Cordycepin Inhibits the Growth of Hepatocellular Carcinoma by Regulating the Pathway of Aerobic Glycolysis

Hepatocellular carcinoma (HCC) is one of the most common malignant tumors in China, with a high incidence and mortality rate. Glucose metabolism reprogramming is a major characteristic of tumor cells. Increasing evidence indicates that aerobic glycolysis is associated with tumor growth and insensitivity to chemotherapy. Cordycepin inhibits the growth of HCC cells, but the mechanism is yet to be elucidated. Herein, in vitro and in vivo methods were utilized to investigate the cordycepin-inhibited growth of HCC by regulating the metabolic pathway of aerobic glycolysis. In vitro analyses using colony formation and flow cytometry revealed that cordycepin inhibits HCC cells' proliferation and promotes apoptosis. In addition, cordycepin reduced the production of lactic acid and pyruvate, reduced the uptake of glucose, and decreased the extracellular acidification in HCC cells. Specifically, cordycepin inhibited the expression of HK2, LDHA, and PKM2 in aerobic glycolysis via the AMPK-Akt pathway. Taken together, these findings revealed that cordycepin reduces the tumor energy supply and decreases lactic acid production, thereby inhibiting the growth of HCC cells by regulating the metabolic pathway of aerobic glycolysis. These findings might provide novel insights into the mechanisms underlying cordycepin-mediated inhibition of tumor growth as well as a new treatment for HCC.


Introduction
Liver cancer is a highly malignant tumor [1]. It is classifed into four diferent pathological types: hepatocellular carcinoma (HCC), intrahepatic cholangiocarcinoma (ICC), hepatoblastoma (HB), and mixed liver cancer (HCC-ICC) [2]. Te incidence of HCC accounts for 75-85% of liver cancers [3]; however, its pathogenesis is yet to be clarifed. Clinical studies demonstrated that in the early stage of cancer, chemotherapeutic drugs can inhibit the growth of cancer cells; however, in the process of tumor development, the migration and invasion of cancer cells weaken the efect years, cordycepin has played a major role in the treatment of cancer. Te underlying mechanisms are summarized as enhancing apoptosis and regulating the cell cycle of tumor cells [8,9], thereby preventing the metastasis of tumor cells and the synthesis of new substances for cell growth [10,11], and inhibiting cell proliferation and angiogenesis [12]. Wang et al. identifed cordycepin as an efective drug to block radiation ulcers in rats/mice by increasing NRF2 nuclear expression and preventing cell senescence. Cordycepin also activates AMPK by binding to α1 and c1 subunits near the self-inhibitory domain of AMPK, promoting p62-dependent Keap1 autophagy degradation, and inducing NRF2 separation from Keap1 and transfer to the nucleus [13]. Although the mechanism of antitumor efects has been reported, the role of cordycepin in inhibiting tumor growth by regulating energy metabolism has not yet been reported.
Adenosine 5-monophosphate-activated protein kinase K (AMPK) regulates intracellular energy metabolism [14]. In higher eukaryotes, AMPK has acquired the ability to sense available energy in cells by directly binding to adenine nucleotides [15]. Once activated, AMPK phosphorylates the key proteins, shifting the metabolism to increased catabolism and reduced anabolism [16]. In addition to direct involvement in the regulation of key enzymes in metabolic pathways, AMPK also reconnects cell metabolism for the long term by targeting transcriptional regulators [17]. Protein kinase B (Akt) is a serine/threonine kinase downstream of AMPK, which can be activated by a wide range of growth signals; also, the biochemical mechanism underlying Akt activation is elucidated [18]. Once activated, Akt regulates many downstream proteins related to cell proliferation, migration, and glucose metabolism. Akt is the hub of multiple signaling pathways that are often regulated in various human cancers [19].
Furthermore, for the occurrence and development of cancer, the cells need to reprogram their catabolism and anabolism to obtain sufcient energy and synthetic biomass for cell growth [20]. Warburg et al. demonstrated that even under oxygen-rich conditions, tumor cells metabolize glucose through the glycolysis pathway, i.e., aerobic glycolysis [21]. Te rapid metabolism of glucose results in a local anoxic environment outside the tumor cells [22]. Aerobic glycolysis is also considered an adaptive mechanism that supports the metabolic microenvironment of tumors and facilitates rapid biosynthesis. Previous studies have confrmed that phosphofructokinase (PFK), hexokinase 2 (HK2), lactate dehydrogenase A (LDHA), and pyruvate kinase 2 (PKM2) are the key proteins in the metabolic pathway of tumor aerobic glycolysis, which are potential targets for tumor drug therapy. Te expression of HK2 is upregulated in many cancers [23]. Jian et al. found that in animal models, the absence of HK2 decreased the proliferation of cancer cells without obvious side efects, indicating that targeting HK2 is a feasible cancer treatment strategy. Inhibition of HK2 downregulated aerobic glycolysis, thus afecting various pathways of central metabolism and destabilizing the outer membrane of the mitochondria, resulting in cell death [24].
In summary, the disruption of energy metabolism promotes the proliferation of tumor cells. Te Warburg efect is one of the main characteristics of the energy metabolism of tumor cells and a target for cancer treatment. Cordycepin inhibits cancer cell growth; however, the mechanism is not yet clarifed. Herein, we hypothesized that cordycepin might inhibit the aerobic glycolysis pathway, thereby inhibiting the production of lactic acid and pyruvate and reducing glucose uptake. Te present study confrmed that cordycepin reduces the tumor energy supply and decreases lactic acid production to inhibit HCC growth by regulating the metabolic pathway of aerobic glycolysis. Tus, these fndings might provide novel insights into the mechanisms of cordycepin-mediated inhibition of tumor growth.

Cytotoxicity
Analysis. Te cytotoxicity of cordycepin was confrmed on HepG2 and HCC-LM3 cells using the CCK-8 method. Te cells were seeded in 96-well plates at a density of 10000 cells/well and treated with 0, 25, 50, and 100 μg/mL cordycepin, respectively, at 37°C for 24 h, followed by the addition of the CCK-8 reagent (C0038, Beyotime, Shanghai, China). Te reaction was incubated with cells for 1 h before measuring the absorbance at 450 nm on a microplate reader (Termo Fisher Scientifc, USA). Subsequently, the tumor cell growth inhibition ratio (TGI) was calculated according to the following formula: where T is the average absorbance value of the treated groups and C is the average absorbance value of the control group. 2.6. Glucose Uptake Assays. Glucose uptake was detected using the glucose (GO) assay kit (GAGO20, Sigma-Aldrich, St. Louis, MO, USA). HCC-LM3 and Hep-G2 cells were plated in 6-well plates at a density of 1 × 10 6 cells/well and cultured with cordycepin at diferent concentrations for 24 h. Ten, the cells were washed with PBS and cultured in glucose-free DMEM for 6 h. Subsequently, the supernatants were collected to measure the glucose uptake, following the manufacturer's instructions. Te glucose consumption was calculated by deducting the measured glucose concentration in the medium from the original glucose concentration.

Pyruvate Kinase Activity.
Pyruvate kinase activity was detected using the kit (Solarbio). HCC-LM3 and Hep-G2 cells were plated at a density of 1 × 10 6 cells/well in 6-well plates and treated with diferent concentrations of cordycepin for 24 h. Subsequently, the cells were counted in the precipitate, and the supernatant collected by centrifugation was transferred to a 1.5 mL tube. All estimations were performed according to the manufacturer's instructions. accordance with the guidelines established by the National Institutes of Health Guide for the Care and Use of Laboratory Animals. HCC-LM3 cells were cultured until the cell density reached 5 × 10 7 /mL. Ten, 1 × 10 7 cells were inoculated in a 200 μL volume on the right back of the nude mice. Once the tumors reached an average volume of 100 mm 3 , the mice were randomly divided into fve groups (each group, n � 5) and treated as follows: group I: PBS as the control; group II: 10 mg/kg cordycepin; group III: 20 mg/kg cordycepin; group IV: 10 mg/kg 5 FU; and group V: 10 mg/ kg cordycepin + 10 mg/kg 5 FU. Te intraperitoneal injection was administered once a day. Te tumor size was measured using Vernier calipers every 3 days, and the volume was calculated using the formula: 0.5 × (Length × Width 2 ), wherein L represents the long diameter, and W represents short diameter. After 30 days, the mice were sacrifced.

18 F FDG-PET/CT.
Before imaging, the rats were fasted for 6-8 h and weighed. About 5 mBq 18 F-FDG developer was injected into nude mice by caudal vein injection. After 40 min, 2% isofurane gas was inhaled, and after 5 min, whole-body PET/CT (Mediso, Hungary; Bioscan, USA) scanning was conducted. Subsequently, PET images were acquired on the computer. Te PET images of nude mice were captured on the workstation, and SUVmax was measured (continuously at fve tumor levels, and the mean value was the fnal SUVmax of the tumor).

Immunohistochemical (IHC) Staining.
Te parafn sections were sliced into 3-5 mm thin slices on a tissue slide and baked at 65°C for 1 h to prevent peeling. Te baked slices were placed in xylene for 10 mins, followed by 95% ethanol and 75% ethanol for 5 min each. Ten, the slices were rinsed in double-distilled water, immersed in PBS bufer (pH 7.2), and soaked in a 3% hydrogen peroxide-methanol solution (90 mL methanol/10 mL 30% hydrogen peroxide solution) at room temperature for 15 mins to eliminate the efects of endogenous peroxidase. After PBS washes, the slices were sealed in a humid chamber with 20% FBS for 30 mins. For antigen retrieval, the slides were heated in citrate bufer. Ten, the slides were incubated with primary antibodies against PKM2 (1 : 800; 4053S; Cell Signaling Technology), HK2 (1 : 50, 22029-1-AP, Proteintech), and LDHA (1 : 800; 3582S; Cell Signaling Technology) at 4°C, overnight, followed by incubation with secondary antibody at 37°C for 1 h. After DAB color development (Dako, Carpinteria, CA, USA), the slides were stained with hematoxylin (Riblology, Shanghai, China) for 15 s, washed under tap water, dried at 65°C, and sealed with a coverslip and neutral gum before being examined under the microscope.
2.14. TUNEL. Dewaxing: Same as IHC Staining. Te TUNEL staining kit was purchased from Beyotime. Te wet sections were incubated with protease K solution at 37°C for 10-30 mins, followed by PBS washes. Ten, the slides were dried before adding 50 μL TDT enzyme working solution, and the reaction was facilitated at 37°C for 60 mins in the dark. Subsequently, 50 μL of streptavidin-HRP was added to the slides and incubated at 37°C for 30 mins in the dark. Finally, a fuorescence quenching and sealing solution was used to seal the slides before observing them under a fuorescence microscope.

Statistical Analysis.
Te group results were expressed as the mean ± standard deviation of at least three repeated samples. Te diferences between groups were analyzed using GraphPad Prism and SPSS statistical software (SPSS, Chicago, IL, USA). P < 0.05 indicated statistically signifcant results.

Cordycepin Inhibits the Proliferation of HCC Cells.
To detect the efects of cordycepin on tumor cell proliferation, CCK-8 assays and plate cloning experiments were performed on HCC-LM3 and Hep-G2 cells. Te CCK-8 assays showed that cordycepin inhibits the proliferation of HCC-LM3 and Hep-G2 cells after treatment with 25, 50, and 100 cordycepin and 10 μg/mL 5 FU as a positive control for 24 h, and the cell proliferation was signifcantly decreased with the increase in cordycepin concentration (Figures 1(a)  and 1(b)). Next, we detected the efects of cordycepin on the colony formation of HCC-LM3 and Hep-G2. Compared to the control group, the size of the colonies in the cordycepin treatment group was smaller, and their number decreased signifcantly (Figure 1(c)).

Cordycepin Induces Apoptosis in HCC Cells.
Apoptosis is a form of programmed cell death. Apoptotic, live, and necrotic cells can be distinguished using AnnexinV/ PI double staining. HCC-LM3 and Hep-G2 cells were treated with diferent concentrations of cordycepin for 48 h. Te apoptotic rate of HCC-LM3 and Hep-G2 cells was detected by fow cytometry. Te results showed that, compared to the control group, 50 μg/mL cordycepin had no signifcant efect on HCC cells ( * P > 0.05), while 100 μg/mL cordycepin induced apoptosis in HCC-LM3 cells. Te apoptotic rate of Hep-G2 cells was 10.7%, which was signifcantly higher than that of the control group ( * * P < 0.01) (Figure 2(a)). Te efects of cordycepin on apoptosis-related proteins in HCC-LM3 and Hep-G2 cells were detected by Western blot assay. Ten, HCC-LM3 and Hep-G2 cells were treated with 50 μg/mL and 100 μg/mL cordycepin for 24 h, and the expression of apoptotic proteins Bcl-2 and cleaved caspase3 was detected by Western blotting. Te results showed that the expression of Bcl-2 in HCC-LM3 and Hep-G2 was downregulated and that of the cleaved caspase3 protein was upregulated after 100 μg/mL cordycepin treatment (Figures 2(b) and 2(c), * P > 0.05 or * * P < 0.01). Tese fndings confrmed that cordycepin induces apoptosis in HCC-LM3 and Hep-G2 cells.
Autophagy is another form of programmed cell death. We would make sure whether cordycepin-induced cell death could result from autophagy. So, we measured and compared the expression of autophagy-regulated target molecules and autophagosomes. Te results showed no signifcant diference between the control and cordycepin groups in the expression of ATG5, Beclin1, or LC3 in HCC-LM3, Hep-G2 cells, and HCC-LM3 tumor-bearing nude mice tissues ( Supplementary Figures 1(a), 1(b), 1(d), and 1(e)). In addition, the number of autophagosomes could not be induced obviously after 100 μg/mL cordycepin treatment in Hep-G2 compared with the control group, as determined by transmission electron microscopy ( Supplementary  Figure 1(c)). All these results confrmed that Cordycepin does not induce the autophagy cell death pathway.

Cordycepin Regulates Aerobic Glycolysis in HCC Cells.
Lactic acid is the primary end-product of glucose metabolism, and a proportion is excreted out of the cell. Te metabolic level of aerobic glycolysis is indirectly refected by estimating the content of the extracellular lactic acid. After Hep-G2 and HCC-LM3 cells were treated with 50 μg/mL and 100 μg/mL cordycepin, the lactic acid content was detected using a lactic acid detection kit. Te results showed that the extracellular lactic acid production of HCC-LM3 and Hep-G2 cells treated with 50 μg/mL and 100 μg/mL cordycepin was signifcantly decreased compared to the controls. Tis phenomenon suggested that cordycepin limits Evidence-Based Complementary and Alternative Medicine the production of lactic acid in the metabolic process of aerobic glycolysis in HCC cells (Figure 3(a)). Pyruvate is the intermediate metabolite of the glycolysis pathway, and a proportion of pyruvate can be excreted from the cell while some content is retained for the subsequent enzymatic reaction to produce lactic acid.
After treating Hep-G2 and HCC-LM3 cells with 50 and 100 μg/mL of cordycepin for 24 h, the cells were lysed, and the pyruvate content was estimated using a pyruvate detection kit. Te results showed that the content of pyruvate in Hep-G2 and HCC-LM3 cells treated with 50 μg/mL and 100 μg/mL cordycepin was signifcantly     Evidence-Based Complementary and Alternative Medicine lower than that in control cells ( * * * P < 0.001). Tis phenomenon suggests that cordycepin decreases the production of pyruvate in the metabolic process of aerobic glycolysis in HCC cells (Figure 3(b)). After Hep-G2 and HCC-LM3 were treated with 50 μg/ mL and 100 μg/mL cordycepin, the glucose content in the culture medium of each group was measured using a glucose detection kit. Te results showed that, compared to the control group, the glucose uptake of HCC-LM3 and Hep-G2 cells treated with 50 μg/mL and 100 μg/mL cordycepin was signifcantly decreased, indicating that cordycepin inhibits glucose uptake in HCC cells (Figure 3(c)).
Te glucose metabolism produces lactic acid, which leads to the production and excretion of protons in the extracellular culture medium. In this study, the acidity of the culture medium was directly detected by the hippocampal XF24 extracellular fux analyzer, i.e., the ECAR. Te extracellular acidifcation of hepatoma cells (LM3 and Hep-G2) in each group was assessed on XF24 at 50 μg/mL and 100 μg/mL cordycepin. Te results showed that, compared to the control group, the ECAR of the 50 and 100 μg/mL cordycepin treatment groups signifcantly was decreased (Figures 3(d) and 3(e)). Tis suggests that cordycepin inhibits the ECAR rate of HCC cells.

Cordycepin Regulates the Aerobic Glycolysis
Pathway through the AMPK-Akt-HK2/PKM2/LDHA Axis. In order to further explore the mechanism of cordycepin on aerobic glycolysis, Hep-G2 and HCC-LM3 cells, treated with 50 μg/ mL and 100 μg/mL cordycepin for 24 h, were selected, and the key target genes related to glucose metabolism and lipid metabolism were detected by RT-qPCR assay. Te results confrmed that the mRNA levels of HK2, PKM2, and LDHA were signifcantly reduced in HCC-LM3 and Hep-G2 cells after treatment with 50 μg/mL and 100 μg/mL cordycepin (Figures 4(a) and 4(b)).
Furthermore, the upstream regulatory targets of cordycepin regulating the metabolic pathway of aerobic glycolysis were investigated. Tus, Hep-G2 and HCC-LM3 cells were treated with 50 μg/mL and 100 μg/mL cordycepin, respectively. Te expression of AMPK, p-AMPK, Akt, p-Akt, HK2, PKM2, and LDHA, the key targets of aerobic glycolysis, was detected in these cells. Te expression of p-

Cordycepin Suppresses Tumor Growth In Vivo by Regulating Aerobic
Glycolysis. Te transplanted tumor model of HCC-LM3 tumor-bearing nude mice was established and randomly divided into fve groups (the control group, the 10 mg/kg cordycepin group, the 20 mg/kg cordycepin group, the 10 mg/kg 5 FU group, the 10 mg/kg cordycepin group, and the 10 mg/kg 5 FU group) after the tumor grew up to day 10 (n � 5 nude mice/group). 5 FU is one of the most commonly used chemotherapeutic agents for gastric cancer. 5 FU has a direct cytotoxic efect on tumors, and drug combinations are a commonly used clinical treatment for tumors. In this study, we designed a combined administration group. Te tumor volume and analysis of nude mice in the diferent treatment groups are shown in Figure 5(a). We found that the combination group was more efective than cordycepin or 5 FU alone. Te results showed that the tumor volume in the 10 mg/kg and 20 mg/ kg cordycepin treatment groups decreased signifcantly compared to the control group. PET/CT was used for in vivo imaging of 18 F-FDG-labeled glucose to observe its metabolism in various parts of the body; the glucose uptake value was expressed as SUVmax. Most of the studies have shown that tumors have a high uptake of 18 F-FDG. Te white arrow indicates the location of the tumor. Te current results showed that the SUVmax of the 10 mg/kg and 20 mg/kg cordycepin treatment groups was signifcantly lower than that of the control group ( Figure 5(b)), suggesting that cordycepin reduces glucose uptake in tumor-bearing nude mice. TUNEL staining was used to detect the apoptosis of tumor tissues. Te results showed that the number of apoptotic bodies increased signifcantly in the 10 mg/kg and 20 mg/kg cordycepin treatment groups ( Figure 5(c)). We also found that the combination treatment was more efective than cordycepin and 5 FU alone. Te parafn sections of tumor tissues of nude mice were stained with IHC to detect the expression of HK2, PKM2, and LDHA, the key enzymes of aerobic glycolysis. Te results showed that, compared to the control group, the protein expression levels of HK2, LDHA, and PKM2 in the cordycepin treatment group were decreased ( * P < 0.05 or * P < 0.01; Figure 5(d)). Tis phenomenon suggests that cordycepin regulates the expression of key enzymes of aerobic glycolysis metabolism in HCC. Some tumor tissues from nude mice were randomly selected in each group, and the expression of HK2, PKM2, and LDHA proteins in tumor tissues was detected by Western blotting. Te results showed that, compared to the control group, the expression of HK2 was downregulated in the 10 mg/kg and 20 mg/kg cordycepin groups, and the expression of PKM2 and LDHA was signifcantly downregulated in the 20 mg/kg cordycepin group ( * P < 0.05 or * * P < 0.01; Figure 5(e)). Terefore, we verifed that cordycepin suppresses tumor growth in vivo by regulating aerobic glycolysis. According to the results, cordycepin inhibits tumor growth by mediating aerobic glycolysis, as shown in Figure 6.

Discussion
Liver cancer results in almost one million deaths annually worldwide, most of which are associated with metastasis [25]. Te treatment of liver cancer is predominantly based on surgical resection combined with radio-, chemo-, and targeted-therapy, which inhibits the malignant proliferation of liver cancer cells [26]. Normal cells undergo the four processes of proliferation, senescence, apoptosis, and death. For normal cells, senescence and apoptosis belong to the cell fault protection program, which is conducive to a functional life. Tumor cells are diferent from normal cells as they suppress cell fault protection and hyperproliferate. Terefore, the proliferative activity of tumor cells is a critical prognostic index for the diagnosis of cancer, and the measurement of cell proliferation can provide valuable information for the prognosis of patients [27].
Te Warburg efect suggests that tumor cells provide energy and substances for their own growth through aerobic glycolysis. Tis is the metabolic characteristic of most tumors [28], and the feature provides a direction for tumor treatment. Glucose is not only the substance needed by the human body for survival but also the most important source of energy. Tus, reducing glucose uptake and inhibiting the aerobic glycolysis pathway can indirectly inhibit tumors. Combining the current results and the relevant literature, the detection of glucose metabolites in this study revealed that cordycepin afects the expression level of glucose metabolites in cells after administration. Tus, it is speculated that the efect of cordycepin on HCC cells might be mediated through the glycolysis pathway; the specifc mechanism needs to be further explored.
Cordycepin is considered a natural agonist of AMPK, a key regulator of energy metabolism and mitochondrial diagenesis [29]. Moreover, AMPK has attracted wide attention as a potential therapeutic target for metabolic diseases, including type 2 diabetes and cancer [30]. Xx et al. demonstrated a signifcant protective efect of cordycepin on mice under conditions of hepatic steatosis, infammation, liver injury, and metabolic stress by activating the AMPK signaling pathway [31]. Wang et al. found that cordycepin prevents radiation ulcers by inhibiting cell senescence via NRF2 and AMPK in rodents, and activation of AMPK or NRF2 might be the therapeutic targets for preventing cell senescence and radiation ulcer [13]. In various diseases, cordycepin has been shown to activate AMPK and regulate the downstream metabolic pathways. Terefore, it is a promising natural agonist of AMPK. Based on the fnding that cordycepin regulates glucose metabolism in liver cancer cells, this study verifed that cordycepin regulates the glucose metabolism pathway in tumor cells through phosphorylation of AMPK and downstream Akt. Te specifc target molecules of the glucose metabolism pathway have not been explored in this study, which is essential in the future. In vivo studies demonstrated that the combination of cordycepin and 5 FU had a better anticancer efect than either of the drugs alone, which provided directions for follow-up studies. Te in vitro and in vivo experiments showed that cordycepin inhibits the proliferation and induces the apoptosis of HCC. In vitro mechanism studies showed that cordycepin inhibits the growth of HCC cells through the AMPK/Akt-HK2/PKM2/LDHA regulatory axis. In conclusion, cordycepin inhibits the growth of liver cancer by regulating the aerobic glycolysis pathway. Tese fndings provide a new mechanism for cordycepin in antitumor therapy and a new research strategy for traditional Chinese medicine and tumor metabolism.

Data Availability
All remaining data are available within the article and supplementary fles or available from the authors upon request.

Ethical Approval
All animal experiments were performed in accordance with the National Institutes of Health guide for the care and use of laboratory animals and protocols approved by the Animal Ethics Committee of the Shanghai University of Medicine and Health Sciences.

Disclosure
Ya Chen and Yuan Jiang are the co-frst authors.

Conflicts of Interest
Te authors declare that they have no conficts of interest.