Adiponectin suppresses tumor growth of nasopharyngeal carcinoma through activating AMPK signaling pathway

Adiponectin is an adipocyte-secreted cytokine that enhances insulin sensitivity and attenuates inflammation. Although circulating adiponectin level is often inversely associated with several malignancies, its role in the development of nasopharyngeal carcinoma (NPC) remains unclear. Here, we investigated the clinical association between circulating adiponectin level and NPC, and examined the impact of adiponectin, as well as the underlying mechanisms, on NPC growth both in vitro and in vivo. The association between circulating adiponectin level and the risk of developing NPC was assessed in two different cohorts, including a hospital-based case–control study with 152 cases and 132 controls, and a nested case–control study with 71 cases and 142 controls within a community-based NPC screening cohort. Tumor xenograft model, cell proliferation and cycle assays were applied to confirm the effects of adiponectin on NPC growth in cultured cells and in xenograft models. We also investigated the underlying signaling mechanisms with various specific pharmacological inhibitors and biochemistry analysis. High adiponectin levels were associated with a monotonic decreased trend of NPC risk among males in both the hospital-based case–control study and a nested case–control study. In vitro, recombinant human full-length adiponectin significantly inhibited NPC cell growth and arrested cell cycle, which were dependent on AMPK signaling pathway. The growth of xenograft of NPC tumor was sharply accelerated in the nude mice carrying genetic adiponectin deficiency. An adiponectin receptor agonist, AdipoRon, displayed strong anti-tumor activity in human xenograft models. These findings demonstrated for the first time that circulating adiponectin is not only inversely associated with NPC, but also controls the development of NPC via AMPK signaling pathway. Stimulation of adiponectin function may become a novel therapeutic modality for NPC.


Background
Nasopharyngeal carcinoma (NPC) has a high prevalence in southeast Asia, especially in southern China [1,2]. Although Epstein-Barr virus (EBV) infection is the most well-characterized risk factor [3][4][5], other potential genetic and environmental factors have also been suggested to contribute to the pathogenesis of this malignancy [6,7]. Obesity has been suggested by some studies as a risk factor of NPC for decades, but the findings from different groups have been inconsistent [8][9][10]. Recent published studies have proposed that altered levels of adipose-derived adipokines, such as adiponectin, leptin, and resistin, may have contributed to the development of various malignancies [11,12].
Adiponectin is an adipocytokine almost exclusively secreted by the adipose tissue [13,14]. Circulating levels of adiponectin are paradoxically reduced in obesity and diabetes [15]. Mounting evidence has clearly shown its intimate involvement in the regulation of cardiovascular function, glucose/lipid metabolism, and chronic inflammation [16][17][18]. A series of clinical studies have also revealed that circulating adiponectin is inversely associated with the risk of several malignancies, such as multiple myeloma, prostate, breast, colorectal, and pancreatic cancers [19][20][21][22]. In fact, adiponectin elicits anti-proliferative effects in different tumor histocytes in vivo and in vitro, including breast, prostate, hepatocellular, and endometrial carcinomas [23][24][25].
It is still unclear if adipose tissue, through the secreted adiponectin, plays a role in controlling the development of NPC. Herein, we set out to examine the relationship between blood concentrations of adiponectin and the risk of developing NPC in two cohorts from Guangdong province, including a hospital-based case-control study with 152 cases and 132 controls, and a nested case-control study with 71 cases and 142 controls within a community-based NPC screening cohort. Importantly, we also investigated whether, and by what mechanisms, adiponectin directly regulates the growth of NPC cells.

Animal breeding and subcutaneous transplantation
All animal experimental procedures were approved by the Experimental Animal Academic Ethics Committee of Guangdong Pharmaceutical University (gdpulacspf2017064).
Adiponectin deficient mice were generously provided by Professor Philipp Scherer of the University of Texas Southwestern (Dallas, TX, USA). Male nude mice were purchased from the GemPharmatech (Nanjing, Jiangsu, China), and crossed with APN−/− female mice to generate three genotypes of nude mice: APN+/+, APN+/−, and APN−/−. Mice were kept in the Laboratory Animal Center of Guangdong Pharmaceutical University (Guangzhou, Guangdong, China), and maintained in specific pathogen-free conditions with stationary temperature of 23-25 °C and 12-h light/dark cycles.
1 × 10 6 CNE-2 or 5-8F cells were resuspended in 100 µL PBS and subcutaneously injected into the right armpit region of five-to six-week-old male nude mice. Tumors were measured using digital Vernier calipers every day, with tumor volume calculated using the formula [sagittal dimension (mm) × cross dimension (mm)] 2 /2 and expressed in cm 3 . All animals were sacrificed, tumor tissues were collected, imaged, and weighed.
For AdipoRon administration, four days after injection NPC cells, the mice were randomly allocated into two groups (Vehicle and AdipoRon groups) of 6 mice per group. In the AdipoRon group, mice were intragastrically administered 50 mg/kg AdipoRon suspended in corn oil every other day. In the Vehicle group, mice were administered solvent alone in corn oil.
Recombination human full-length adiponectin was dissolved in deionized water to prepare a working stock solution of approximately 0.5 mg/mL (BioVendor, Brno, Czech Republic). AdipoRon was dissolved in DMSO to prepare a working stock solution of approximately 50 mM (Selleck Chemicals, Houston, TX, USA). Compound C was purchased from MedChem Express (Monmouth Junction, NJ, USA), and was prepared as a stock concentration at 10 mM in DMSO and stored at − 80 °C.

Cell viability and proliferation assays
Cell viability was measured using cell counting kit-8 (CCK-8) (Sangon Biotech, Shanghai, China). Cells were cultured in 96-well plates, with six duplicate wells in each group, and pre-treated in 100 μL medium with or without different concentrations of inhibitors for 1 h, followed by solvent alone, AdipoRon or APN for the indicated period. After incubation, CCK-8 solution was added to each well followed by a further 2 h incubation under 5% CO 2 at 37 °C. Absorbance was automatically measured at 450 nm with a microplate reader (Infinite F50, Tecan Group Ltd., Mannedorf, Switzerland). The relative cell viability was calculated as the percentage of untreated cells. Cell proliferation was measured using plate clone formation and 5-ethynyl-2'-deoxyuridine (EdU) assays. CNE-2 cells were plated in 12-well plates and treated with human recombinant adiponectin or AdipoRon. Then, the culture medium was replaced with fresh medium containing adiponectin every 3 days. After 7 days' treatment, the medium was removed, and cell colonies were fixed and stained with crystal violet (Sangon Biotech). Images were taken with a digital camera, colonies contained more than 50 cells in each well were counted. The EdU assay were preformed according to manufacturer's instructions (Ribo-Bio, Guangzhou, Guangdong, China). The EdU-positive rate was calculated as EdU-positive cells/Hoechst-stained cells × 100%. The assays were repeated in triplicate.

Transient transfection with small interfering RNA
The small interfering RNA (siRNA) oligos against Adi-poR1, AdipoR2 and scrambled control siRNA were commercially synthesized by RiboBio (Guangzhou, Guangdong, China), and transfected with riboFECT CP transfection reagent (RiboBio, Guangzhou, Guangdong, China) according to the manufacturer's protocol. The siRNA duplexes used for this study are listed in Additional file 1: Table S2. Two days after transfection, the cells were subjected to total RNA isolation and viability assays.

Cell cycle assay
CNE-2 cells were incubated in serum-free medium overnight, and then cells were treated with adiponectin or Adi-poRon. Cells were collected, washed, and suspended in cold PBS. Cells were then fixed in 70% cold ethanol at 4 °C overnight. After fixation, the cells were washed with PBS twice, resuspended in 0.2 mL PI/RNase staining buffer (BD Biosciences, San Jose, CA, USA) for 30 min at room temperature. The cell cycle distribution was determined by the DxP Athena flow cytometry system (Cytek Biosciences, Fremont, CA, USA), and the percentages of different phases of cell cycle were determined using ModFit LT 5.0 (Verity Software house, Topsham, ME, USA).

Cell apoptosis assays
PE Annexin V apoptosis detection kit (BD Biosciences, #559763) was used to determine cell apoptosis. Cells treated with the indicated drug concentrations. After treatment, we harvested the cells, washed them twice with PBS, and stained them using Annexin V-PE and 7-AAD for 15 min in the dark, followed by analysis using the DxP Athena flow cytometry system (Cytek Biosciences). The upper right quadrant represents late apoptotic cells, and the lower right quadrant represents early apoptotic cells. The assessment of the apoptosis rate was the sum of early and late apoptosis.

RNA extraction and qRT-PCR
Total RNA was extracted from cell by using Trizol reagent (Sigma; T9424). The quantity and quality of RNA were determined using a ScanDrop2 nano-volume spectrophotometer (Analytik Jena), and reversely transcribed based on the HiScript II Q RT kit (Vazyme; R223) according to the manufacturer's instructions. Amplification and real-time detection were performed on a qTOWER3 G real-time PCR system (Analytik Jena) by using ChamQ Universal SYBR qPCR Master Mix (Vazyme; Q711) in 20 μL reaction. The relative expression levels of each targeted gene were normalized by subtracting the corresponding mouse β-actin threshold cycle (CT) values by using the ΔΔCT comparative method. Three biological replicates per group were used for qPCR. Primers were synthesized by Sangon Biotech (Shanghai, China). Sequences of all primers used are provided in Additional file 1: Table S3.

Statistical analysis
Data were expressed as mean ± SD. Statistical analyses were performed using GraphPad Prism 7.0 (GraphPad Software, La Jolla, CA, USA). The statistical significance between groups was assessed by Student's t test or by analysis of variance (ANOVA) with Sidak's multiple comparisons test. A value of P < 0.05 was considered statistically significant.

Association of circulating adiponectin with the risk of NPC
The baseline characteristics of the two cohorts are presented in Table 1. In both case-control studies, patients with NPC and controls had similar gender and age distributions. In the hospital-based case-control cohort, the median circulating adiponectin was significantly lower in the cases versus the controls (1.76 vs 2.94 μg/ mL, P = 0.003) (Additional file 1: Table S1). Similarly, the median circulating adiponectin also showed a trend of lower value in the cases (vs. the controls) in the nested case-control cohort (1.62 vs 2.21 μg/mL, P = 0.06) (Additional file 1: Table S1). These differences were statistically significant among men in both the hospital-based casecontrol study (1.72 vs 3.50 μg/mL, P = 0.002) and the nested case-control study (1.45 vs 2.09 μg/mL, P = 0.021) but not among women (Additional file 1: Table S1).
Analyses of the risk of NPC associated with circulating adiponectin level were stratified by gender. Among men, we confirmed there was a strong inverse association with risk of NPC in the hospital-based case-control study (highest vs. lowest tertile: OR = 0.45; 95% CI 0.27-0.89, P trend = 0.014) and in the nested case-control study (OR = 0.28; 95% CI 0.10-0.80, P trend = 0.018) ( Table 1). Based on multivariable models, the significance of association was sustained by adjusting for a set of other risk factors for NPC, such as age, family history, smoking, and EBV antibody levels at least in the nested case-control cohort (OR = 0.11; 95% CI 0.01-0.70, P trend = 0.019) ( Table 1). Taken together, these data show that high adiponectin levels were associated with a statistically significantly reduced future risk of NPC among men, suggesting that altered adiponectin levels may play a role in the etiology of NPC.

Adiponectin-deficiency exacerbates the growth of NPC cells in vivo
The outcomes from these clinical studies warrants further investigation if there exists a causative effect of adiponectin diminution on NPC growth. To this end, we established an NPC tumor model by subcutaneously implanting human NPC (CNE-2) cells-derived xenograft in the immunodeficient nude mice carrying adiponectin deficiency. Compared to the control nude mice, heterozygotic and homozygotic adiponectin deficiency both sharply accelerated tumor growth within even 10 days, and caused significantly increased tumor weight (Fig. 1A, B), indicative of a strong suppressive effect of adiponectin on NPC progression. Consistent with these results in the animal model, co-incubation of recombinant human full-length adiponectin in cultured human CNE-2 and C666-1 cells strongly inhibited the proliferation of cells, as revealed in both the plate clone formation assay and 5-ethynyl-2'-deoxyuridine (EdU)-staining assay (Fig. 1C, D).
The observation that adiponectin suppressed the growth of human NPC-xenograft led us to examine whether adiponectin might modulate cell cycle progression in NPC cells. Flow cytometry analysis showed that co-incubation with adiponectin significantly increased cell population arrested at G 0 /G 1 stage and decreased the population at G 2 /M stages (Fig. 1E, F). In line with the flow cytometric results, Western blot assays revealed marked reduction of cell cycle regulators, particularly cyclin B1 and cyclin D1, and parallel increase in p21 and p27, following the treatment of CNE-2 and C666-1 cells with adiponectin (Additional file 1: Fig.  S1A, B). Such modification of cell cycle regulator appeared to be specific to a sub-group as the levels of some other cell cycle regulatory proteins such as CKD2 and CKD4 were not significantly altered by adiponectin treatment (Additional file 1: Fig. S1A, B). In the same Western blot assay, we also found that adiponectin markedly promoted the activation of AMPK as reflected by the increase of AMPK-phosphorylation (Fig. 1G, Additional file 1: Fig. S1C). LKB1 is known to act as an upstream kinase, directly phosphorylating and activating AMPK [35]. Our results showed that although adiponectin did not affect the level of total LKB1, it increased the level of p-LKB1 (Fig. 1G, Additional file 1: Fig. S1C). Furthermore, adiponectin suppressed the phosphorylation of ERK1/2 (Fig. 1G, Additional file 1: Fig. S1C).
To test whether the inhibitory effect of adiponectin on cell cycle progression in NPC cells is dependent on AMPK activation, NPC cells were incubated with adiponectin in the absence or presence of Compound C (ComC), a specific AMPK inhibitor. ComC treatment completely released the NPC cells from the arrest at the G0/G1 phase ( Fig. 2A, B). Consistent with such observation, ComC treatment also essentially neutralized the anti-proliferative effect of adiponectin on CNE-2 and C666-1 cells (Fig. 2C). Blockade of AMPK signaling with ComC also neutralized adiponectin-dependent reduction of cyclin D1 as well as the induction of CKIs (p21 and p27) (Fig. 2D, E, Additional file 1: Fig. S2). Collectively, these results demonstrated that adiponectin can directly inhibit the proliferation of human NPC cells by regulating cell cycle-regulatory proteins via activation of AMPK.

Inhibition of NPC growth by an adiponectin receptor agonist
To examine whether the anti-proliferative effect of adiponectin is mediated by adiponectin receptors (AdipoRs) in the human NPC cells, we examined cell lines expressed AdipoR1 and AdipoR2 (Fig. 3A, B). Following reduction of AdipoR1 or AdipoR2 expression via small interfering RNA (siRNA) (Fig. 3C, D) blocked the inhibitory effect of adiponectin on proliferation (Fig. 3E). Under the same condition, adiponectin-induced AMPK phosphorylation was also diminished by the knockdown of AdipoR1 or AdipoR2 expression in NPC cells (Fig. 3F).
We further determined if boosting the activity of both adiponectin receptors, AdipoR1 and AdipoR2, will stall the growth of nasopharyngeal carcinoma, if so, what is the impact of altering AdipoRs signaling on cell cycle progression of NPC cells. Initial analysis genomic data in The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) found the expression of both Adipo-R1and AdipoR2 in various types of head & neck squamous cancers (HNSC) as well as NPC tissues (Fig. 4A, B). Interestingly, further analysis showed that both receptors displayed higher expression levels in NPC than those in the normal nasopharyngeal epithelium (Fig. 4C), which might reflect a compensatory elevation in response to the reduced circulating level of adiponectin in NPC patients. These bioinformatic findings prompted us to explore if a specific adiponectin receptor agonist, AdipoRon, could inhibit the proliferation of human NPC cells.
In NPC cells, AdipoRon treatment of CNE-2 and C666-1 cells generated dose-dependent suppression of cell viability with an IC 50 at 50.79 μM and 40.69 μM, respectively (Fig. 5A). Even at the doses substantially below IC 50 , AdipoRon still strongly inhibited the proliferation of NPC cells as evident in both the plate clone formation assay (Fig. 5B) and the EdU-staining assay (Fig. 5C).
Flow cytometric analysis showed that co-incubation with AdipoRon significantly elevated cellular population arrested at G0/G1 phase after 24 h treatment (Fig. 5D, E), while inducing apoptosis in the NPC cells at 48 h post treatment (Additional file 1: Fig. S3 Fig. 5F). In complete agreement with the results seen in adiponectin-treated NPC cells, AdipoRon induced significant reduction in the expression of cell cycle regulators, particularly cyclin B1 and cyclin D1, and a parallel increase in p21 (Fig. 5G, Additional file 1: Fig. S4A). Although the levels of CDK2 and CDK4 was not altered in the 24 h treatment duration, their levels were significantly decreased after 48 h (Fig. 5G, Additional file 1: Fig. S4A). In the analysis of these cellular samples, AdipoRon treatment also elicited strong AMPK signaling (Fig. 5H, Additional file 1: Fig.  S4B). Taken together, these data demonstrated that stimulation of adiponectin receptor signaling via its specific agonist, AdipoRon, could not only inhibit proliferation and but also induce cell death in NPC cells. The observations described above led us to investigate tumor-suppressive effects of AdipoRon in animal models. To this end, we evaluated the growth of subcutaneously implanted CNE-2-and 5-8F-dervied tumor in the nude mice. From day 4 after inoculation, the mice were orally administered 50 mg/kg AdipoRon every other day. Following the delivery, we observed a significant retardation of tumor growth (Fig. 6A-F, Additional file 1: Fig.  S5A, 5C) without body weight loss (Additional file 1: Fig.  S5B, 5D). Ki-67 staining of the sections prepared from AdipoRon-treated tumors revealed a sharp reduction of tumor cell proliferation in the treated vs. the untreated tumors (Fig. 6G, H). In addition, we noticed a significant decrease of vessel density, suggesting either indirect or direct inhibition of angiogenesis by AdipoRon (Fig. 6G, I). Collectively, these findings suggest that boosting adiponectin activity through oral administration of Adi-poRon exhibited a strong suppressive effect against human NPC cells derived xenograft tumors.

Discussion
The findings derived from the retrospective and prospective case-control studies established for the first time the inverse relationship between adiponectin and the risk of NPC. This newly identified inverse relationship is completely independent of other well known risk factors, such as age, EBV infection status, family history, suggesting an independent regulation of NPC development by an adipocyte-derived metabolic hormone. The current study did not stratify the correlation against body weight owing to the lack of such data in both cohorts. However, previous studies have largely ruled out the association of body weight with risk of NPC [8][9][10]. Thus, we do not believe that this issue will change the outcomes of this analysis. Interestingly, we did not find evidence of an association among women, which might be due to the relatively few female NPC cases. Extending investigation is required to confirm our findings, and to better elucidate sex hormones affect the relationship between adiponectin and NPC.
In corroborating the outcomes of such strong clinical associations, we further established the causative effects as well as the underlying mechanisms, of adiponectin on human NPC development. Co-incubation with adiponectin or adiponectin-receptor agonist suppressed the growth of human NPC cells, arrested cell cycle via Adi-poR1-and AdipoR2-mediated AMPK activation. Importantly, adiponectin-deficiency significantly accelerated, while administration of adiponectin-receptor agonist inhibited, the growth of human NPC cell-derived xenografts in the nude mice. Taken together, these results unequivocally solidified that adiponectin is not just a correlative circulating factor but also a direct regulatory factor in the development of NPC.
Initially discovered as a crucial regulator of inflammation, energy balance, glucose/lipid metabolism [15,18], adiponectin has been reported to have direct anti-proliferative effects in several malignant cell lines [36]. In this study, we demonstrated that adiponectin could directly suppress the growth of NPC cells by arresting cell cycles at the G0/G1 phase through regulating the expression of several cell cycle key regulators, with the activation of AMPKα as the most likely initiating signaling event. Tumor suppressor LKB1 as the critical upstream kinase, its phosphorylation leads to the activation of AMPK [35]. We found adiponectin treatment did not affect the level of total LKB1, it significantly increased the level of p-LKB1, further demonstrating that adiponectin suppresses NPC growth through activating LKB1/AMPK signaling. Furthermore, adiponectin induced AMPK activation probably suppressed the phosphorylation of ERK1/2, further enhancing the anti-proliferative effect. Recent studies have shown that AMPK signaling can play a critical role in the regulation of cancer cell proliferation via induction of apoptosis and cell cycle arrest [37][38][39][40]. Several tumor suppressor genes, such as p53, mTOR, and p27, are considered as the downstream signaling components of AMPK activation [41,42]. 5-Aminoimidazole-4-carboxamide ribonucleoside (AICAR), a pharmacological activator of AMPK, suppresses cell growth of head and neck squamous cell carcinoma [43]. Moreover, metformin has recently received attention as an anti-tumor drug, since it induces inhibition of cancer cell proliferation via activation of AMPK signaling [44]. Our previous study has already shown that adiponectin could arrests endometrial cancer cells at the G0/G1 stage, possibly by activating AMPK [24]. Such result is well corroborated by the observation that the inhibitory effect of adiponectin on NPC cell proliferation was neutralized by inhibition of AMPK activity with a specific AMPK inhibitor, ComC, particularly adiponectin-controlled cell cycle progression in NPC cells. Besides AMPK, other molecular mechanisms could also play critical. For instance, chronic inflammation is an important even in propelling the development of NPC [3]. Owing to its wellestablished anti-inflammatory function, adiponectin can also prevent NPC development by suppressing proinflammatory cytokines such as IL6, tumor necrosis factor-α, and interferon γ [45], as well as inducing the expression of anti-inflammatory cytokines such as IL-10 and IL-1RA [46]. From a translational perspective, we have tested if stimulation of adiponectin receptor activity would attenuate the growth of human nasopharyngeal carcinoma, and applied AdipoRon, the first oral adiponectin receptor agonist capable of binding and activating both AdipoR1 and AdipoR2 [47], in the human NPC model. AdipoRon has emerged as a possible candidate for the treatment of different pathological conditions, including metabolic, cardiovascular, and cognitive dysfunction of Alzheimer's disease, specifically comorbidity between depression and obesity [47][48][49][50]. In this study, we have demonstrated that oral administration of AdipoRon exhibited a robust anti-cancer effect against human NPC derived xenograft tumors, and the dosing of AdipoRon applied in our study (at 50 mg/kg) closely matched those reported in other mouse models [47,49,51,52]. With multitudes of mechanisms, AdipoRon may represent a therapeutic agent that can be applied towards the treatment of human NPC.

Conclusions
In conclusion, our findings from this study shed some new light on the pathogenesis of NPC, highlighting the importance of an adipocyte-derived endocrine hormone, adiponectin, as a crucial inhibitor to NPC tumorigenesis via AMPK activation. Further investigations are needed to establish the linkages between other adipocyte-derived endocrine hormones, in addition to adiponectin, with the progression and pathological features of NPC, such as tumor grade, vascular invasion, and metastasis. Our findings herein may provide knowledge of adiponectin as a novel therapeutic target in NPC therapy.