Cordyceps cicadae induces G2/M cell cycle arrest in MHCC97H human hepatocellular carcinoma cells: a proteomic study

Background Cordyceps cicadae is a medicinal fungus that is often used for treating cancer. However, the anticancer mechanisms of C. cicadae are largely unknown. This study aims to investigate the anticancer mechanisms of C. cicadae against hepatocellular carcinoma cells in vitro using a proteomic approach. Methods Human hepatocellular carcinoma MHCC97H cells were treated with a water extract of C. cicadae (0, 100, 250, 500, and 1000 μg/mL) for 48 h and harvested for cell viability assays. The significant differences in protein expression between control and C. cicadae-treated cells were analyzed by two-dimensional gel-based proteomics coupled with matrix-assisted laser desorption ionization-time of flight mass spectrometry. Flow cytometry analysis was employed to investigate the cell cycle and cell death. The anticancer molecular mechanism was analyzed by whole proteome mapping. Results The water extract of C. cicadae (0, 100, 250, 500, and 1000 μg/mL) inhibited the growth of MHCC97H cells in a dose-dependent manner via G2/M phase cell cycle arrest with no evidence of apoptosis. Among the identified proteins with upregulated expression were dynactin subunit 2, N-myc downstream-regulated gene 1, heat shock protein beta-1, alpha-enolase isoform 1, phosphatidylinositol transfer protein, and WD repeat-containing protein 1. Meanwhile, the proteins with downregulated expression were 14-3-3 gamma, BUB3, microtubule-associated protein RP/EB family member 1, thioredoxin-like protein, chloride intracellular channel protein 1, ectonucleoside triphosphate diphosphohydrolase 5, xaa-Pro dipeptidase, enoyl-CoA delta isomerase 1, protein-disulfide isomerase-related chaperone Erp29, hnRNP 2H9B, peroxiredoxin 1, WD-40 repeat protein, and serine/threonine kinase receptor-associated protein. Conclusion The water extract of C. cicadae reduced the growth of human hepatocellular carcinoma MHCC97H cells via G2/M cell cycle arrest.

effectively prevent liver cancer, but the treatment of liver cancer is still difficult [3,4].
Chinese medicine (CM) has been widely used in conjunction with chemotherapeutic drugs for liver cancer treatment in China with positive outcomes [5]. Cordyceps is a genus of ascomycete fungi belonging to the Clavicipitaceae family. All Cordyceps species are endoparasitoids, and most of them parasitize insects and other arthropods. The Cordyceps genus includes nearly 400 species, and some of them have potential anticancer effects. Cordyceps sinensis is a medicinal fungus that has been used for cancer treatment in CM and Traditional Tibetan medicine since the 15th century [6,7]. The anticancer properties of C. sinensis via cancer cell apoptosis induction, proliferation inhibition, or both in various types of cancers, including leukemia, melanoma, Leydig tumor, breast cancer, and human hepatocellular carcinoma (HCC) have been investigated [8][9][10]. C. sinensis inhibited tumor metastasis in vivo [11,12]. Another well-known species, Cordyceps militaris, showed anticancer effects on leukemia, lung cancer, and breast cancer in vitro and in vivo [13][14][15][16][17], and its bioactive compound, cordycepin, exhibited cytotoxic and reactive oxygen species-generating activity in relation to cancer proliferation inhibition [18][19][20][21][22].
Cordyceps cicadae, another well-known medicinal mushroom that grows on larvae of Cicadae flammata, has been used in CM in China for the prevention and treatment of various diseases, including kidney disease, immune disease, and cancer [23][24][25]. C. cicadae exhibited immunoregulatory effects on human T lymphocytes and modulated the growth of mononuclear cells [26][27][28][29]. C. cicadae inhibited the growth of lung adenocarcinoma and melanoma in vivo and in vitro [30,31]. However, the anticancer mechanism of C. cicadae for liver cancer is still unknown.
In the last decade, proteomics has been widely used in medical studies for clinical biomarker identification, pathogenesis investigation, new drug discovery, pharmacological research, toxicological examination, and so on [32]. Most biological functions are transmitted via proteins such as enzymes, receptors, and structural components. Therefore, comprehensive proteomic analyses help us to understand the molecular modifications in physiological conditions [33]. This study aims to investigate the anticancer mechanisms of C. cicadae against HCC in vitro by two-dimensional gel-based proteomics coupled with matrix-assisted laser desorption ionization-time of flight (MALDI-TOF/TOF) mass spectrometry (MS), flow cytometry analysis, and proteome mapping.

Cell culture and reagents
The MHCC97H cell line was purchased from the Liver Cancer Institute of Fudan University (China). MHCC97H cells were cultured in DMEM (Gibco BRL, USA) supplemented with 10% fetal bovine serum (Gibco BRL) in a humidified incubator containing 5% CO 2 in air at 37°C, and subcultured with 0.25% trypsin-0.02% EDTA (Gibco BRL). A lyophilized hot water extract of wild-type C. cicadae (BioAsia Co., China) was dissolved in phosphate-buffered saline (PBS) and adjusted to a final concentration of 10 mg/mL.

Cell proliferation assay
The dose-dependent effect of C. cicadae on cell viability was assessed by the MTT assay. Briefly, suspended MHCC97H cells (1 × 10 5 cells/mL; 100 μL) were dispersed into the wells of 96-well microtiter plates. After 24 h of incubation, various concentrations of C. cicadae were added to each well and incubated for 48 or 72 h. Next, 10 μL of MTT solution (5 mg/mL 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide dissolved in c-DMEM) (USB Corporation, USA) was added to each well, and incubated for 3 h at 37°C. The MTT solution was then removed and the insoluble purple formazan crystals formed were dissolved in 50 μL of isopropanol in 0.1 M HCl (MERCK, Germany). The optical density (OD) of each well was measured using a Bio-Rad 550 Microplate Reader (Bio-Rad, USA) at 595 nm with a reference wavelength of 655 nm. The percentage of cell viability was expressed as (A treatment / A control ) × 100%.

Cell cycle analysis
The dose-dependent effect of C. cicadae on the cell cycle distribution was assessed by flow cytometry as described in our previous report [34]. Briefly, MHCC97H cells (1 × 10 5 cells/mL) were treated with various concentrations of C. cicadae (0, 100, 250, 500, and 1000 μg/mL) for 48 h, and the cells were then harvested, fixed in 70% ethanol (MERCK, Germany), and stored at −20°C for 24 h until further analysis. Next, the cells were washed twice with ice-cold PBS, and incubated with RNase and propidium iodide (PI) (Sigma-Aldrich, USA) for 30 min. The PI-stained cells were excited at a wavelength of 488 nm and emitted at a maximum wavelength of 617 nm. Acquisition of 10,000 events was chosen for measurement of the DNA cell cycle distribution using a COULTER XL Flow Cytometer (Beckman and Coulter, USA). The distribution of cells in the different cell cycle phases shown in the DNA histograms was analyzed using Becton Dickinson Cell Fit Software (BD, USA).

Sample preparation for proteomic analysis
MHCC97H cells were seeded in 100-mm culture dishes at 1 × 10 6 cells/dish (1 × 10 5 cells/mL), incubated overnight, and then treated with or without 500 μg/mL C. cicadae for 48 h. The cells were harvested by trypsinization, washed three times with PBS, and centrifuged in a Beckman Spinchron DLX (Beckman and Coulter) at 400 × g for 5 min. The cell pellet was resuspended in lysis buffer [1% Triton X-100 (USB Corporation), 25 mmol/L 4-

Two-dimensional electrofocusing and polyacrylamide gel electrophoresis
The 2-DE procedures were carried as described in our previous report with modifications [35]. The samples were examined in duplicate and 12 gels (6 for control cells and 6 for C. cicadae-treated cells) were used in total. For the first-dimensional electrophoresis of proteins, protein samples (150 μg) were mixed with 350 μL of rehydration buffer [9.5 mol/L urea, 2% CHAPS, 0.28% DTT, 1% immobilized pH gradient buffer with pH 3-10 (Bio-Rad), 0.002% bromophenol blue (Sigma-Aldrich)] and then rehydrated for 10 h before isoelectric focusing via the following program: (a) linear increase up to 500 V in 1 h; (b) held at 500 V for 2 h; (c) linear increase up to 10,000 V in 4 h; (d) linear increase up to 10,000 V in 3 h; (e) held at 10,000 V to reach a total of 90,000 V × h. Focused immobilized pH gradient (IPG) gel strips were equilibrated for 15 min in a solution [50 mmol/L Tris-HCl, pH 8.8 (Bio-Rad), 6 mol/L urea, 30% glycerol, 2% sodium dodecyl sulfate (SDS; USB Corporation), containing 20 mmol/L DTT], followed by incubation with the same buffer containing 20 mmol/L iodoacetamide (Sigma-Aldrich) for another 15 min. The second-dimensional separation was performed in 1-mm-thick 12.5% polyacrylamide gels by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) at 15-mA current for 30 min followed by 30-mA current for the rest of the analysis.

Image acquisition and analysis
After the electrophoresis, the gels were stained with SYPRORuby Protein Stain (Bio-Rad), scanned with a Molecular Imager PharosFX Plus System (Bio-Rad), and analyzed using PDQuest 8.0 (Bio-Rad) according to the following procedures: background subtraction, spot detection, and spot matching. The expression levels, expressed as the percentage volume (% vol), were exported. The relative intensities of spots were used for comparisons between the two groups and only those spots with a significant difference (>1.5-fold increase or decrease, P < 0.05) were selected for protein identification.

Protein identification
Spots with differential expression (P < 0.05) after 500 μg/mL C. cicadae treatment were sent to the Genome Research Centre (The University of Hong Kong, Hong Kong) for protein identification by MALDI-TOF/TOF MS (4800 MALDI TOF/TOF Analyzer; Applied Biosystems, USA) analysis after trypsin digestion. The identification was performed by Mascot peptide mass fingerprinting, which can identify proteins from the NCBInr database with taxonomy limited to Homo sapiens. Mascot reported the molecular weight search (MOWSE) score, which was calculated as −10 × log 10 (P), where P is the probability that the observed match is a random event. P was limited by the size of the sequence database being searched (limited by taxonomy), with conditions such as peptide fixed modification and variable modification and the settings for trypsin digestion. Each calculated value that fell within a given mass tolerance of an experimental value counted as a match. The threshold was accepted if an event occurred at random with a frequency of < 5%. In this study, a protein match with a score of > 71 was regarded as significant.

Statistical analysis
In cell proliferation assay, experiments were repeated three times, measurements within an experiment were done in six duplicate, and in cell cycle analysis, experiments were repeated three times. Data from three independent experiments were expressed as mean ± standard deviation (SD). The significance of differences in data was determined by Kruskal-Wallis one-way analysis of variance (ANOVA) followed by Dunnets post hoc test for determination of dose-response dependence. In proteomic analysis, six individual samples were prepared and run for 2D gel in each group (i.e., n = 6) the significance of differences in data was determined by a two-tailed Student's t-test with Levene's test for equality of variances. All analyses were performed using SPSS 17.0 (IBM, USA) with P < 0.05 as the significance level.

Results
Effects of C. cicadae on MHCC97H cell proliferation MHCC97H cells were grown in medium containing various concentrations of C. cicadae (0, 100, 250, 500, and 1000 μg/mL) for 48 or 72 h to investigate the effects of C. cicadae on cell proliferation. Dose-dependent inhibition of cell growth was observed with treatment of C. cicadae (P < 0.05) (Figure 1).
Effect of C. cicadae on the cell cycle distribution of MHCC97H cells Figure 2A shows the effects of C. cicadae (0, 100, 250, 500, and 1000 μg/mL) treatment for 48 h on the cell cycle phase distribution of MHCC97H cells. As shown in Figure 2B, treatment with C. cicadae (100, 250, and 500 μg/mL) for 48 h resulted in cell accumulation in G2/M phase at 23.8%, 30.2%, and 38.6%, respectively, compared with the control cells (19.8%), and decreased the cell percentages in G0/G1 phase by 48.2%, 43.9%, and 35.6%, respectively, compared with the control cells (53.3%). Treatment with C. cicadae at 100, 250, and 500 μg/mL for 48 h did not affect the cell population in S phase (27.9%, 25.8%, and 25.9%, respectively, compared with the control cells (26.91%). Furthermore, most cells (64.74%) were arrested in G2/M phase with the highest dose of 1000 μg/mL C. cicadae. The sub-G1 apoptotic fraction of cells showed no significant changes when C. cicadae (100-500 μg/mL) was added, suggesting that no apoptosis occurred during the period of examination.
CDK1/cyclin B complex is involved in the G2/M phase checkpoint by regulating the process of M phase [36]. Figure 3 shows that while the CDK1 protein level was decreased in MHCC97H cells after C. cicadae treatment, the cyclin B1 protein level was unaffected.
Western blot analyses of a few randomly selected proteins were performed for proteomic data validation. In agreement with the proteome changes identified by the proteomic analysis, the expression levels of Hsp27, peroxiredoxin 1, and WD 40 were upregulated in the cancer cells with C. cicadae treatment at 500 μg/mL Figure 1 The effects of C. cicadae on MHCC97H cell viability. The cell viability was tested by MTT assay. Data were expressed as mean ± SD (n = 3 separate experiments). **,***, P < 0.01, 0.001 vs. control in 48 h treatment group, and ##, ###, P < 0.01, 0.001 vs. control in 72 h treatment group, respectively. compared with the control cells ( Figure 5). These results were similar to those obtained in the proteomic analysis.

Discussion
Species in the genus Cordyceps can reduce cancer cell proliferation [14,15,[55][56][57][58], but only limited studies on liver cancer are available. Several polysaccharides isolated from C. cicadae exhibited antitumor and immunomodulatory properties [14,15,[55][56][57][58] but the mechanisms have not been fully explored. In the present study, we showed that C. cicadae inhibited the growth of the human HCC cell line MHCC97H via G2/M cell cycle arrest. In support  of the cell cycle measurements, several differentially expressed proteins for cell cycle regulation, mitosis, and protein synthesis were identified by the proteomic analysis.
Western blot analyses revealed a decrease in the level of CDK1, which is a G2/M checkpoint cyclin. The cell cycle measurements also indicated accumulation of cells in G2/M phase rather than induction of apoptosis with C. cicadae. To our knowledge, this is the first study to demonstrate that the anticancer mechanisms of C. cicadae involve G2/M arrest. Other studies showed that C. militaris could induce G2/M cell cycle arrest in human colon cancer HT-29 cells [37]. Figure 6 summarizes the magnitude changes in expression of the identified proteins with reference to the control cells. 14-3-3 gamma, BUB3, DCTN2, MAPRE1, GLRX3, and CLIC1 are thought to be involved in cell cycle regulation, G2/M checkpoint control, microtubule polymerization, and spindle organization during mitosis etc. Although their mechanisms related to the G2/M cell cycle arrest induced by C. cicadae were not clear, 14-3-3 gamma is important for cell cycle regulation and 14-3-3 gamma expression change-induced cell cycle deregulation has been implicated in cancer formation [37]. 14-3-3 gamma blocked premature mitotic entry into G2/M phase after DNA damage [59,60], while 14-3-3 gammaknockdown cells showed inhibition of cdc25C function, which induced overriding of both the incomplete S phase and the G2 DNA damage checkpoint [61]. Overexpression of 14-3-3 gamma in HCC patients was associated with extrahepatic metastasis and worse survival [62,63]. Our proteomic data showed that C. cicadae treatment reduced the expression of 14-3-3 gamma in the human HCC cell line MHCC97H, suggesting that the anticancer and G2/M phase arrest effects of C. cicadae could be attributed to 14-3-3 gamma expression inhibition.
BUB3, DCTN2, and MAPRE1 are involved in spindle checkpoint and mitosis regulation. BUB3 is a mitosis checkpoint protein for early stages of mitosis to prevent premature sister chromatid separation, missegregation, and aneuploidy via regulation of anaphase-promoting complex/cyclosome (APC/C) [38,64]. DCTN2 is involved in functions for the microtubule anchor dynactin complex, inactivation of the spindle checkpoint through kinetochore disassembly, and recruitment of cell cycle regulators to centrosomes [40,65]. MAPRE1 is involved in microtubule polymerization, microtubule dynamics control [39], and mitotic spindle function [66], and is dominantly expressed in poorly differentiated HCCs [67]. The expression changes of BUB3, DCTN2, and MAPRE1 in C. cicadae-treated MHCC97H cells indicate a mitosis-regulatory effect of C. cicadae, which may contribute to G2/M phase arrest.
GLRX3 maintains redox homeostasis in living cells [41]. In GLRX3-knockout mice, embryonic fibroblasts were arrested at G2/M phase [68]. CLIC1 is a nuclear chloride ion channel and involved in G2/M cell cycle regulation [42]. These findings may suggest a reason why the expression levels of GLRX3 and CLIC1 were changed in C. cicadae-treated MHCC97H cells.
ENTPD5 is an endoplasmic reticulum (ER) enzyme that hydrolyzes purine nucleoside diphosphates and is essential for liver functions, and ENTPD5-deficiency resulted in hepatocellular neoplasia [69]. Overexpression of ENTPD5 was observed in prostate cancer [43,70]. NDRG1 is a member of the NDRG gene family, and exhibited anticancer and metastasis-suppression effects in pancreatic cancer, colon  Protein score: generated by the MS identification system. 4 Expression Quantity were calculated by the PDQuest software, data expressed as mean ± SD. 5 Mr and pI: relative molecular mass (Mr) and isoelectric point (pI) generated by the MS system. The SMN complex plays an essential role in spliceosomal snRNP assembly in the cytoplasm and is required for pre-mRNA splicing in the nucleus [53,54] cancer, cervical and ovarian cancer, prostate cancer, and breast cancer [44,[71][72][73]. MHCC97H cells exhibited an 80% pulmonary metastatic rate when injected into nude mice [74]. HSPB1 is a subunit of HSP27, and suppressed cell apoptosis in MHCC97H cells via an NF-kappaB pathway [45]. The elevation of this protein indicated some resistance of the cancer cells, and could be responsible for the absence of apoptosis in the C. cicadae-treated cancer cells. ENO1, a glycolysis module protein, was overexpressed in HCC cells [46,75], and its increased level suggests that C. cicadae treatment increased the glycolytic activity of the cancer cells.
A reduction in PEPD is involved in collagen and GSH metabolism of C. cicadae. PEPD is a homodimeric iminodipeptidase that releases carboxy-terminal proline or hydroxyproline residues from oligopeptides, and is used as a marker of non-alcoholic liver fibrosis [47,76]. ERP29 works in the early secretory pathway in the ER, and overexpression of ERP29 resulted in higher expression of HSP27 in breast cancer cells [48]. The reduced expression of ERP29 is likely to be attributed to reduced ER stress with treatment of C. cicadae.
STRAP is involved in spliceosomal mRNP assembly in the cytoplasm and pre-mRNA splicing in the nucleus [53]. Overexpression of STRAP was reported in several cancers [53,77]. A STRAP-activated p53-related apoptosisinduction effect and higher expression of STRAP in drugtreated HCC cells were observed [54,78]. In the present study, the expression of STRAP was reduced, which may be related to the non-apoptotic effect of C. cicadae.
PRDX1 is a member of the peroxiredoxin family involved in redox regulation in cells, regulation of hydrogen peroxide signaling through its peroxidase activity, and a protein chaperone function. The aspect of whether the reduced expression of PRDX1 contributes to reduced survival of MHCC97H cells via failure in redox regulation against reactive oxygen species remains to be determined. Elevated expression of PRDX1 was found in various cancers [52,79,80].
The present study has identified novel molecular signatures in C. cicadae-induced cell cycle arrest and cell death in MHCC97H cells using proteomics. Among the Figure 5 Western blotting analysis of differential expressed proteins. With 500 μg/mL of C. cicadae treatment, the expression of Hsp27 was upregulated and Peroxiredoxin 1 and WD 40 were upregulated in the cancer cells as compared to control (500 μg/mL of C. cicadae).

Figure 6
Summary of the differential expressed proteins in response to C. cicadae treatment in MHCC97H. The proteins with increased expression level after 500 μg/mL C. cicadae treatment were plotted on the right hand side of axis, and proteins with decreased expression level were plotted on the left hand side. differentially expressed proteins identified, 14-3-3 gamma, BUB3, DCTN2, MAPRE1, GLRX3, and CLIC1 may play some significant roles in the C. cicadae-induced G2/M phase arrest in MHCC97H cells. The identified proteins will further enhance a molecular explanation for the cell cycle arrest process induced by C. cicadae. Further validation of these markers in gene knockout studies would greatly improve our understanding of the molecular mechanisms behind the liver cancer therapy potential of C. cicadae.