Background
Oral squamous cell carcinoma (OSCC) is a major subtype of head and neck carcinoma with many pathological differences from cancers found at other sites in the head and neck region and is one of the most widely prevalent cancers throughout the world [
1]. OSCC accounts for over 90% of malignant neoplasms of the oral cavity. Its mortality rate has remained largely unchanged over the past decade, with a 5-year survival rate under 50% [
2]. However, the molecular and cellular mechanisms underlying the pathogenesis of OSCC are relatively poorly understood. Surgical resection is feasible for OSCC patients, but it is not effective for late-stage metastatic tumors; thus, adding chemotherapy, radiotherapy or both (chemoradiotherapy, CRT) as adjuvant or as definitive treatment are acceptable modalities [
3]. Cancer therapy has increasingly focused on novel treatment strategies combining radiotherapy and chemotherapy. Within the context of CRT, the optimal dose for OSCC irradiation has not been clearly defined. To diminish the damage to normal tissue, treatment with chemical modifiers as radiosensitizers in combination with lower dose irradiation (IR) may augment its overall therapeutic efficacy [
3,
4]. Recently, the anti-cancer drug bortezomib, the first proteasome inhibitor approved by the U.S. Food and Drug Administration for the treatment of multiple myeloma, has attracted attention for its ability to treat solid tumors alone or in combination with radiotherapy [
5,
6]. We have demonstrated previously that a proteasome inhibitor combined with radiation possesses synergistic anti-pancreatic cancer potency both in vitro and in an orthotopic murine model [
7]; nevertheless, the effects and the precise mechanism of combined treatment of bortezomib and radiation against OSCC remain unclear.
There are two major protein degradation pathways in eukaryotic cells: the ubiquitin (Ub)-proteasome system (UPS) and the autophagy-lysosome system (hereafter autophagy) [
8]. The UPS is a selective proteolytic system in which substrates are recognized and tagged with ubiquitin for degradation. This pathway has essential functions in homeostasis, which include preventing the accumulation of misfolded or deleterious proteins [
5,
8]. To ensure appropriate destruction of those proteins that are no longer needed, the components of this system must act in a highly coordinated manner through definitive steps that include polyubiquitylation, deubiquitylation, and degradation of the target protein [
9]. In general, ubiquitins are conjugated via a lysine residue at position 48 to target proteins for degradation [
10]. Whereas for proteins tagged with lysine 63 (K63)-linked polyubiquitin chains of ubiquitin, instead of being targeted for proteasomal degradation, they alter protein function or localization and thereby regulate signaling activation including receptor endocytosis, protein trafficking, kinase activity and DNA repair [
5,
10]. For example, tumor necrosis factor receptor-associated factor 6 (TRAF6), a critical regulator of NF-κB signalling, has been identified as an E3 ligase for K-63-linked polyubiquitination of PKB/Akt. This polyubiquitination promotes membrane recruitment of PKB/Akt and its phosphorylation and activation upon growth factor stimulation [
11]. Recently, TRAF6 has also been corroborated to be an oncoprotein involved in cancer development and progression in multiple cancers [
12]. Tumor cells overexpressing TRAF6 produce proteins that promote cell growth and survival while meanwhile inhibiting the mechanisms of cell death, and thus, it represents a potential therapeutic target for the treatment of cancer through inhibition of UPS TRAF6-mediated K63-linked polyubiquitination as a means to shift this fine equilibrium towards cell death [
5,
13].
Autophagy is a bulk degradative system that uses lysosomal hydrolases to degrade long-lived proteins and damaged or old organelles; it involves membrane formation followed by fusion of the vesicle with lysosomes [
14]. Interestingly, autophagy appears to have a dual role in cancer therapy [
15]. The proper amount of autophagy promotes cancer cell survival, whereas a high level of autophagy results in autophagic cell death [
16]. Whether autophagy has pro-survival or pro-death effects depends on different factors, including cancer cell type/phase, stress context and the microenvironment. Therefore, life or death of the cell is context dependent. Autophagy and UPS were previously thought to be independent of each other in components, action mechanisms, and substrate selectivity. Notably, recent studies suggest that a single proteolytic network in cells comprised of the autophagy and UPS systems functionally cooperate with each other to maintain proteostasis, and the central feature common to both autophagy and UPS is ubiquitination [
8]. Ubiquitination can either generate degradation signals on substrates delivered for destruction by proteasomes or lysosomes, or modulate their non-proteolytic processes [
17]. The inhibition of the UPS results in the compensatory activation of autophagy, implying there is a crosstalk between autophagy and UPS [
8]. It has been reported that bortezomib can induce proteasome independent degradation of TRAF6 in myelodysplastic syndrome [
18]. Since TRAF6 is necessary for maintaining the survival of cells, its degradation by bortezomib-induced autophagy contributes to cell death [
18]. However, whether TRAF6 plays a critical role in the cross talk between UPS and autophagy in OSCC cancer cells remains undetermined.
In the present study, we investigated the anticancer effect of combined IR and bortezomib treatment on human OSCC cancer cells both in vitro and in a xenograft murine model. The types of cell death, especially autophagic cell death, and the underlying mechanisms, including ubiquitination and phosphorylation of signaling regulators, were examined. In addition, the clinical impact of TRAF6 in oral cancer patients was also investigated.
Methods
Additional procedures are described in detail in the Additional file
1.
Irradiation treatment, cell viability and synergistic interaction analysis
Irradiation was performed with 6 MV X-rays using a linear accelerator (Digital M Mevatron Accelerator, Siemens Medical Systems, CA, USA) at a dose rate of 5 Gy/min. An additional 2 cm of a tissue-equivalent bolus was placed on the top of the plastic tissue-culture flasks to ensure electronic equilibrium, and 10 cm of tissue-equivalent material was placed under the flasks to obtain full backscatter. After IR treatment, cells were treated with bortezomib immediately. The treated cells were centrifuged and resuspended with appropriate amount of PBS. For cell viability assay, 20 μl cell suspension was mixed with 20 μl Trypan blue solution (0.4% in PBS). Placing the mixture on a hemocytometer, and the blue-stained cells were counted as nonviable. The effect of the combination treatment was evaluated by the combination index (CI) method using CalcuSyn software (Biosoft), which is based on the median effect model of Chou and Talalay [
19]. The experimental data were entered into the CalcuSyn interface and CI values were calculated. CI < 1, CI = 1, and CI > 1 indicate synergism, additive effect, and antagonism, respectively.
Clonogenic assay
The cells were irradiated using the dosages of 2, 4, 6 or 8 Gy. Bortezomib was added to the cells at concentrations of 25 or 30 nM. The cells were trypsinized and counted. Known numbers of cells were subsequently replated in 6-cm culture dishes and returned to the incubator to allow for colony development. After 2 weeks, colonies (containing≥50 cells) were stained with 0.5% crystal violet solution. The plating efficiency (PE) is the ratio of the number of colonies to the number of cells seeded in the nonirradiated group. Calculation of survival fractions (SFs) was performed using the equation: \( \mathrm{SF}=\frac{\kern0.75em \mathrm{colonies}\ \mathrm{counted}\kern0.75em }{\kern0.75em \mathrm{cells}\ \mathrm{seeded}\times \mathrm{PE}\kern0.75em } \).
Early apoptosis and autophagy detections
Apoptosis was assessed by observing the translocation of phosphotidyl serine to the cell surface, as detected with an Annexin V apoptosis detection kit (Calbiochem, San Diego, CA, USA), according to our previous report [
20,
21]. For autophagy analysis, cell staining with acridine orange (Sigma Chemical Co.) was performed according to published procedures [
22,
23], adding a final concentration of 1 μg/ml for a period of 20 min. Flow cytometry was used to detect Annexin V-positive cells and acidic vesicular organelles (AVOs).
Stable knockdown clone selection
For generation of a TRAF6-knockdown stable cell line, SAS cells were transfected with lentiviral vector containing short hairpin RNA (shRNA) purchased from the National RNAi Core Facility located at the Institute of Molecular Biology/Genomic Research Center, Academia Sinica. The clone is identified as TRCN0000007348, which targeted the human TRAF6 transcript sequences, 5′-CGGAATTTCCAGGAAACTATT-3′. We added the lentivirus to cells in a growth media containing 8 μg/ml polybrene (MOI = 3). After 16 h post infection, we removed the media and replaced it with media containing puromycin (0.4 μg/ml), and then amplified the cells.
shRNA transfection
The clone (TRCN0000040123) of shRNA targeting ATG5 was purchased from the National RNAi Core Facility located at the Institute of Molecular Biology/Genomic Research Center, Academia Sinica. We used TransIT-X2 transfection reagent (Mirus Bio Corporation, Madison, WI) to transfect ATG5 shRNA into SAS cells. For 10-cm dish, the total volume of medium and cells per well prior to transfect should be 10 ml. In an eppendorf tube, combined the serum-free medium for 1 ml and plasmid DNA for 10 μl of a 1 μg/μl stock. Added 30 μl TransIT-X2 to the diluted DNA mixture. Pipetted gently to mix completely and incubated at room temperature for 30 min, added total of complex to 10-cm dish for Incubate for 24-48 h. SAS cells were harvested 48 h after shRNA transfection for Western blotting.
Subcutaneous xenograft in vivo model
Male NOD-SCID mice (5- to 7-weeks-old) were acquired from the National Cheng Kung University Laboratory Animal Center (Taiwan). The animals were housed 5 per cage at 23 ± 2°C with 60% ± 5% relative humidity and subjected to a 12-h light/12-h dark cycle. The animals were adapted to the environment 1 week before the start of the experiments. SAS cells (2 × 106 cells in 0.1 ml of PBS) were subcutaneously inoculated into the right back of the mice. Seven days post injection, the mice were randomized into 5 groups (n = 5 for each group): (1) control (99.9% DMSO): mice were injected (intraperitoneally, i.p.) with DMSO. (2) Bortezomib group: i.p. injections with 1 mg/kg Bortezomib twice a week for 3 weeks. (3) IR group: a single dose of 6 Gy IR. (4) Bortezomib + IR: combination therapy with 1 mg/kg Bortezomib twice a week and a single dose of 6 Gy IR at the beginning of the first week. For the TRAF6-knockdown stable clone animal model, the mice were randomized into 4 groups (5 mice per group): normal, control, SAS shTRAF6#1 and SAS shTRAF6#2. Individually, SAS, shTRAF6#1 and shTRAF6#2 cell lines (2 × 106 cells in 0.1 ml of PBS) were subcutaneously inoculated into the right back of the mice and the mice were sacrificed after 3 weeks. Volume estimations were determined using the following formula: \( \mathrm{volume}=\frac{\uppi \times \mathrm{width}\times \mathrm{length}\times \mathrm{height}}{6} \). Mouse body weight was measured once per week and was used as an indicator of the systemic toxicity of the treatment. During the experimental period, no deaths occurred in the treatment groups. Mice were sacrificed via CO2 exposure. After the mice were sacrificed, the tumor tissues were formalin fixed and paraffin embedded for immunohistochemistry.
Clinical samples and ethics statement
Clinical tissues were collected from patients who received curative surgery for oral squamous cell carcinoma at Cheng Kung University Hospital, Taiwan. Further adjuvant radio/chemo-radiotherapy was suggested depending on disease status according to head-and-neck cancer treatment guidelines. Histological sections of all cases were reviewed by the pathologist specializing in head and neck cancer, who was blinded to the clinical outcome. The following criteria were used to score the staining: 0: negative (no detectable staining); 1: weakly positive (light yellow staining in cytoplasm); 2: strongly positive (brown cytoplasmic staining). The study was approved by the Institutional Review Board of National Cheng Kung University Hospital (NCKUH-10307004).
Kaplan-Meier analysis
The cancer specific survival was analyzed with the Kaplan-Meier method by SPSS Ver.17. Univariate analyses of patient and disease characteristics were tested by the log-rank test. Multivariate analysis was calculated by the Cox regression model. P values less than 0.05 were considered as statistically significant.
Statistical analysis
We evaluated the differences in the differences in continuous variables (presented as mean ± standard deviation [SD]) between groups using the two-sample t-test or one-way analysis of variance carrying with a post-hoc Bonferroni test. We performed all statistical analyses using the SPSS 17.0 statistical software (SPSS Inc., Chicago, IL, USA). All statistical tests were performed at a two-sided significance level of 0.05.
Discussion
The primary objective of combination treatments is to exploit the synergistic effects between the agents, and these optimized combination regimens might help to broaden the applicability of single agent treatment. We provided evidence that combined bortezomib and IR treatment resulted in a synergistic cell-killing effect in SAS human oral cancer cells in vitro and it possessed potent antitumoral activity in a xenograft animal model in vivo (Figs.
1 and Additional file
2: Figure S2). Furthermore, bortezomib alone or in combination with IR caused no detectable toxicity as determined by either biochemical examination or in terms of the loss of body weight (Additional file
2: Figure S2a and Additional file
3: Table S1). A recent review article indicated that the therapeutic effect of radiotherapy can be influenced by regulating autophagy, and the induction of autophagy has become more popular than the use of autophagy inhibitors [
33]. Radiation sensitization can be enhanced when the level of autophagy is higher than the tolerance of cells, which can lead to autophagic cell death [
34]. It has been shown that an increase in autophagic flux, which is defined as the quantity of degradable material transported from the autophagosome to the lysosome, may be the key factor that modulates autophagy towards cell death [
16,
35]. We found that combining bortezomib with IR treatment induced a significant amount of autophagy but only a small amount of apoptosis in SAS, a human oral cancer cell line (Fig.
2). Our in vivo study further demonstrated that the induction of autophagy could be observed in SAS xenograft tumors, in which LC3 was increased following a combined treatment of bortezomib with IR (Additional file
2: Figure S2). Pre-treatment with 3-MA, an inhibitor of autophagy, followed by combined bortezomib with IR treatment of SAS cells, resulted in a significant reduction in cytotoxicity (Fig.
2f). In addition, our current results also showed the combined treatment-induced accumulation of LC3-II in the presence of BAF (Fig.
2h). Thus, the combined treatment of bortezomib with IR increased autophagic flux and induced autophagic cell death in human oral cancer cells. Previous studies have demonstrated that p62 decreased levels can be observed when autophagy is induced [
36]. However, data accumulated by us and others have revealed that induction of autophagy was accompanied by increased expression of p62 [
20,
37,
38]. After a rapid increase in p62/SQSTM1 expression upon the combination of the phorbol ester PMA and p38MAPK inhibitor SB202190 stimulation, the level of p62 gradually decreased at 24 h [
37]. In the present study, we found that the expression levels of the p62 proteins increased with IR and bortezomib alone or in combination at 24 h.
A deeper understanding of the unique functions of autophagy is required for the development of more effective treatments of human cancer [
15]. As indicated in our previous study, TRAF6 could serve as a direct E3 ligase for K63-linked ubiquitination of oncogenic Akt, leading to its membrane recruitment and phosphorylation in cells treated with insulin-like growth factor-1 (IGF-1). [
11] Since Akt activation is a well-known key event in tumorigenesis, TRAF6 represents a potentially important therapeutic target for human cancers. More recently, we also demonstrated that a proteasome inhibitor (MG132) combined with IR enhanced its anti-pancreatic tumor effects through the induction of autophagy and the downregulation of TRAF6 [
7]. Furthermore, a reduced TRAF6 protein level was found in bortezomib-induced autophagy and subsequent cytotoxicity in myelodysplastic syndrome/acute myeloid leukemia [
18]. In addition, bortezomib not only affects TRAF6 but also other targets. Previous research has shown that bortezomib inhibits NF-kappaB activity in malignant mesothelioma cells and induces cell cycle blockade and apoptosis [
39]. Bortezomib induced mitochondrial apoptosis through regulation of BAK and NOXA [
40]. Nonetheless, the potential role of TRAF6 as a therapeutic target in OSCC during chemoradiotherapy combining a proteasome inhibitor with IR has never been reported before. We demonstrated that the TRAF6 expression level in the 3 human oral cancer cell lines was higher than in hNOK, a human normal oral keratinocytes cells (Additional file
2: Figure S1a). Bortezomib alone or in combination with IR inhibited TRAF6-mediated NF-κB/Akt activation through polyubiquitination, and reduced TRAF6 protein levels through autophagy-mediated lysosomal degradation (Figs.
3,
4 and Additional file
2: Figure S1). Ubiquitin chains not only generate signals that induce acute degradation of tagged proteins by the proteasome but they also play an essential role for proteins and their associated subcellular organelles by regulating autophagic degradation [
41,
42]. We found an increased ubiquitination of TRAF6 and consequent autophagic degradation in OSCC cells treated with bortezomib alone or in combination with IR (Fig.
3). Cells may employ multiple types of heterotypic chain linkages in both autophagy and UPS; however, many questions remain to be answered. For example, what is the molecular decision-making process when the same protein substrates are delivered to the autophagy or UPS? How do cells modulate the activities of the autophagy and UPS in response to various stresses? Understanding the functional relationship between the autophagy and UPS will contribute to the development of therapeutic strategies through modulating proteostasis and removal of pathogenic protein species [
8].
In the current study, transfection of TRAF6 shRNA resulted in decreased NF-κB/Akt activation, increased autophagy and reduced viability of SAS cells compared with control shRNA (Figs.
5a and
b). Using the TRAF6 knockdown SAS cancer xenograft model system, we were able to demonstrate a decreased tumor growth rate and decreased TRAF6 protein expression in the cancer xenografts when compared with parental SAS cancer xenografts (Fig.
5). The overexpression of TRAF6 oncoprotein has been reported to be associated with increasing tumorigenicity and metastasis of esophageal squamous cell carcinoma in vivo [
43,
44]. Tao et al. also found that high TRAF6 expression levels had significantly poorer survival when compared with those with low levels (
P < 0.05) in 135 patients with colon cancer [
45]. However, the clinical impact of TRAF6 in oral cancer patients has seldom being investigated. In our current study, risk groups defined by expression of TRAF6 and level of differentiation were the only predictors of cancer specific survival of OSCC patients. Therefore, TRAF6 is potentially a prognostic marker of OSCC, especially in patients with well differentiated tumors.