Dihydroartemisinin as a Putative STAT3 Inhibitor, Suppresses the Growth of Head and Neck Squamous Cell Carcinoma by Targeting Jak2/STAT3 Signaling

Developing drugs that can effectively block STAT3 activation may serve as one of the most promising strategy for cancer treatment. Currently, there is no putative STAT3 inhibitor that can be safely and effectively used in clinic. In the present study, we investigated the potential of dihydroartemisinin (DHA) as a putative STAT3 inhibitor and its antitumor activities in head and neck squamous cell carcinoma (HNSCC). The inhibitory effects of DHA on STAT3 activation along with its underlying mechanisms were studied in HNSCC cells. The antitumor effects of DHA against HNSCC cells were explored both in vitro and in vivo. An investigation on cooperative effects of DHA with cisplatin in killing HNSCC cells was also implemented. DHA exhibited remarkable and specific inhibitory effects on STAT3 activation via selectively blocking Jak2/STAT3 signaling. Besides, DHA significantly inhibited HNSCC growth both in vitro and in vivo possibly through induction of apoptosis and attenuation of cell migration. DHA also synergized with cisplatin in tumor inhibition in HNSCC cells. Our findings demonstrate that DHA is a putative STAT3 inhibitor that may represent a new and effective drug for cancer treatment and therapeutic sensitization in HNSCC patients.


Introduction
Signal transducer and activator of transcription (STAT) proteins are a family comprised of seven members, includingSTAT1, STAT2, STAT3, STAT4, STAT5a, STAT5b, and STAT6. It is now clear that STAT3, originally considered as an acutephase response protein [1], is a latent cytoplasmic protein that can be activated by various extracellular polypeptides and other stimuli. These include cytokines (such as IL-6) and growth factors (such as EGF) [2], and hypoxia stress [3], etc. Activation of STAT3 involves phosphrylation of specific tyrosine on STAT3, which in turn induces STAT3monomer's homodimerization and/or heterodimerization with STAT1 or STAT5 through reciprocal Src homology 2 (SH2) domain/phosphotyrosine interactions. Consequently, dimerized STAT3 translocates to the nucleus and binds to specific DNA sequences, and regulates transcription and expression of downstream genes that are associated with cell survival and proliferation [4], cell cycle regulation [5], apoptosis [6] and angiogenesis [7].
Persistent phosphorylation of STAT3 has been found in numerous malignant neoplasms [8][9][10][11][12][13][14], such as head and neck cancers [8]. In fact, frequency of persistentSTAT3activation is more than 95% in head and neck cancers. It has been reported that activation of STAT3 plays a crucial and pivotal role in initiation of malignant transformation [9], immune evasion and suppression [10], as well as cancer invasion and metastasis, suggesting that therapeutic interventions specifically targeting STAT3 can convert the effects of STAT3 activation from pro-tumor to anti-tumor events. Therefore, several novel small-molecule compounds have been developed to inhibit STAT3 phosphorylation [11,12], but their poor solubility and ambiguous aftereffects to host preclude them from clinical trials and practical uses in cancer treatment.
Dihydroartemisinin (DHA) is a semi-synthetic derivative and main active metabolite of the artemisinin, a natural product isolated from a Chinese medicinal herb (Artemisia annua). It is one of first-line antimalarial drugs recommend by World Health Organization in regions where Plasmodium falciparumis becomes resistant to traditional drugs. Moreover, DHA has been shown to exert antibacterial [13] and antiviral [14] effects. In addition to these efficacies, evidence from epidemiological, pharmacological and case control studies has suggested that DHA possess antitumor activity and selective cytotoxicity to various malignancies [15][16][17][18]. Notably, its low toxicity to host and easy solubility in water is the major incentive for developing the compound as an anticancer agent.
A most recent preliminary study [19] revealed that in T cells of contact hypersensitivity mouse model, artemisinin exerted a strikingly inhibitory effect on IL-17 production, and diminished the level of IL-6, which effects were accompanied with a significant reduction of STAT3 activation, suggesting that reduced STAT3 activation is a result of IL-16 expression inhibition. However, there is no further and direct evidence for proving artemisinin/DHA to be a putative STAT3 inhibitor, and little is known about inhibition effects of DHA on proliferation of HNSCC cells. In the present study, we tested the possibility of DHA as a putative STAT3 inhibitor. Therapeutic potency of DHA against HNSCC cells was validated in vitro and in vivo. It is for the first time that we identified DHA as a putative inhibitor of STAT3, and thus the compound represents a promising therapeutic agent against HNSCC.

Western blot analysis
HNSCC cells from culture or from tissue specimens were lysed in Radio Immunoprecipitation Assay (RIPA) lysis buffer (Beyotime Institute of Biotechnology, Haimen, China) and Western blotting was performed using previously described procedures [20]. Briefly, equivalent amounts of proteins were separated by 10% or 12%SDS-PAGE, and then transferred onto a PVDF membrane (Milipore Corporation, Temecula, CA). After blocking with TBS plus 5% non-fat milk for 1 h at room temperature, the membrane was incubated with indicated primary antibodies overnight. Membranes were incubated with a horseradish peroxidase-conjugated secondary antibody (Zhongshan Goldenbridge Co, Ltd, Beijing, China) for 1 h at room temperature. Immunoreactive proteins were visualized with an enhanced chemiluminescence detection system (GE Healthcare Life Sciences, Amersham, UK).

MTT assay
In brief, cells were seeded in 96-well culture plates (10×10 3 cells /200 μl per well) and incubated overnight, and then treated for 24 or 48h with various concentrations of DHA or with the diluting vehicle (DMSO). After incubation, freshly prepared 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) (20μl; 5 mg/ml; Sigma) was added to each well and incubated for 4 h at 37°C. The culture medium was removed from the wells and substituted with the reduced MTT solublized in 150μl/well DMSO. The prepared plates were subjected to examination with an Enzyme-linked Immunosorbent Detector (Model 550, Bio-Rad, Hercules, CA, USA) with absorbance values of 490 nm to evaluate cell viability.
Wound healing assay 1×10 5 HNSCC cells were seeded in 24-well plates and starved for 12 h in serum-free medium. Then a scratch in the cell monolayer was made using a sterilized pipette tip (10μl). Thereafter, all wound-healing processes were performed in serum-free conditions. After washing three times with 0.01M phosphate-buffered saline (PBS, pH 7.4), the plates were incubated with DHA (40 or 20μM) or with the DMSO, and images were captured at 0, 12, 24 and 48 h using digital camera (Canon, Tokyo, Japan) connected to an inverted microscope (Olympus, Tokyo, Japan). The distance of wound sealing was calculated with Image J software (National Institutes of Health, MD, USA). Flow cytometry 4×10 5 HNSCC cells were seeded in each well of 6-well plates in triplicates, incubated overnight, and treated with various concentrations of DHA, or DMSO for 24 h. For cell cycle analysis, the treated cells were harvested, washed twice with ice-cold PBS and fixed in 70%methanol, the samples were exposed to RNase A and stained with propidium iodide (PI) for analyzing the DNA content by FACS Calibur flow cytometer (Becton Dickinson, San Diego, CA). For assessing cell apoptosis, we used the Annexin V-FITC Apoptosis Detection Kit (Becton Dickinson, San Diego, CA) and performed the assessment according to manufacturer's instructions. Percentage of apoptotic cells was analyzed by FACS Calibur flow cytometer.

Murine xenograft model and tumor treatment
Four to 6weeks old BALB/c male mice weighing18 to 20 g were obtained from Vital River Laboratory Animal Technology Co. Ltd. (Beijing, China) and were maintained in an air-conditioned room with constant temperature (22-24°C) and a dark-light cycle (12 h/12 h), and housed in plastic cages, maximum 5 mice per cage. They were fed a standard chow and tap water ad libitum. All animal experiments were reviewed and approved by the ethics committee of Bethune International Peace Hospital. To establish a xenograft tumor, 1×10 7 Cal-27 cells in 200 μl culture medium were inoculated subcutaneously into the left inguinal area of each nude mouse. The general conditions of animal, including mental state, diet and defecation were observed every day after tumor implantation. Furthermore, the bitten wound and locally cutaneous ulcer were also paid close attention. Tumor-bearing mice were randomly assigned to either treatment group or control group when average tumor diameter reached 5 mm. The treatment group received intraperitoneal injection of DHA at a dosage of 50 mg/kg once daily, 5 times per week, for 4 weeks. The control group was given DMSO. Tumor size and body weight was measured every 4days throughout the study. Tumor volume was calculated by the formula: V (cm 3 ) = width 2 (cm 2 ) × length cm)/2. Tumor growth inhibition rates were calculated using the formula (1-average tumor weight of experimental group/average tumor weight of control group) × 100%. In our experiments, no mice were observed to be died of tumor loading. All animals were pre-euthanized with CO 2 and sacrificed by cervical dislocation at the termination of experiments, and the tumors were excised and weighed. The portions of each tumor were taken for Western blot and pathological analyses.

Statistical analysis
All in vitro experiments were repeated at least three times and data were expressed as mean ± SD. Student's t test or on-way ANOVA was used for statistical analysis when indicated. P values < 0.05 were considered to indicate statistical significance.

DHA, as a putative STAT3 inhibitor, effectively blocks activation of STAT3 in HNSCC cells
First, DHA inhibits constitutive phosphorylation and activation ofSTAT3 in HNSCC cells. As observed, three HNSCC cell lines (Fadu, Cal-27 and Hep-2) expressed certain amount of p-STAT3 under normal cultural conditions, typical of constitutive or persistent activation of STAT3 (S1 Fig). DHA significantly inhibited activation of STAT3 in the three HNSCC cell lines in dose-and time-dependent manners ( Fig 1A). Second, DHA effectively inhibited activation of STAT3 induced by hypoxia and IL-6 in HNSCC cells. When hypoxia or IL-6 was given, activation of STAT3 was substantially induced, which was effectively inhibited by DHA ( Fig  1B and 1C). Finally, inhibition of STAT3 activation by DHA was also confirmed in vivo as observed on expression of p-STAT3 in the xenograft animal tumors (Fig 1D). Taken together, DHA inhibits STAT3 activation under different circumstances.
The efficacies of DHA on STAT3 inhibition were compared with those of AZD1480 and AG490, two specific inhibitors of Jak2/STAT3 signaling, which are commonly used in the experimental studies. The doses of AZD 480 and AG490 were chosen based on a preliminary dose-escalation study (S2 Fig). All of the three inhibitors remarkably downregulated levels of p-Jak2 and p-STAT3in HNSCC cells (Fig 2). The inhibitory effects of DHA on STAT3 activation were comparable with those of AZD 480 and AG490, which was most likely more effective than that of AG490 in Hep-2 cells (Fig 2, the right panel).

Inhibition of STAT3 activation by DHA in HNSCC cells is attributed to selective blockade of Jak2 phosphorylation
After treating Fadu, Cal-27 and Hep-2 cells with different concentrations of DHA, inhibition of STAT3 activation was prominent, but little or no inhibition of Akt and ERK phosphorylation was observed (Fig 3A); the expression of p-Jak2 in three HNSCC cell lines was markedly reduced in a dose-dependent manner; however, neither inhibition of p-SRC nor downregulation of p-EGFR was observed (Fig 3B). It was of particular interest to note that, in the xenograft animal tumors, application of DHA resulted in significant simultaneous decrease of p-STAT3 and p-Jak2 expression, but the levels of p-SRC and p-EGFR remained unchanged (Fig 3C). Furthermore, treatment with DHA and/or dominant negative Jak2 (DN-Jak2) rather than abrogation of EGFR and SRC activation by their dominant negative constructs (DN-EGFR and DN-SRC) reduced the expression of p-Jak2 and p-STAT3 in Cal-27 cells (Fig 3D). However, DHA did not altered the level of overexpressed p-STAT3, apoptosis, and cell cycle in Cal-27 cells transfected with constitutively active STAT3 (CA-STAT3) construct ( Fig 3E). Collectively, inhibition of STAT3 activation by DHA is through selective blockade of Jak2 phophorylation and activation.

DHA inhibits proliferation, growth and migration of HNSCC cells in vitro
DHA significantly inhibited the proliferation of HNSCC cells. Fadu cells were the least sensitive to DHA while other two cell lines were more sensitive to the compound comparatively ( Fig 4A). The DHA-induced growth inhibition involved apoptosis induction. The percentage of the apoptotic cells in HNSCC cells treated with DHA was greater than that in untreated (control) cells (Fig 4B). It was also observed that DHA reduced the migration of all three cell lines in a time-dependent manner (Fig 4C, left and middle panels) (p<0.01). As demonstrated in the cell migration assay, expression of two important metastasis-associated proteins (MMP-2 and -9), were remarkably decreased after treatment with DHA (Fig 4C, right panel). Other proteins downstream of STAT3 including Mcl-1, Bcl-xl, Cyclin-D1 and VEGF were also downregulated in the three HNSCC cell lines after DHA treatment (Fig 4D).

DHA inhibits tumor growth in the xenograft animal models of human HNSCC
Compared to the vehicle (DMSO), DHA significantly decreased tumor size, overall tumor weight, and mean tumor volume (Fig 5A-5C). However, the general conditions and the body weight of the animals treated with DHA showed nearly no change for the period of DHA application, which implies that the compound has no obvious toxicity to experiment animals ( Fig  5D). As expected, the expressions of the above-mentioned functional proteins downstream of STAT3 were all inhibited in xenograft animal tumors treated with DHA. Apart from inhibition of p-Jak2 and p-STAT3 in DHA-treated tumors (Fig 3C), expression of downstream proteins of STAT3 including Mcl-1, Bcl-xl, Cyclin-D1, VEGF, MMP-2, and MMP-9 was significantly suppressed in DHA-treated tumors (Fig 5E).

DHA synergistically potentiates the antitumor activity of cisplatin in HNSCC cells
The synergistic effects between the two drugs were seen in different combinations in separate cell lines (Fig 6A). For Cal-27 cells and Hep-2 cells, synergistic effects were observed when cells were treated with 10 μM DHA and 10 μM cisplatin, as well as 20 μM DHA with 5 and 10 μM cisplatin. In Fadu cells, synergistic effects were noted when cells were cultured with 10 or 20 μM DHA and5 or 10μM cisplatin. The CIs of these combinations were all less than 1, suggestive of a synergistic activity between the two drugs in proliferation inhibition on HNSCC cells (Fig 6B). Inhibition of STAT3 activation resulted in reversion of the cell cycle distribution patterns induced by upregulation of p-STAT3. It was revealed that, after treatment with DHA,

Discussion
Searching novel therapeutic agents that target some specific signaling molecules is crucial for developing revolutionized and promising treatment modalities to cure HNSCC [21]. Of particular interest, STAT3 may serve as a potential target for this purpose. It has been demonstrated that activation of STAT3 is necessary for the growth of HNSCC cell lines [22], and the status of STAT3 activation is a marker to predict the survival and prognostic outcomes in HNSCC patients [23]. In this regard, targeting activation of STAT3 would be an ideal strategy for preventing pathogenesis and progression of HNSCC.
Up till now, several compounds have been screened and shown to inhibit the STAT3 activation. These STAT3-targeting agents are currently divided into two main groups, direct action group and indirect action group. The former group, which includes Stattic [24], ST3-H2A2 [25], S3I-1757 [26] and Diindolylmethane [27], interacts with structural domain of STAT3 and disrupts the process of STAT3 phosphorylation. By contrast, the indirect action group alters the status of STAT3 activation via acting on the upstream proteins that dominate the STAT3 activation. In fact, Pentoxifylline [28], AZD1480 [29], AG490 [30], JSI-124 (cucurbitacin I) [31], Indirubin [32] and LBH589 [33] belong to the indirect inhibitors of STAT3. Although the before-mentioned agents have been shown to possess substantial activities on inhibiting STAT3 activation, they are mostly used for experimental purposes, and only a very few have been approved for clinical trials by FDA as a STAT3 inhibitor due to major toxicity-related safety concerns. Very recently, a first-in-man phase I study was conducted in patients with refractory solid malignancies [34] to evaluate the safety, pharmacokinetics, pharmacodynamics and tumor inhibition efficacy of OPB-51602, a novel small-molecule and direct inhibitor of STAT 3 phosphorylation at Tyr705 and Ser727 sites. Although the preliminary tumor inhibition effects were observed in non-small cell lung carcinoma (NSCLC), the side effects and dose-related toxicities of the chemical were obvious, which need further in-depth and longterm clinical investigations.
As a highly effective treatment drug of falciparum malaria, DHA has been proved to be a safe, well tolerated, and widely used in clinic. Moreover, a few studies have reported that the efficient antimalarial drug DHA shares strong antitumor activities in different human cancer cells, such as colorectal carcinoma cells, T-lymphoma cells, and leukemia cells [35][36][37]. However, there has been no available evidence for proving the effects of DHA on growth of HNSCC cell inhibition to date. In the present study, it is for the first time we demonstrated that DHA possesses antitumor properties against a variety of human HNSCC cells both in vitro and in vivo. The underlying mechanism involves selective inhibition of Jak2/STAT3 signaling and its downstream target proteins, thereby producing proliferation inhibition and inducing cell apoptosis. treated with DHA as described. Expression of p-EGFR, p-Jak2, p-SRC and p-STAT3 in tumor tissues was analyzed by Western blotting. (D) DHA specifically blocked Jak2 activation in HNSCC cells. Cal-27 cells were transfected with DN-EGFR, DN-Jak2, DN-SRC, CA-STAT3, or empty vector, and exposed to 80 μM DHA for 24 h. Expression of p-EGFR, p-Jak2, p-SRC and p-STAT3 were studied by Western blotting. (E) CA-STAT3attenuated the cell cycle arrest induced by DHA. Cal-27 cells were transfected withCA-STAT3or empty vector and exposed to 80 μM DHA for 24 h. Cell cycle and cell apoptosis were analyzed by flow cytometry. All experiments were performed in triplicates.  Our results also demonstrated that the DHA-induced inhibitory effects on STAT3 signaling are selective and specific, as we summarized in Fig 7. In the present study, we made it clear that inhibition of STAT3 by DHA depended on blockade of Jak2 kinase rather than on inhibition of EGFR tyrosine and SRC family kinases in HNSCC cells. Moreover, DHA inhibited phosphorylation of STAT3, but did not affect the constitutive activation of Akt and ERK, the key proteins of PI3K/Akt and MAPK/RAS oncogenic signaling pathways. Our observations further confirmed the notion that Jaks are central mediators of STAT3 signaling in solid tumor cells [29].
In the present investigation, the efficacy of STAT3 inhibition by DHA was also confirmed. DHA exhibited substantial effects on inhibiting STAT3 activation both in vitro and in vivo. Inhibition of STAT3 activation by DHA exerted functional impacts on HNSCC cells, including decrease in cell viability and migratory capability, G0/G1 phase cell cycle arrest and apoptosis in HNSCC cells. As postulated in Fig 7, these effects are most likely to be attributed to downregulation of the cycle regulator cyclin D1, antiapoptotic proteins Bcl-xL and Mcl-1, growth factor VEGF, and metastasis-associated proteins MMP-2 and MMP-9, most of which are major downstream targets of STAT3 [33,38].
Apart from direct inhibition of tumor growth, blocking STAT3 activation is of potential value in combating chemoradiotherapeutic resistance of HNSCC. For instance, activation of STAT3 is associated with resistance of laryngeal carcinoma cells to ionized radiation; blockade of STAT3 signaling by shRNA sensitized the laryngeal carcinoma cells to radiotherapy both in vitro and in vivo [39]. Moreover, activation of STAT3 is also associated with chemoresistance of HNSCC [40]. In the present study, we demonstrated for the first time that DHA potentiates the antiproliferative effects of cisplatin in HNSCC cells. It would be of great interest to further elucidate whether DHA has synergized effects with other chemotherapy drugs in killing HNSCC cells.
STAT3 activation is also responsible for HNSCC resistance to some molecular targeted therapies. In fact, activation of STAT3 is related to resistance of HNSCC to EGFR monoclonal antibodies, such as Cetuximab [41]. Targeting STAT3 with a STAT3 decoy reduced cellular viability and the expression of STAT3 target genes in EGFR inhibitor (Cetuximab) resistance models; the addition of a STAT3 inhibitor to EGFR blocking strategies significantly enhanced antitumor effects of Cetuximab in vivo [42]. Activation of STAT3 also constitutes a cause of insensitivity of HNSCC cells to proteasome inhibitor. For instance, Bortezomib up-regulates STAT3 and synergizes with inhibitors of STAT3 to promote cell death in HNSCC [43]. Therefore, addition of DHA, as a putative and effective STAT3 inhibitor, to the above-mentioned targeting therapies may greatly enhance the treatment efficacies and thus outcomes of HNSCC patients.
Owing to the scarcity of the available STAT3 inhibitors that can be used in clinic, defining DHA as a putative STAT3 inhibitor is of profound clinical implications. Because our results are preclinical, further controlled clinical trials are due to carry out to confirm the efficacy of DHA in improving the outcomes of HNSCC or some other human malignancies.
(C)DHA induced inhibition of migration of HNSCC cells. Cells were incubated with 40μM (FaDu) or 20μM (Cal-27and Hep-2) DHA or DMSO. The wound healing capacity was measured at 0, 12, 24, and 48 h. Data were expressed as means ± SD (left and middle panels). Simultaneous determination of levels of MMP-2, MMP-9 and p-STAT3 was conducted (right panel). (D) Effects of STAT3 inhibition by DHA on expression of downstream proteinsMcl-1, Bcl-xl, Cyclin-D1 and VEGF in HNSCC cells as determined by Western blotting. All experiments were performed in triplicates. doi:10.1371/journal.pone.0147157.g004