Cordycepin induced MA-10 mouse Leydig tumor cell apoptosis by regulating p38 MAPKs and PI3K/AKT signaling pathways

The p38 MAPKs play important roles in the regulation of balance between cell survival and cell death on the development of various cancers. However, the roles of p38 MAPKs regulating apoptotic effects on Leydig tumor cells remain unclear. In the present study, we showed that cordycepin (3′-deoxyadenosine) selectively induced apoptosis in MA-10 mouse Leydig tumor cells through regulating the p38 MAPK and PI3K/AKT signaling pathways. Cordycepin reduced viability in MA-10, TM4, and NT2/D1 cells, but not cause cell death of primary mouse Leydig cells on moderate concentration. Cordycepin increased reactive oxygen species (ROS) levels, which is associated with the induction of apoptosis as characterized by positive Annexin V binding, activation of caspase-3, and cleavage of PARP. Inhibition of p38 MAPKs activity by SB203580 significantly prevented cordycepin-induced apoptosis in MA-10 cells. Co-treatment with wortmannin or the autophagy inhibitor 3-methyladenine (3-MA) elevated levels of apoptosis in cordycepin-treated MA-10 cells. Moreover, cordycepin activated p53, p21 and TGFß; and downregulated CDK2. The antitumour activity of cordycepin-treated MA-10 cells was significantly distinct in severe combined immunodeficiency (SCID) mice in vivo. These results suggested that cordycein is a highly selective treatment to induce MA-10 cells apoptosis via p38 MAPKs signaling.


Cordycepin induced apoptosis in MA-10 mouse Leydig tumor cells. The MTT test was used
to investigate the effect of cordycepin on cell viability in primary mouse Leydig cells, MA-10 mouse Leydig tumor cells, and TM4 Sertoli tumor cells (Fig. 1a). After 24 h of cordycepin treatment (10 μ M to 1 mM), cell death was seen in MA-10 and TM4 tumor cells in a dose-dependent manner, but was only seen in primary mouse Leydig cells at the highest doses of cordycepin (500 and 1,000 μ M). To further test the effect of cordycepin on cell death in mouse Leydig cells, we used the LDH release assay and western blotting. Treating primary mouse Leydig cells with 100 μ M cordycepin did not release a significant amount of LDH compared to control (Fig. 1b). Western blot analysis showed that cordycepin upregulated cleavage of caspase-3 and PARP in MA-10 cells (p < 0.001), but did not significantly affect levels of cleaved proteins in primary mouse Leydig cells (p > 0.05; Fig. 1c,d). These results indicate that 100 μ M cordycepin did not cause cell death in primary mouse Leydig cells. However, it should be noted that high concentration of cordycepin (1,000 μ M) did induce cell death in primary mouse Leydig cells.
Pluripotent embryonal carcinomas (EC) are the malignant counterparts to embryonic stem cells and are considered the stem cells of testicular germ cell tumors (TGCT) 31 . The data indicated that cordycepin significantly reduced cell viability in NT2/D1 cells as etoposide effects ( Figure S1A). NT2/D1 cells were treated with cordycepin for 24 h, and upregulated protein levels of cleavage of caspase-3 and PARP were assessed by western blotting ( Figure S1B). The G2M and subG1/G0 phase cells number were markedly increased among cordycepin-treated NT2//D1 cells, as shown in Figure S1C and S1D. These observations suggest that cordycepin significantly suppressed the viability of testicular cancer cells, including testicular germ cell tumors, MA-10 mouse Leydig tumor cells, and TM4 Sertoli tumor cells.
To confirm the cordycepin-induced caspase signaling cascades in MA-10 cells, the following caspase inhibitors were tested: Z-VAD-FMK (general caspase inhibitor), Z-IETD-FMK (caspase-8 inhibitor), and Z-LEHD-FMK (caspase-9 inhibitor). MA-10 cells were treated with caspase inhibitors for 1 h and then cordycepin (100 μ M) was added and cells were treated for an additional 12 h. The general caspase inhibitor Z-VAD-FMK significantly attenuated cleavage of caspase-3, -8, and -9 (p < 0.05; Fig. 2e), and specific inhibitors also affected levels of the appropriate cleaved caspases (p < 0.05; Figure S2A,B). Moreover, the MTT assay was used to determine whether the general caspase inhibitor could rescue the viability of MA-10 cells treated with cordycepin. Cordycepin (100 μ M) reduced MA-10 cell viability to 37% (relative to control) after 24 h of treatment (p < 0.001), but only Z-VAD-FMK improved the viability of cordycepin-treated MA-10 cells by 14% (p < 0.01; Fig. 2f and Figure S2C). Results also show that caspase-3 siRNA significantly reduced cleaved caspase-3 expression and recovered cell viability with cordycepin treatment in MA-10 cells (p < 0.05; Figure S2D,F). These results suggest that activation of caspase pathways was important for cordycepin to induce apoptosis in MA-10 mouse Leydig tumor cells.
Cordycepin induced apoptosis by activating p38 signaling. The caspase signaling pathways involved in MAPKs activate apoptosis 1 . We therefore examined the effect of cordycepin on phosphorylation of ERK, JNK, and p38 by western blotting. A high dose of cordycepin (1 mM) promoted phosphorylation of ERK with the optimal effects seen after 5 min, 15 min, 6 h, 12 h or 24 h, while intermediate time points (30 min, 1 h, or 3 h) exhibited minor effect (p < 0.05; Figure S3A,B). Although there is no statistical difference ( Figure S3B), an increasing trend of phosphorylated ERK blotting could be observed by 100 μ M cordecypin at 6 hr treatment ( Figure S3A). Moreover, treatment with 100 μ M cordycepin for 12 h or 1 mM cordycepin for 6, 12, or 24 h induced phosphorylation of JNK (p < 0.05), and treatment with 100 μ M cordycepin for 12 h or 1 mM cordycepin for 12 or 24 h induced phosphorylation of p38 (p < 0.05; Figure S3A,B).
To confirm whether MAPK signaling pathways were directly activated by cordycepin, MA-10 cells were pre-treated with inhibitors of ERK (PD98059), JNK (SP600125), and p38 (SB203580) 1 h, then 100 μ M cordycepin was added and cells were treated for an additional 12 h. Addition of inhibitors significantly reduced cordycepin-induced elevation of p-ERK, p-JNK, and p-p38 (p < 0.05; Fig. 3a and Figure S3C,D). To elucidate which MAPK signaling mediated cordycepin-induced apoptosis, the MTT assay was performed to determine cell viability. Treating MA-10 cells with 100 μ M cordycepin alone for 24 h reduced cell viability to 40% of control. Inhibition of p38 partially restored the viability of cordycepin-induced cell death in MA-10 cells (p < 0.01; Fig. 3b), but not JNK or ERK inhibitors ( Figure S3E).
To determine whether activation of p38 correlates to cordycepin-induced MA-10 cell apoptosis, we labeled cells with annexin V and PI and used flow cytometry. MA-10 cells were pre-treated with inhibitors of general caspase (Z-VAD-FMK) and p38 (SB203580) 1 h, then 100 μ M cordycepin was added and cells were treated for an additional 24 h. Treating MA-10 cells with Z-VAD-FMK and SB203580 did not increase a significant amount of apoptotic cells and change morphology compared to control (p > 0.05; Fig. 3c,d and Figure S3F). The general caspase inhibitor significantly reduced amount of apoptotic cells associated with 100 μ M cordycepin treatment, but only pretreatment with p38 inhibitor restored control levels of apoptotic cells (p < 0.001; Fig. 3c,d). Western blotting for cleaved caspase-3 and PARP demonstrates that the target of cordycepin-induced caspase activation is attenuated when the p38 inhibitor SB203580 is added 1 h prior to cordycepin treatment (p < 0.001; Fig. 3e,f). These data indicate that cordycepin exerted its effect in cordycepin-induced MA-10 cell apoptosis through p38 MAPKs signaling.

Cordycepin induced apoptosis and autophagy by regulating PI3K/AKT/mTOR signaling.
PI3K/AKT/mTOR signaling pathways and related signaling networks could promote cell growth and survival 33 . To investigate whether cordycepin could regulate PI3K/AKT/mTOR pathways to induce apoptosis in MA-10 tumor Leydig cells, we first examined levels of AKT and mTOR by western blotting. Treatment of MA-10 cells with 100 μ M cordycepin for 1 or 3 h or 1 mM cordycepin for 30 min to 12 h significantly reduced phosphorylation of AKT (p < 0.05; Fig. 4a,b). Similar regimens of cordycepin treatment also reduced phosphorylation of mTOR in MA-10 cells (p < 0.05; Fig. 4a,b). However, levels of p-AKT were elevated (relative to controls) after 24 h of 1 mM cordycepin treatment, whereas p-mTOR levels were elevated after 12 or 24 h of 100 μ M cordycepin treatment or 24 h of 1 mM cordycepin treatment. These results demonstrated a bimodal response to cordycepin treatment in PI3K/Akt/mTOR signaling.
To determine whether PI3K signaling pathways contribute to cordycepin-induced apoptosis of MA-10 cells, cells were pretreated with the PI3K/AKT inhibitor wortmannin (1 μ M) for 1 h prior to cordycepin (10 μ M to 1 mM) treatment, and cells were treated for an additional 24 h. The PI3K/AKT inhibitor significantly reduced levels of p-AKT associated with 100 μ M or 1 mM cordycepin treatment, as well as levels of p-mTOR associated with 1 mM cordycepin treatment (p < 0.05; Fig. 4c,d). Interestingly, treatment with wortmannin and cordycepin for 24 h increased expression of cleaved caspase-3 (p < 0.05; Fig. 4c,d). The MTT assay was therefore used to determine whether PI3K/AKT signaling affected the viability of MA-10 cells treated with cordycepin. The PI3K/AKT inhibitor further reduced the viability of cells in the presence of 10 μ M cordycepin for 24 h (p < 0.01; Fig. 4e). Moreover, we have published that  blocking phosphorylation of AKT by wortmannin further decreased cordycepin-induced cell viability 24 , suggesting the involvement of PI3K/Akt pathway in the regulation of cordycepin-induced apoptosis in MA-10 cells.
After MA-10 cells were treated with 1 mM cordycepin for 12 or 24 h, LC3 II upregulation was detected (p < 0.05; Fig. 4a,b), suggesting that autophagy was involved in this apoptotic response. The MTT assay was therefore used to determine whether autophagy affected the viability of MA-10 cells treated with cordycepin. The autophagy inhibitor 3-methyladenine further reduced the viability of cells in the presence of 100 μ M cordycepin for 24 h (p < 0.01; Fig. 4F). Thus, the inactivation of PI3K/AKT/ mTOR signaling pathways and upregulation of LC3 II, which is related to autophagy, may be involved in cordycepin-induced apoptosis of MA-10 cells.
Cordycepin led to ROS accumulation. We next asked whether cordycepin affected the generation of ROS in MA-10 mouse Leydig tumor cells. MA-10 cells were preincubated with 2.5 mM N-acetyl-l-cysteine (NAC) for 1 h, then 100 μ M cordycepin was added and cells were treated for an additional 3 h. ROS production was determined by fluorescence microscopy (Fig. 5a) and flow cytometry using DCF-DA (Fig. 5b). Treatment with cordycepin alone for 3 h increased ROS levels, but pretreatment with NAC fully reversed this effect (Fig. 5a,b). Quantitative analyses further showed that 100 μ M cordycepin alone increased ROS levels by 1.4-fold after 24 h treatment (p < 0.05), but pretreatment with NAC (2.5 mM) restored control levels of ROS (p < 0.01; Fig. 5c). Taxol, which is known to induce ROS 34 , was used as a positive control. Taxol and cordycepin had similar effects on ROS in MA-10 cells (p < 0.01; Fig. 5b,c). To elucidate which ROS accumulation mediated cordycepin-induced apoptosis, the MTT assay was performed to determine cell viability. Results show that NAC significantly recovered cell viability with cordycepin treatment in MA-10 cells (p < 0.05; Fig. 5d). These data strongly suggest that an important step involved in cordycepin-induced apoptosis of MA-10 mouse Leydig tumor cells is generation of ROS.
Cordycepin induced apoptosis by regulating CDK activity through the p53 and TGFβ signaling pathways. The p53 signaling pathway plays important roles in regulating cell growth and apoptosis 35 .
Since TGFβ signals could regulate cell apoptosis in the context of the cell cycle 36 , TGFβ signaling pathways may regulate CDK2 activity and contribute to MA-10 cell apoptosis. After cells were treated with cordycepin for 3 or 6 h, levels of TGFβ 2 and Smad4 were upregulated, but TGFβ 3 was not affected (p > 0.05; Fig. 6e,f). Moreover, the MTT assay was used to determine whether the blocked p53 signaling could rescue the viability of MA-10 cells treated with cordycepin. Results show that p53 siRNA recovered cell viability with cordycepin treatment in MA-10 cells (p < 0.05; Fig. 6g). These results suggest that the p53 and TGFβ signaling pathways may play critical roles in cordycepin-induced cell cycle arrest and apoptosis of MA-10 mouse Leydig tumor cells.
Cordycepin inhibited tumor growth in an allograft model of testicular cancer. Based on the ability of cordycepin to induce apoptosis of MA-10 mouse Leydig tumor cells in vitro, we next tested the in vivo effect of cordycepin using tumor allografts in SCID mice. Male SCID mice were injected subcutaneously with MA-10 cells. After 1 week, mice were given daily injections of vehicle or cordycepin (0.1% in PBS; 20 mg/kg). Antitumor effects were observed in tumor-bearing mice treated with cordycepin compared with controls (n = 6 each group; Fig. 7a). Animals treated with cordycepin exhibited a significant decrease in tumor volume compared with control mice beginning on day 21 (616.5 vs. 1201 mm 3 ; p < 0.05; Fig. 7b), with maximal differences seen on day 33 (3091 vs. 4956 mm 3 ; p < 0.001; Fig. 7b). Importantly, there was not a significant difference in body weight between control and cordycepin-treated mice, indicating that cordycepin treatment had an acceptable safety profile (p > 0.05; Fig. 7b inset). Moreover, immunohistochemical analysis of tumors revealed increased levels of cleaved caspase-3 and decreased levels of p-AKT in cordycepin-treated tumors compared to control mice, indicating cellular apoptosis (Fig. 7c). Thus, cordycepin effectively inhibited tumor growth in both in vitro cellular system and in vivo animal model of tumorigenesis.

Discussion
As Leydig tumors occur most frequently on children and patients will expect long term survival, elucidation of new chemotherapeutic agents without having off-target effects may benefit patients suffering from Lydig tumors and other testicular cancers. In this study, our results indicated that cordycepin reduced the cell viability in MA-10, TM4, and NT2/D1 cells; activated caspases, induced cell cycle arrest, regulated p38 MAPKs signaling, increased ROS levels, and inactivated PI3K/AKT signaling in MA-10 mouse Leydig tumor cells. These results indicate an important role for the cordycein to induce MA-10 cell apoptosis via mediating p38 MAPKs signaling.  We have found that 100 μ M or 1 mM cordycepin induced cleavage of caspase proteins in MA-10 cells. However, different doses of cordycepin affected different caspases. Previous studies have shown that a single factor can activate different patterns of caspase proteins under different treatment regimens (i.e., temporal or dosage variations) 37 , as we have seen here. We further showed that caspase inhibitors improved the viability of cordycepin-treated MA-10 cells, confirming that cordycepin activated caspase cascades to induce the apoptosis of MA-10 cells.
Inhibiting of p38 MAPKs effectively downregulated apoptosis of MA-10 cells treated with cordycepin. Thus, the p38 pathway may play a critical role in cordycepin-induced apoptosis of MA-10 mouse Leydig tumor cells. Previous studies have shown that p38 is important for induction of apoptosis in human breast cancer cells and colon cancer cells 38,39 , which supports our current observation.
PI3K/mTOR inhibition increases the effectiveness of therapeutic drugs in several cancers 40 . Here, cordycepin suppressed levels of p-AKT and p-mTOR. Indeed, cordycepin combined with a PI3K/AKT inhibitor promoted cleavage of caspase-3 in MA-10 cells, indicating that the PI3K/AKT/mTOR signaling pathways also play an important role in cordycepin-induced apoptosis of MA-10 cells. Interestingly, levels of p-AKT and p-mTOR were significantly higher when cells were treated with 1 mM cordycepin for 24 h. This suggests that a protective effect may be activated by high dosages of cordycepin. This warrants further investigation, as these studies may reveal mechanisms by which cells develop resistance to chemotherapeutic agents.
Autophagy is essential for cancer cells to survive under conditions of nutrient starvation, hypoxia, or chemotherapeutic stress 41 . Here, high doses of cordycepin upregulated levels of LC3 II, which indicates autophagy. In addition, the inhibition of autophagy increased the cytotoxic effects of cordycepin in MA-10 cells. These results again suggest that a protective effect (this time related to autophagy) may have been activated in cells exposed to high levels of cordycepin.
Excessive generation of ROS, however, causes mitochondrial dysfunction related to apoptosis 42 . In the current study, excessive generation of ROS was detected in cordycepin-treated MA-10 cells, but treatment with NAC inhibited ROS production and generated cells with normal morphology. Thus, our data show that cordycepin-dependent apoptosis of MA-10 cells depended on ROS production.
The activation of p53 by therapeutic agents results in p53-mediated gene regulation (involving CDKIs, for example) to induce apoptosis and inhibit growth of cancer cells 43 . Our data showed that cordycepin upregulated levels of p53 and p21 and downregulated CDK2 in MA-10 cells. However, knockdown of p53 did not affect p21 or CDK2 levels in cordycepin-treated cells. Thus, alternative pathways activated by cordycepin may regulate p21 and CDK2. In fact, TGFβ could upregulate p21 through Smad4 44 . Here, cordycepin upregulated TGFβ 2 and Smad4. It is therefore highly likely that p21 and CDK2 were regulated by p53 and/or TGFβ pathways to induce apoptosis in MA-10 cells.
Based on these results, we generated a model illustrating how cordycepin treatment is able to induce the apoptosis of MA-10 cells (Fig. 8). The activation of p38 is able to induce a cascade of events that include cleaved caspase-3 and PARP activation resulting in the induction of apoptosis. Cleaved caspase-3 also can be triggered by either death receptor-dependent or mitochondrion-dependent signaling pathways to promote the terminal phase of apoptosis. It should be also noted that treatment of cordycepin promotes MA-10 cell apoptosis through the activation of p53-dependent pathways and accumulation of ROS level. These signaling pathways associate and induce cell cycle arrest, leading to the final stages of apoptosis. Cordycepin also induced autophagy by suppressing PI3K/AKT/mTOR pathways. The Cordycepin treatment is able to induce the activation of p38. The activation of p38 is able to induce a cascade of events that include cleaved caspase-3 and PARP activation resulting in the induction of apoptosis. Cleaved caspase-3 also can be triggered by either death receptor-dependent or mitochondrion-dependent signaling pathways to promote the terminal phase of apoptosis. It should be also noted that treatment of cordycepin promotes MA-10 cell apoptosis through the activation of p53-dependent pathways and accumulation of ROS level. These signaling pathways associate and induce cell cycle arrest, leading to the final stages of apoptosis. Cordycepin also induced autophagy by suppressing PI3K/AKT/mTOR pathways. The molecular mechanism triggered by cordycepin in the MA-10 mouse Leydig tumor cells and this mechanism may be a model for future targets of cancer treatment. Potential major pathways are indicated by large arrowhead, and potential minor pathways are indicated by slight arrowhead. molecular mechanism triggered by cordycepin in the MA-10 mouse Leydig tumor cells and this mechanism may be a model for future targets of cancer treatment.
Finally, we used allograft tumors to show in vivo that cordycepin reduced Leydig tumor growth, decreased levels of p-AKT, and increased levels of cleaved caspase-3. These results are similar to those achieved with cisplatin and bleomycin therapies on testicular germ-cell tumors 45 . We found no evidence of toxicity or side effects associated with cordycepin treatment, although side effects and resistance have been reported for cisplatin 46 . The inhibition of AKT and ERK in colorectal cancer cells prolongs survival in an allograft model of the disease 47 . Although these mechanisms were not explored in our model, cross-talk between different activators of AKT and ERK may have contributed to our findings. To our knowledge, this is the first study to assess the antitumor efficacy of cordycepin on testicular cancer in vivo.
In conclusion, we have demonstrated that cordycein induced MA-10 cell apoptosis via mediating p38 MAPKs signaling and reduced the growth of Leydig cancer cells in vivo. Specifically targeting PI3K/ AKT/mTOR or autophagy pathways increased levels of cordycepin-induced apoptosis in MA-10 cells.
These results indicate that cordycepin is a potential therapeutic agent for further development for the management of testicular cancer.
Cell culture. MA-10 cells were a gift from Dr. Mario Ascoli (University of Iowa, Iowa City, IA) 48 . Cells were maintained in Waymouth MB 752/1 medium (Sigma-Aldrich, St. Louis, MO) containing 10% fetal bovine serum (Invitrogen, Grand Island, NY) and incubated in a humidified atmosphere containing 95% air and 5% CO 2 at 37 °C. The murine TM4 Sertoli-like cell line and NT2/D1 cells were obtained from the American Type Culture Collection (Manassas, VA) 49 . TM4 Sertoli cells were maintained in Dulbecco's modified eagle medium, nutrient mixture F-12 (DMEM/F-12; Invitrogen, Grand Island, NY) containing 10% fetal bovine serum and incubated in a humidified atmosphere containing 95% air and 5% CO 2 at 37 °C. NT2/D1 cells were maintained in Dulbecco's modified eagle medium with high glucose (DMEM/ HIGH GLUCOSE; Invitrogen, Grand Island, NY) containing 10% fetal bovine serum and incubated in a humidified atmosphere containing 95% air and 5% CO 2 at 37 °C.

Isolation of Leydig cells.
Male B6 (C57BL/6NCrj) mice, 5-7 weeks old, were purchased from the National Cheng Kung University Animal Center (Tainan, Taiwan). All animals were housed in groups of four in 29 × 18 × 13-cm polyethylene cages. The animal room was maintained at 22-24 °C under a 12-h light/12-h dark cycle. Purina mouse chow (Ralston-Purina, St. Louis, MO) and water were provided ad libitum. The procedure for sacrificing animals was approved by the counselors of the National Cheng Kung University Animal Center. Testes were removed from sacrificed mice and decapsulated in M199 medium (pH 7.35) containing 4 mM NaHCO 3 , 25 mM HEPES, 0.06 g penicillin, 0.05 g streptomycin, and 0.2% bovine serum albumin (BSA). After decapsulation, testes were incubated in M199 medium containing 1% BSA and 100 U/mL collagenase for 15 min in a shaking water bath (120 cycles/min) at 37 °C. Cold M199 medium was then added to arrest the collagenase activity. Seminiferous tubules were separated from interstitial cells by gravity sedimentation. Cells were then collected by centrifugation (300 × g, 4 °C, 6 min) and resuspended in 2 mL M199 medium containing 0.1% BSA, to give an interstitial cell suspension containing 20-30% Leydig cells. A gradient of 10 mL isotonic Percoll solution and 15 mL M199 medium with 0.1% BSA and 25 mM HEPES was prepared by centrifugation (25 000 × g, 40 min). The interstitial cell preparation was layered onto the gradient and centrifuged (800 × g, 4 °C, 20 min). An 1-mL fraction of the gradient was collected from the top. Mouse Leydig cells were found primarily in fractions 23-25. 3 beta hydroxyl-delta-6-steroid dehydrogenase (3β -HSD) staining solution-which consisted of 1 mg/ml nitroblue tetrazolium, 3 mg/ml β -nicotinamide adenine dinucleotide, 2.88 mg/ml dihydroepiandrosterone in 10% acetone, and 1.6 mg/ml nicotinamide in 0.07 M phosphoric acid solution at pH 7.4-was used to identify the purity of Leydig cells. After isolation, cells were treated with 3β -HSD staining solution at 37 °C for 1 hr. The total number of cells and the percentage of 3β -HSD-positive cells were determined for the Leydig cell preparation 50 . The purity of the Leydig cells was reached 85-95%.
MTT cell viability assay. MA-10 and TM4 cells were seeded into 96-well plates (Techno Plastic Products AG, Trasadingen, Switzerland) at 1 × 10 4 cells in 100 μ L medium per well. After reaching 70-80% confluence, cells were treated with cordycepin and/or kinase inhibitors. MTT was added to a final concentration of 0.5 mg/mL and the cells were incubated for 4 h at 37 °C. The medium was removed, 50 μ L dimethyl sulfoxide were added per well, and crystals were dissolved by gently shaking in the dark for 20 min. Optical density at 570 nm (OD 570 ) was determined using an ELISA microplate reader (VersaMax, Molecular Devices, Sunnyvale, CA).
Lactate dehydrogenase (LDH) release assay. Primary mouse Leydig cells were seeded into 96-well plates at 5 × 10 4 cells/well and incubated for 24 h at 37 °C in the presence or absence of cordycepin. LDH activity was measured in the culture medium using an LDH cytotoxicity assay kit (BioChain Institute, Heyward, CA). The red formazan product was measured at 490 nm using an ELISA microplate reader. Each treatment condition was assayed in triplicate wells. Mean values were expressed as a percentage of control (100% cytotoxicity).
Annexin V staining and flow cytometry. MA-10 cells were treated with 10 μ M, 100 μ M, or 1 mM cordycepin for indicated amounts of time at 37 °C. Negative controls were incubated in plain medium, and positive control were incubated with 50 nM taxol for 24 h. Levels of apoptosis were determined by incubating cells with fluorescently labeled annexin V (Annexin V-FITC apoptosis detection kit, Strong Biotech, Taipei, Taiwan). After incubation with annexin V, floating and adherent cells were trypsinized, pooled, and centrifuged (1 000 × g, 10 min). Pelleted cells were washed in cold phosphate-buffered saline (PBS), centrifuged again (1 000 × g, 10 min), and resuspended in 100 μ L of 1× annexin-binding buffer (10 mM HEPES, 140 mM NaCl, 2.5 mM CaCl 2 , pH 7.4) to yield a cell density of 1 × 10 6 cells/mL. Next, 2 μ L annexin V conjugate and 2 μ L of 100 μ g/mL propidium iodide (PI) working reagent were added to each 100 μ L cell suspension and cells were incubated at room temperature for 15 min. Finally, 500 μ L of 1× annexin-binding buffer was added, and cells were gently mixed. Labeled cells were analyzed by flow cytometry with fluorescence emissions measured at 515 nm (FACScan, Becton-Dickinson, Mountain View, CA). The percentage of cells was calculated using Cell-Quest software (Becton-Dickinson, Mountain View, CA).
Western blotting. MA-10 cells were cultured in 6-cm dishes at 6 × 10 5 cells in 2 mL medium. After treatment, cells were rinsed with cold PBS and harvested using 100 μ L lysis buffer (20 mM Tris-base, 150 mM NaCl, 1 mM ethylenediaminetetraacetic acid, 1 mM ethylene glycol tetraacetic acid, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM β -glycerophosphate, 1 mM Na 3 VO 4 ). The lysate was centrifuged (12 000 × g, 12 min, 4 °C) and the supernatant stored at − 20 °C for later use. Protein concentration was determined using a Micro BCA protein assay kit (Thermo Fisher Scientific, Waltham, MA). Total proteins were separated by sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis, using a 12.5% gel and a running buffer of 24 mM Tris-HCl, 0.19 M glycine, 0.5% SDS, pH 8.3. Proteins were transferred to polyvinylidene difluoride membranes in transfer buffer (20 mM Tris-HCl, 150 mM glycine, 10% methanol, 0.05% SDS). The membranes were incubated in blocking buffer (Tris-buffered saline containing 5% nonfat dry milk and 0.1% Tween-20) at room temperature for 1 h. After washing, the membranes were incubated with primary antibodies (1:4 000) overnight at 4 °C. The membrane was washed three times (10 min each) with Tris-buffered saline containing 0.1% Tween-20, and then incubated with horseradish peroxidase-conjugated antibodies against mouse or rabbit. The immunocomplex was visualized using an enhanced chemiluminescence detection kit (EMD Millipore Corporation, Billerica, MA), and protein levels were quantified using a UVP EC3 imaging system (Upland, CA). β -actin served as a loading control. Small interfering RNA (siRNA) design and transient transfection. Block-iT RNAi design software (Invitrogen, Grand Island, NY) was used to design siRNAs targeting p53 (CAC CUC ACU GCA UGG ACG AUC UGU U or GAG UAU CUG GAA GAC AGG CAG ACU U or GGG ACA GCU UUG AGG UUC GUG UUU G) and caspase-3 (UUA UUA UGC AUA UGC CCA UUU CAG G or UAG AAU CAC ACA CAC AAA GCU GCU C or ACU ACU GCC GGA GUC UGA CUG GAA A). For transfections, 2.0 × 10 5 MA-10 mouse Leydig tumor cells were seeded into 6-well plates and grown to 60-70% confluency without antibiotics. Cells were transfected with siRNAs (10 nM) using 7.5 μ L lipofectamine RNAiMAX (Invitrogen, Grand Island, NY) in a total volume of 2.5 mL. Cells were transfected with siR-NAs for 6 h, then the transfection medium was replaced with fresh Waymouth medium containing 10% FBS, and cells were incubated for an additional 24 h before protein samples were collected for western blotting. Each siRNA experiment was performed in triplicate. A negative control involving a siRNA that was not homologous to any known genes (Thermo Fisher Scientific, Waltham, MA) was used to control against nonspecific effects of oligonucleotides. The transfection efficiency associated with each siRNA was confirmed using SiGLO Red transfection indicators (Thermo Fisher Scientific, Waltham, MA). Uptake of the fluorescent oligos, which correlates with uptake of Stealth ™ RNAi, was used to measure transfection efficiency.

ROS measurements.
Animal treatment. Male CB17/Icr-Prkdc scid /CrlNarl and severe combined immunodeficiency (SCID) mice (12 weeks old) were purchased from National Cheng Kung University Animal Center (Tainan, Taiwan). To induce tumor growth, mice were injected subcutaneously with 1 × 10 7 MA-10 mouse Leydig tumor cells suspended in 0.1 mL PBS. Tumors were allowed to grow for 1 week and mice were randomly assigned to vehicle control or cordycepin treatment (6 mice per group). Injections (0.1% cordycepin in PBS; 20 mg/kg) were performed once daily. Tumors were measured once daily and tumor volumes were calculated using the formula: length × width × depth × 3.14/6. Data were expressed as average tumor volume ± SEM. After animals were sacrificed, tumors were harvested to evaluate protein localization by immunohistochemistry. We have followed the ARRIVE guidelines regarding animal studies 51 .
Ethics statement. The animal study is approved by the institutional animal care and use committee (IACUC) of National Chen Kung University with approval No. 100111. The methods were carried out in accordance to the approved guidelines.