Puerarin Exerts a Delayed Inhibitory Effect on the Proliferation of Cardiomyocytes Derived from Murine ES Cells via Slowing Progression through G2/M Phase

Pueraria lobata (Willd) Ohwi (kudzu) has been utilized as a food source, medicine, and fodder in Southeast Asia for thousands of years [1]. Puerarin [7-hydroxy-3-(4-hydroxyphenyl)-1-benzopyran-4-one 8-(β-D-Glucopyranoside)] is the main isoflavone extracted from the root of it. The pharmacological properties of puerarin, as well as its beneficial and protective effects on cardiovascular diseases, have been widely investigated [2]. Puerarin exerts its myocardial protection effect by regulating cardiac Smad3 and Smad7 transcripts in spontaneously hypertensive rat [3], inducing vascular endothelial growth factor and endothelial nitric oxide synthase mediated angiogenesis in rats with myocardial infarction (MI) [4], inhibiting oxidative stress via regulation of superoxide dismutase activity in acute ischemic myocardial injury [5], or indirectly relieving myocardial ischemic damage by blocking the signaling transmission via P2X3 in SG and DRG [6].

Cardiomyocytes (CMs) derived from embryonic stem (ES) cells (ES-CMs) provide an alternative platform to study the mechanism of pathological remodeling, development and maturation of heart, and thus can serve as a relatively simple tool to investigate the detailed mechanism explaining pharmacological protective effects of puerarin on the heart. Previously, we reported that puerarin treatment suppresses the self-renewal of murine ES cells [7], and significantly enhances the population of functional murine ES-CMs [8]. Importantly, puerarin facilitates t-tubule development of these cells via up-regulating t-tubule development and formatting the related genes caveolin-3, amphiphysin-2 and junctophinlin-2 [9]. Accumulated data show that CMs are remodeled or "dedifferentiated" in pathological conditions, i.e. heart failure and other cardiovascular diseases, presenting enhanced proliferation ability and morphological and functional properties similar to embryonic CMs [10,11]. Based on the myocardial protection effects of puerarin reported previously, we hypothesized that puerarin might also inhibit CM proliferation, thus further support its potential effect on reversing cardiac phenotypic remodeling in heart diseases.

It is widely accepted that the proliferation of CMs mainly occurs during prenatal development [12], and essentially stops early after birth. After birth, CMs further grow in volume and maturity in physiological function. Contemporary studies mostly focus on the mechanisms regulating cell cycle activity of CMs in vitro and in vivo, which are significant for the cardiovascular regenerative medicine community. Humoral factors have been shown to upregulate CM proliferation via critical signaling pathways, such as PI3K [13,14], ERK [14], Npr3 and cAMP [15], and JAK/STAT [16] signaling pathways. In neonatal rat CMs, nuclear import of Cyclin D1 and cyclin-dependent kinase 4 (Cdk4) stimulates proliferation by activating the retinoblastoma regulatory pathway [17]. Inactivation of microRNA-34a enhances post-MI remodeling by dispelling its regulation on cell activity and apoptosis via target genes including Cyclin D1, Bcl2, and Sirt1 [18]. Thus CM proliferation, which plays a crucial role in in vitro cardiac differentiation [12], is related to several signaling pathways and regulated by cyclins.

Herein, we aimed to investigate the effects of puerarin on in vitro cardiac proliferation. This study may supplement the explanation of the myocardial protection effect of puerarin on reversing cardiac phenotypic remodeling in heart diseases.

The murine genetically modified D3 ES cell line αPIG (clone 44) was maintained as previously described [8,19]. Briefly, undifferentiated ES cells were cultured in high glucose Dulbecco's modified Eagle's medium (DMEM), supplemented with 15% fetal bovine serum (FBS), 1% nonessential amino acids, 2 mM L-glutamine, 100 U/mL penicillin, 0.1 mM β-mercaptoethanol, 100 µg/mL streptomycin and 1000 U/mL leukemia inhibitory factor (LIF, Chemicon). After removal of LIF, which inhibits the differentiation of ES cells [20], cardiac differentiation was performed by formation of embryonic bodies as shown in Fig. 1A. Puromycin (10 µg/mL, InvivoGen) was added at day 9 and day 12 of differentiation to collect purified ES-CMs clusters [19]. Puerarin (50, 100 or 200 µM, the National Institute for Food and Drug Control, China) or DMSO (0.05%, Sigma) was added from day 10 to the endpoint of the experiment (Fig. 1B).

ES-CMs clusters were collected at day 10 of differentiation, and dissociated into single cells using 0.05% Trypsin/EDTA (Biosharp). 5×10⁴ to 1×10⁵ cells per well were seeded in a gelatin-coated 24-well-plate with or without 100 µM puerarin supplemented with 10 µM EdU (5-Ethynyl-2'-deoxyuridine) from the Click-iT® EdU Imaging Kits C10338 (Invitrogen). After 48 h or 72 h cultivation, the staining of ES-CMs with EdU was performed according to the instructions of the manufacturer. ES-CMs were identified by staining with mouse IgG anti-α-sarcomeric-actinin antibody (1:200, Sigma) overnight at 4°C. The secondary goat anti-mouse FITC-conjugated IgG (1:200, Proteintec) was used for 1 h at room temperature. Nuclei were stained with Hoechst33342 (5µg/mL, Invitrogen). Images were taken randomly using a fluorescence microscope (Nikon, TE2000-S). The total number of ES-CMs was determined by counting α-actinin/Hoechst33342 double positive cells. The percentage of proliferating ES-CMs was determined by dividing the number of α-actinin/EdU/ Hoechst33342 triple positive cells by the total number of ES-CMs. The counting was carried out independently by three operators. The proliferation assays were performed by 5 to 6 independent experiments, and total 3002 of control cells and 3166 of puerarin treated cells were counted.

ES-CM clusters with or without puerarin treatment were collected at day 13 of differentiation, and dissociated into single cells using 0.05% Trypsin/EDTA. Cells were washed with cold PBS, and fixed in ice-cold 70% ethanol for 2 hours or longer at 4°C. The cells were then stained with propidium iodide solution prepared according to the instruc tions of the manufacturer (Beyotime Biotehnology). After incubation at 37°C for 0.5 h in the dark, the samples were analyzed for DNA content by Flow Cytometry Caliber instrument (BD Biosciences, LSR Ⅱ). Cell distribution of ES-CMs among cell cycle phases and the percentage of apoptotic cells were evaluated as previously described [21]. The cell cycle distribution is shown as the percentage of cells containing 2n (G1 phase), 4n (G2 and M phases), and 4n>3>2n DNA amount (S phase) estimated by PI staining. The apoptotic cells are defined by the percentage of cells with DNA content lower than 2n (sub/G1phase) [22].

Total RNA was extracted from ES-CMs using Trizol (Life technology) according to the instructions of the manufacturer at day 13 of differentiation. The total RNA (5 µg) was reversely transcribed into cDNA using oligo (dT) primer and mmlv reverse transcriptase (Invitrogen). Real-time PCR was performed in the CFX Connect™ Real-Time PCR Detection System (Bio-Rad) using SYBR Premix Ex Taq (TOYOBO). CT values were automatically obtained. Relative expression of mRNA amount was calculated using the ΔΔCT method [23]. The primers used for real-time PCR are listed in Table 1.

ES-CMs at day 10 of differentiation were pretreated in Opti-MEM (Invitrogen) for 2 h, and proteins were extracted after DMSO (0.05%) or puerarin (100 µM) treatment for different time periods (0 min, 10 min, 30 min, 1 h and 2 h). Protein concentration was determined using a BCA Protein Assay Kit (BOSTER). 25 µg protein samples were subjected to 12% SDS-PAGE and analyzed by blotting with antiphospho-ERK-1/2 MAb (1:2000, Cell Signaling Technology) or antiphospho-Ser⁴⁷³ Akt MAb (1:5000, Abcam) to detect the phosphorylation status of MERK/ERK and PI3K/Akt in ES-CMs.

Specific inhibitors were further used to investigate whether MERK/ERK and PI3K/Akt signaling pathways mediated the effect of puerarin on the transcription levels of cell cycle relevant genes. The PI3K inhibitor Wortmannin (1 µM, Sigma-Aldrich) or the EKR1/2 inhibitor PD0325901 (1 µM, Selleck) was added to differentiation medium, respectively. Total mRNA was extracted from cells at day 13 of differentiation, and the transcription levels of Cyclin A2, Cyclin B1 and Cdk1 were assessed by real-time PCR.

Unless otherwise stated, the results are expressed as mean ± SEM for at least three independent myocardium differentiations. Statistical significance was evaluated using paired or unpaired t-test where appropriate, p < 0.05 was considered as statistically significant.

CMs rapidly proliferate during fetal life but exit the cell cycle soon after birth in mammals, which limits the ability of the heart to restore function after any significant injury. To explore the effects of puerarin on the proliferation of CMs, ES-CMs at day 10 of differentiation were selected as study objects. The morphology of ES-CMs was comparable after cultivation with or without 100 µM puerarin for 48 h and 72 h, demonstrating that puerarin does not change the viability of ES-CMs (Fig. 2A). In the proliferation assay using EdU labeling, the percentage of proliferating ES-CMs was determined by dividing the number of α-actinin/EdU/ Hoechst33342 triple positive cells by the total number of ES-CMs. We found that puerarin treatment for 48 h did not significantly affect the proliferation of ES-CMs (23.44 ± 1.97% in puerarin treated group vs. 23.50 ± 2.46% in control group, n = 6, p = 0.3540) (Fig. 2A and B). However, after 72 h puerarin treatment, the percentage of proliferating ES-CMs was attenuated compared with control group (25.61 ± 1.20% in puerarin treated group vs. 31.28 ± 1.17% in control group, n = 5, p = 0.0096) (Fig. 2A and B), indicating that puerarin exerts a delayed inhibitory effect on the proliferation of ES-CMs at the early-stage differentiation.

To obtain further insight into the mechanism of proliferation inhibitory effect of puerarin, we assessed cell cycle distribution of ES-CMs by flow cytometry. Flow cytometric analysis showed that the majority of the ES-CMs were in G0/G1 phase in the puerarin treatment group (82.9 ± 1.5%) and the control group (84.4 ± 1.3%). In addition, puerarin treatment for 72 h increased the percentage of the cells in G2/M phase compared with control group (11.1 ± 0.5% in the puerarin treatment group vs. 8.6 ± 0.4% in the control group, n = 5, p = 0.0045) at day 13 (Fig. 3A and B). Moreover, there is no significant difference between the percentage of the apoptotic ES-CMs in the puerarin treatment group (2.2 ± 0.9%) and that in the control group (2.6 ± 1.2%) (n = 5, p = 0.8236) (Fig. 3A and C). These data suggest that with or without puerarin treatment, the majority of CMs are arrested in G0/G1 phase naturally. More importantly, puerarin slows the progression through G2/M phase without inducing apoptosis of ES-CMs.

The cell cycle, which is controlled by lots of cytokines and growth factors, plays an important role in proliferation of CMs [12]. We next analyzed the effect of various concentrations of puerarin on the transcription levels of cell cycle relevant genes in ES-CMs by real-time PCR. As shown in Fig. 4, the transcription of Cyclin A2 was dose-dependently up-regulated by puerarin treatment. The mRNA expression levels of Cyclin B1 and Cdk1 in the puerarin treatment group were significantly elevated by puerarin at 100 µM and 200 µM. The significant increases in transcription of Cyclin B2 and Cdk2 were also observed in 200 µM puerarin treatment group. Puerarin treatment did not change the transcription level of Cyclin D1.

To explore mechanism how puerarin affects transcription of cell cycle relevant genes, western blot analyses of ERK1/2 and PI3K/Akt signaling pathways were performed. We found that the total-ERK1/2 levels and total-Akt levels were comparable at all observed time points in both control and puerarin treatment groups. The phospho-ERK1/2 (Thr202/Tyr204, p-ERK1/2) and phospho-Akt (Ser473, p-Akt) were detected even after 2 h pretreatment with medium without FBS. We normalized the p-ERK1/2 and p-Akt levels at different time points by the p-ERK1/2 or p-Akt level at 0 min, respectively, in both groups. Compared with the 0 min group, the p-ERK1/2 level was significantly raised after 2 h treated with or without puerarin (p < 0.05 vs. 0 min group), and even greater in puerarin treatment group (Fig. 5A and B, p < 0.05 vs. control group). The p-Akt levels at all observed time points in control group were almost comparable to p-Akt level at 0 min. However, the p-Akt level was increased after 30 min of puerarin treatment (p < 0.05 vs. 0 min group or control group), and maintained at high level up to 2 h, the endpoint of study (Fig. 5A and C, p < 0.05 vs. 0 min group). Consistently, as shown in Fig. 6, we found that solo application of ERK1/2 specific inhibitor PD0325901 (PD, 1 µM) or the PI3K specific inhibitor Wortmannin (W, 1 µM) increased the mRNA expressions of Cyclin A2 and Cyclin B1, respectively, but had no effect on Cdk1. For this reason, we failed to block the effect of puerarin on the transcription of Cyclin A2 and Cyclin B1 by applying PD or W. Nevertheless, both PD and W successfully reversed puerarin-induced up-regulation of Cdk1 (Fig. 6).

The key findings of this study are: (1) puerarin exerts a delayed inhibitory effect on the proliferation of ES-CMs at the early-stage differentiation; (2) Puerarin slows the progression through G2/M phase without inducing apoptosis of ES-CMs; (3) puerarin up-regulates the transcription of Cyclin A2, Cyclin B1 and Cdk1 in ES-CMs; (4) the ERK1/2 specific inhibitor PD and the PI3K specific inhibitor W successfully reversed puerarin-induced up-regulation of Cdk1 but not Cyclin A2 and B1.

Before and after birth of mammals, the heart size increases through proliferation and hypertrophy in CMs, respectively. The cell cycle, which comprises four phases (G1, S, G2 and M), plays a key role in proliferation of CMs. Compared with the high proliferation level during early embryogenesis, the proliferation of mouse CMs starts to decrease significantly around embryonic day 10 to day 12 [12]. The cell cycle of almost all adult CMs is arrested in G1 phase [24], and only very few of the CMs can proliferate [25], which limits the cells' ability of self-repairing after damage. Considering in vitro myocardium differentiation is relatively delayed, we chose ES-CMs at day 10 of differentiation as study objects. We found that one third of early-stage ES-CMs keep proliferating, and puerarin treatment decreases the percentage of proliferating cells about 18.13% relative to control group. The majority of the early-stage ES-CMs were in G0/G1 phase. Meanwhile, puerarin treatment resulted in an elevated percentage of cells in G2/M phase without inducing cell apoptosis. Even though the variations in absolute values of the EdU and flow cytometry assays seem small, the effects of puerarin should not be ignored since the majority of the CMs are arrested in G1 phase naturally and only one third of early-stage ES-CMs keep proliferating in our study. Nevertheless, future work still need to be made in parallel between embryonic CM proliferation and excessive adult myocyte hypertrophy, to further state the effect of puerarin on CM proliferation under pathological condition. Multiple reports suggest that puerarin might have opposite effects on the cell cycle and the proliferation of different cells. It is reported that puerarin down-regulates the Cyclin D1 and cdc25A transcription by promoting the recruitment of corepressors to estrogen receptor-α, and limiting the recruitment of coactivators in endometriotic stromal cells, and therefore arrests these cells in G1 phase to suppress estrogen-stimulated proliferation without inducing cell apoptosis [26]. Other research found that puerarin exhibits antitumor activity on MCF-7 cells via arresting cells in G2/M phase and inducing apoptosis [27]. Conversely, puerarin decreases human osteoblastic MG-63 cells in G1 phase while increasing the proportion of cells in G2 and S phases, thus significantly promotes MG-63 cells proliferation in a dose-dependent manner [28].

The cell cycle is regulated by cyclins, Cdks and Cdk inhibitors [12]. Cdks are activated by forming complexes with cyclins, following which they phosphorylate complex-specific substrates and allow cells to move past cell cycle check points and finish the cell cycle. Cell cycle phases can be arrested at check points due to abnormal cell cycle events [29]. Cyclin B is a mitotic cyclin and binds to Cdk1. The amount of cyclin B and the activity of the Cyclin B-Cdk1 complex rise through the cell cycle until mitosis. After anaphase, they fall suddenly due to degradation of cyclin B [30]. In the cell cycle phases, Cyclin A2 binds Cdk1 during the transition from G2 to M phase. It has been confirmed that the Cyclin A2-Cdk1 complex triggers Cyclin B1-Cdk1 activation [31], which causes break down of nuclear envelope to allow mitotic spindle to access the chromosomes [32]. At anaphase, highly active anaphase-promoting complexes support the degradation of Cyclin B1-Cdk1 complex and thereby inactivate the Cdk1 which initiates mitotic exit and reestablishment of the interphase state [30,33]. Our study shows that the transcription levels of Cyclin A2, Cyclin B1 and Cdk1 are increased in puerarin treated ES-CMs. Taken together, we inferred that puerarin up-regulates the transcription of Cyclin B1 or activates the Cyclin B1-Cdk1 complex persistently, thus the progress through G2/M phase is delayed, and the proliferation of ES-CMs is inhibited.

Both ERK1/2 and PI3K/Akt signaling pathways are involved in cell cycle regulation. Activation of the PI3K/Akt pathway has been proved to inhibit embryonic CM proliferation [34], and to up-regulate Cdk1 in LNCaP cells treated with osteopontin-c [35]. Research on c-Met deficient hepatocytes showed that persistent ERK1/2 activation accounts for the up-regulation of Cdk1, which facilitates hepatocytes at the G2/M transition [36]. Our study found that the ERK1/2 and PI3K/Akt signaling pathways participate in puerarin-induced up-regulation of Cdk1. As regards increased transcriptions of CyclinA2 and CyclinB1 by puerarin treatment in the present study, further studies on the underlying mechanism are needed.

It should be noted that our study was in vitro, and did not examine effects of puerarin on pathological CMs in vivo. Although the present study and our previous study [9] suggest that the puerarin-induced inhibition of CM proliferation and promotion of CM maturity might attenuate the dedifferentiation process and heart remodeling, thereafter benefit the pumping function of the heart. Future work could be made in parallel between embryonic CM proliferation and excessive adult myocyte hypertrophy, to further study the effect of puerarin on CM proliferation under pathological condition and to better state the effect of puerarin on reversing excessive myocyte hypertrophy.

In summary, our study indicates that puerarin inhibits CM proliferation via slowing progression through G2/M phase during early-stage differentiation. This study does suggest new pharmacological effects of puerarin on ES-CMs, which may supplement the explanation of the myocardial protection effect of puerarin on reversing cardiac phenotypic remodeling in heart diseases.

This study was supported by the National Natural Science Foundation of China (No.31100828), the Natural Science Foundation of Hubei Province (2011CDB363), the Project-sponsored by SRF for ROCS Jiaoya Xi, SEM, the Fundamental Research Funds for the Central Universities (HUST: 2011QN211, 2012ZHYX019 and 2013TS145), the National Undergraduate Training Programs for Innovation and Entrepreneurship, HUST:201410487059, and the Fundamental Research Funds for the undergraduates' of Central Universities, HUST:2013B308.

None.

X. Luo and B. Zhong have equal contribution to this manuscript.