kinases were measured by Western blotting and quantified by densitometry based on immunoblot images. b, c HDFs were pretreated with PD98059 (10–20 µM, 30 min), LY294002 (10–20 µM, 30 min), and DMF (80 µM, 12 h) before H2O2 (200 µM, 3 h) exposure and then incubated in fresh medium for 72 h. SA-β-gal staining was performed (b). The percentages of SA-β-gal positive cells were calculated from 5 fields of view (c). d HDFs were pretreated with LY294002, and DMF before H2O2 (200 µM, 3 h) exposure. The p-Akt and β-actin were measured by Western blotting and quantified by densitometry based on immunoblot images. All the data are presented as the means ± S. D. of 3 three independent experiments. *P < 0.05 vs. control. #P < 0.05 versus H2O2-treated cells

Full size image

Discussion

Natural compounds with anti-aging activities can protect against the aging process and ameliorate age-related developmental diseases, through modulation of various mechanisms, including energy homeostasis, cellular metabolism, and stress resistance [30]. The potential of DMF, a class of flavonoid isolated from D. odorifera, as a natural chemical with anti-oxidant properties, has been demonstrated in various cells types. D. odorifera, an important traditional Chinese medicine, has been widely used to treat blood disorders, ischemia, swelling, necrosis, and rheumatic pain [31]. Extracts of this plant exhibited a wide range of bioactivities such as anti-allergic, anti-inflammatory, and anti-oxidant activities [22, 31]. In recent years, the study of DMF’s effect on cell signal transduction has attracted more attention. Im et al. have reported that DMF has anti-osteoporosis effect; inhibits nuclear factor kappa-B ligand-induced osteoclast differentiation [32]. It has also been reported to protect against glutamate-induced oxidative injury in HT22 cells [22]. In this study, we demonstrated that DMF protects HDFs from H2O2-induced premature senescent phenotype and senescence-associated molecular induction via up-regulation SIRT1 expression. DMF also inhibits Akt activation, characteristic of reducing HDF senescence (Fig. 6).

Fig. 6
figure 6

A schematic illustration summarizing the protective effect of DMF against oxidative stress (H2O2)-induced premature senescence in HDFs. Up-regulation of SIRT1 by DMF inhibited cellular senescence through p53/p21Cip1/WAF1 pathway and p16Ink4α/Rb pathway. Meanwhile, inhibition of PI3 K/Akt signaling pathway by DMF also prevented H2O2-induced premature senescence

Full size image

Because senescent cells are characterized by altered cell cycle arrest, giving rise to typical senescent phenotypes and altered gene expression [1], the effect of DMF was confirmed by cell proliferation, SA-β-gal staining, and expression of related molecules such as ac-p53, p21Cip1/WAF1, and p16 Ink4α, pRb and cyclinD1. In line with our results, Cilostazol reportedly protects endothelial cells from H2O2-induced cellular senescence via reduction of p53 acetylation and SA-β-gal positive cells [33]. Salidroside considerably reversed senescence-like phenotypes in oxidative stress and protected human fibroblast cells from premature senescence, by regulating the expression of senescence-related molecules expression such as p53, p21, and p16 [34]. DMF, Cilostazol, and Salidroside all protect cells from H2O2-induced cellular senescence via reducing SA-β-gal positive cells and abrogating H2O2-induced cell enlargement and flattened morphology. DMF and Cilostazol reduced H2O2-induced acetylation of p53.

In order to explore the mechanism by which DMF prevents premature senescence in HDFs, we investigated the possible involvement of SIRT1, the closest homolog of Sir2, shown to regulate numerous physiological process, including DNA repair, stress resistance, cellular survival, and senescence [35]. Recently, researchers have identified a significant array of phytochemicals from plants and these phytochemicals act as SIRT1 activator to improve various stress-induced senescence. For example, resveratrol, ginsenoside Rb1 protect the human endothelium from H2O2-induced oxidative injury and senescence, via Sirt1 activation [36, 37]. Our study showed that DMF treatment up-regulated SIRT1 expression and activity, concentration- and time-dependently, at a higher than basal level. Interestingly, SIRT1 expression reversed H2O2-induced premature senescence; however, SIRT1 inhibition induced premature senescence-like phenotypes with up-regulated p53 acetylation [16, 36]. This agrees with our results. We found that the SIRT1 siRNA transfection abolished the DMF anti-senescent effect. These results suggested that SIRT1 induction by DMF was essential to preventing H2O2-induced premature senescence.

The Akt and MAPK pathways are well-known mediators of cellular stress in a diverse array of cell types. These kinases play vital roles in senescence process. Akt pathway activation can induce senescence in human cells via mTORC1 and p53; however, its inhibition prevents H2O2-senescent phenotype and cell cycle arrest in normal human diploid fibroblasts [6]. MAPKs have three main pathways: ERK, p38 MAPK, and JNK pathways [17]. ERK-, activated by stress stimuli and oxidative stress, induced senescence correlates with β-galactosidase activity and induction of classical senescence-related genes, p16Ink4α, p53, p21, p14-p19ARF, and senescence-associated heterochromatin and DNA damage foci [38]. Our results showed that the Akt and ERK pathways were selectively activated by H2O2 in HDFs. HDFs treated with Akt kinase inhibitor LY294002, but not with a selective ERK inhibitor, had reduced SA-β-gal positive cell percentages after H2O2 exposure, which is consistent with the previous reports using other cells [6]. Recent studies have discovered that flavonoids can also prevent cellular senescence via Akt activation pathway inhibition. Consistent with our findings, Choi et al. reported that 7,8-dihydroxyflavone attenuates TNF-α-induced skin aging in Hs68 HDFs via Akt signaling pathway down-regulation [39].

In all, our work provides reliable evidence supporting the protective effect of DMF against cellular oxidative stress-induced senescence, through SIRT1 activation and Akt pathway suppression. This finding demonstrates that DMF has the potential to ameliorate the aging process and attenuate age-related diseases in humans.