Prognostic Significance of Altered Blood and Tissue Glutathione Levels in Head and Neck Squamous Cell Carcinoma Cases

Head and neck squamous cell carcinoma (HNSCC) is a major cancer problem in Southern China, Pakistan, Thailand, India and Brazile. Overall prevalence of this disease is 12.3 per 100,000 out of total cancers treated in Pakistan (Bray et al., 2011; Pruegsanusak et al., 2012; Masood et al., 2013). HNSCC occurs through a complex multistage process which may involve exposure to a combination of carcinogens from cigarette smoking, alcohol consumption and chewing tobacco (Loyha et al., 2012). Although the mechanisms by which these risk factors enhance cancer risk remains largely unclear, however oxidative damage is believed to play a key role in carcinogenesis (Burlakova et al., 2010). Free radicals generated in tobacco smoke have been implicated in tobacco-induced carcinogenesis at numerous sites within the HNSCC (Mahjabeen et al., 2011). Among these radicals, reactive oxygen species (ROS) are potential carcinogens that facilitate mutagenesis, tumor promotion and progression. Although the extent of the contribution made by oxidative DNA damage has not been well defined, it appears that ROS-induced DNA damage can not only


Introduction
Head and neck squamous cell carcinoma (HNSCC) is a major cancer problem in Southern China, Pakistan, Thailand, India and Brazile. Overall prevalence of this disease is 12.3 per 100,000 out of total cancers treated in Pakistan (Bray et al., 2011;Pruegsanusak et al., 2012;Masood et al., 2013). HNSCC occurs through a complex multistage process which may involve exposure to a combination of carcinogens from cigarette smoking, alcohol consumption and chewing tobacco (Loyha et al., 2012). Although the mechanisms by which these risk factors enhance cancer risk remains largely unclear, however oxidative damage is believed to play a key role in carcinogenesis (Burlakova et al., 2010). Free radicals generated in tobacco smoke have been implicated in tobacco-induced carcinogenesis at numerous sites within the HNSCC (Mahjabeen et al., 2011). Among these radicals, reactive oxygen species (ROS) are potential carcinogens that facilitate mutagenesis, tumor promotion and progression. Although the extent of the contribution made by oxidative DNA damage has not been well defined, it appears that ROS-induced DNA damage can not only initiate carcinogenesis, but can also facilitate tumor progression (Kruk and Duchnik, 2014;Taya et al., 2014). Production and removal of these ROS within the cell is controlled by specific antioxidant defensive mechanisms (Oyagbemi et al., 2009;Nagamma et al., 2011).
These antioxidant defensive mechanisms include the enzymes superoxide dismutase, catalase, and glutathione peroxidase (GPx), as well as nonenzymatic compounds such as α-tocopherol, β-carotene, vitamin C, and glutathione (Narasimhan et al., 2011;Manoharan et al., 2012;Yao et al., 2012). The major intracellular antioxidant glutathione represents a critical line of defense against oxidative stress and is considered to be a major thiol-disulphide redox buffer of cell. Average GSH concentration in the cytosol is 1-11mmol/l. This is far higher than most other redox active compounds (Schafer and Buettner, 2001). Therefore the redox state of glutathione reduced (GSH), glutathione oxidized (GSSG) has been used as indicator for the redox environment of the cell (Charles et al., 2006). Although there are several methods for estimating the intracellular redox index, the most important technique is the measurement of intracellular concentration of glutathione total (GSHt), glutathione reduced (GSH), glutathione oxidized (GSSG) and the calculation of redox index by GSH/GSSG ratio (Gherghel et al., 2013). The intracellular GSH redox homeostasis is strictly regulated to govern cell metabolism and protect cells against oxidative stress and head and neck carcinogenesis (Richie et al., 2008).
The purpose of these experiments was to investigate extensive array of redox status indices: glutathione total (GSHt), glutathione reduced (GSH), glutathione oxidized (GSSG) and the calculation of redox index in HNSCC blood and tumor samples. Glutathione level and redox index were also correlated with clinical/histopathological parameters to assess the role of glutathione as biomarker of HNSCC.

Apparatus
PerkinElmer series 200 HPLC system (PerkinElmer Life and Analytical Sciences 710 Bridgeport Avenue Shelton, CT 06484-4794 USA) was used in this research. This HPLC system consists of analytical pump, autosampler, column oven, photo diode array (PDA) detector, fluorimeter, totalchrom chromatography data system (CDS) software.

Samples collection and patient identifications
The present population-based case-control study was conducted with a prior approval from ethical committees of both COMSATS Institute of Information Technology and collaborating hospitals. Two HNSCC patient groups were recruited in this study. The cohort 1 consisted of blood of 200 HNSCC cases and 200 matching normal healthy individuals as controls. The observed mean age of patients and controls was 45 (±16.35) and 43 (±32.23) years respectively. Non-significant difference was observed in case of age (p=0.9) and smoking status (p=0.4) between cancer cases and controls. In addition to this, a non-significant difference was observed in HNC patients for area of cancer (p=0.08). The selection procedure for patients included confirmed histological diagnosis of HNSCC, no preoperative therapy and availability of complete follow-up data. The inclusion criterion for the controls was absence of prior history of cancer or precancerous lesions. Patients and controls suffering from any other familial disease (diabetes, blood pressure and cardiovascular impairment) were also excluded from this study.
Cohort 2 included tumors of 50 HNSCC patients who had undergone surgery at Military Hospital (Rawalpindi, Pakistan) between 2009 and 2012. Demographic and clinical characteristics of study cohort 2 showed that median age of HNC cases and controls was 60 years with a range of 20-80. 50 tumors included 24 (48%) cases of SCC of the larynx, 12 (24%) cases of the pharynx and 14 (28%) cases of SCC of the oral cavity .Most of the patients were in stage III of clinical and T staging with stage I of N staging. Most of the HNC patients had well differentiated tumors. All tumors were histologically confirmed as squamous cell carcinoma of head and neck and were graded as well differentiated, moderately differentiated and poorly differentiated. Samples of tumor core, the invasive edge of tumor and microscopically healthy mucosa (control) were obtained from each surgical section. Presence of tumor cells in the collected tissues was rectified by examination of frozen sections following haematoxylin and eosin stain by a consultant pathologist. Samples of control mucosa were obtained from macroscopically confirmed uninvolved healthy area >2cm away from the tumor.

Blood sample preparation
Blood sample was prepared by centrifugation at 600g for 10 minutes. Separated aliquot was washed twice with phosphate buffer saline (PBS). Aliquot of the blood sample (200μl) was taken and 200μl of 5% 5-Sulfosalicylic acid (SSA) solution was added. Each sample was vortexed vigorously and left for 10min at 2-8°C. After this the whole mixture was centrifuged at 10,000g for 10 min.

Tissue sample preparation
The tissue sample (100mg) was homogenized in 3ml buffer solution prepared by mixing of 10mM EDTA, 50mM NACLO4 and 0.1% H3PO4. Total protein was precipitated by adding 0.5ml metaphosphoric acid (5%w/v). The samples were vortexed for 20 mins and centrifuged at 8000g for 10mins at 4˚C. The supernatant was then transferred to auto-sampler vials of HPLC for analysis.

Glutathione total (GSHt) assay
The sample was diluted with internal standard, reaction buffer and derivatization solution (o-phthalaldehyde precolumn) was mixed together. This mixture was incubated for 20 mins at 60°C. 100μl reaction buffer was added to this mixture which was incubated for 10 minutes at 2-8°C and then centrifuged for 10 minutes at 10,000g. 100μl of supernatant was added to 200μl reaction buffer in auto sampler-vials. 20μl of sample was then taken and injected in HPLC-system. One run lasted for 4 mins. The chromatograms were recorded by a fluorescence detector. The quantification was performed with EDTA-blood calibrator; the concentration was calculated by internal standard method and a standard curve of GSHt was then generated using a linear regression program (Microsoft Excel; Microsoft cooperation, Redmond, WA) as reported earlier (Gherghel et al., 2005).

Glutathione reduced (GSH) assay
Reduced glutathione was treated with 100μl reaction buffer and 100μl derivatisation solution (o-phthalaldehyde precolumn). This mixture was incubated for 20 mins at 60°C. 100μl reaction buffer was added to this mixture, incubated for 10 mins at 2-8°C and centrifuged for 10 mins at 10,000g. 100μl of supernatant was added to 200μl reaction buffer in auto sampler-vial and 20μl of sample was then injected in the HPLC-system. One run lasted for 4 mins. The chromatograms were recorded by a fluorescence detector.The quantification was performed with EDTAblood calibrator; the concentration was calculated by internal standard method and a standard curve of GSHt was then generated using a linear regression program (Microsoft Excel; Microsoft cooperation, Redmond, WA) as reported previously (Gherghel et al., 2005).
The oxidized glutathione levels (GSSG= GSHt-GSH) and the redox index (defined as the GSH/GSSG ratio) were then calculated. Blood and tissue glutathione levels (GSHt,GSH, GSSG and redox index) were measured in µmol/l and nmol/mg, respectively.

Statistical analysis
The results for the groups are expressed as mean ±SE. Multivariate analysis was performed to determine the influence of age, gender, smoking status and area of cancer on HNSCC blood samples and tissue glutathione levels. For statistical analysis, the experimental values were compared with their corresponding controls. P-values were calculated by chi-square test and ANOVA. Pearson correlation coefficient was used to calculated the correlation among glutathione levels and clinical/ histopathological parameters (*indicates the level of significance of pearson correlations as calculated by SPSS).
Multiple regression analysis revealed that significantly lower level of GSHt (p<0.01) and GSH (p<0.01 below 45 years and p<0.007 above age 45 years) was observed in HNSCC in all age groups when compared with control groups. In case of gender, significant lower levels of GSHt, GSH, GSSG and redox index were observed in male and female HNSCC patients compared to male and female control samples. Moreover, significantly lower level of GSHt, GSH, GSSG and redox index was observed in smoker and non smoker HNSCC patients compared to smoker and non smoker control samples (Table 2). Significantly lower levels of GSHt (p<0.002), GSH (p<0.001), GSSG (p<0.004) and redox index (p<0.001) was observed in different areas of cancer.
A forward stepwise, pearson correlation was performed which revealed that age has a significant and negative correlation with GSH (r=-0.233*; p<0.05). No significant correlation was observed between age and other forms of blood glutathione (GSHt, GSSG, redox index) in HNSCC patients. In case of gender and smoking status, no significant correlation was observed with different forms of blood glutathione (GSHt, GSH, GSSG and redox index) in cancer patients. Additionally to explore glutathione-glutathione relationship, we observed a positive Pearson correlation between GSHt versus GSH (0.800***; p<0.0001) and GSHt versus GSSG (r=0.636***, p<0.0001) in HNSCC patients. A positive correlation was observed between GSH versus redox index (r=0.343**; p<0.01) and negative correlation between GSSG versus redox index (r=-0.287**; p<0.01). An apparent inverse correlation was observed between GSHt versus redox index (r=-0.171; p=0.09) and GSH versus GSSG (r=-0.046; p=0.72). However, these correlations were statistically non-significant (Table 4).

Discussion
Free radicals are implicated in the pathogenesis of a multistage process of head and neck carcinogenesis, which are proposed to cause DNA base alterations, strand breaks, damage to tumour suppressor genes and an enhanced expression of proto-oncogenes (Nisha et al., 2008;Mahjabeen et al., 2011). In addition to these burst of reactive oxygen species (ROS) and reactive nitrogen species (RNS) has also been implicated in the development of different cancers including HNSCC (Kruk and Duchnik, 2014;Valluru et al., 2014). To safeguard the integrity of genome, humans have well defined defense mechanism to prevent body against stress and ROS, among which the most important one is glutathione which directly reduces reactive oxygen species and acts as first line of defense against oxidative stress (Rasheed et al., 2007;Nisha et al., 2008).
In this present study we assessed the glutathione total (GSHt), glutathione reduced (GSH), glutathione oxidized (GSSG) and redox index in HNSCC and compared with controls. In first part, we observed that mean blood glutathione levels are significantly reduced in head and neck cancer compared with respective controls. Our results are in agreement with earlier studies which have demonstrated that glutathione levels differ in patients of head and neck malignancies and this depletion could be due to imbalance in free-radical and antioxidant homeostasis in body fluids during carcinogenesis (Singh et al., 2008;Waly et al., 2012;Shetty et al., 2013). The other plausible explanation for decreased levels of glutathione observed in head and neck cancer patients is possibly GSH, being exhausted as a main low molecular weight antioxidant in cellular protection (Komatsu et al., 2008;Trachootham et al., 2009). Increased utilization of glutathione can act as an indicator of increased oxidative stress and ROS in HNSCC patients. Our results are in agreement with earlier studies which showed that levels of glutathione decrease in blood of the breast cancer when compared with control subjects (Yeh et al., 2006).
In second part of this study we used 50 HNSCC tissue samples and 50 non cancerous control tissue samples and observed significantly higher levels of GSHt and GSH in HNSCC tissue samples compared with control tissue samples. No significant difference in levels of GSSG and redox index was observed in HNSCC tissue samples when compared with non cancerous control tissues. The elevated levels of GSHt and GSH have also been observed in chronic myelogenous leukemia, acute myelogenous leukemia, or chronic lymphocytic leukemias (Singh et al., 2008) which support our results. The increased levels of glutathione in HNSCC tissues may be result of stress and excess need of glutathione in turn increase tumor load, which reduce anti-apoptotic mechanism and cell death and results in un-controlled cell growth and division resulting in cancer pathogenesis (Balendiran et al., 2004). The increased levels of glutathione in tumor resist radiotherapy and anti-cancer drug resistance (Gupta et al., 2010). Present results from this study have shown that levels of glutathione are decreased in blood of head and neck cancer patients and elevated in tissues when compared to the respective controls. These decreased and elevated levels of glutathione may deregulate cytochrome P450 (Wahid et al., 2013) and glutathione expression which are important mechanisms involved in carcinogenesis and mutagenesis.   GSH is among the most efficient substance which can be used by cells and tissues in their defense against oxidative stress and is representative of redox environment of the cell (Cacciatore et al., 2010). Age and disease, however, act to decrease the amount of GSH available in the organism (Erden-Inal et al., 2002). Our results have shown a significantly negative correlation between age and both GSH levels and redox index in both study groups. In HNSCC patients, age is strongly associated with increased oxidative stress and reduced antioxidant stress. In case of gender and glutathione levels, no difference was observed between male and female patients with respect to blood glutathione levels in either of these study groups. It seems that gender has a less consistent effect on blood glutathione levels than other confounding variables such as age and more research is necessary to further address this issue.
Our study also showed that there was a decrease in tissue GSH level and redox index as the pT-stage, C-stage and grade of HNSCC tumor increased. This decreased level of GSH with increase in stage of HNSCC may be due to increased utilization of GSH and increased level of oxidative stress in advance stages of carcinogenesis (Singh et al., 2008). This would imply that the antioxidant defense system of these patients is defective and unable to cope with free radicals' overload. We also observed that decreased GSH results in increased GSSG with a concomitant decrease in redox index in both HNSCC study cohorts. GSH is depleted in various tissues with increase in levels of GSSG which reduce the GSH/ GSSG ratio. Increased GSSG concentration in tissues and accelerated release from the cell is considered to be a signal of oxidative stress and could contribute to free radical diseases such as HNSCC (Muchova et al., 2007). Variations in GSH, GSSG level and redox status in carcinogenesis are thought to function in control of fundamental processes such as gene expression, cell proliferation or apoptosis. Thus if redox index can be detected in vivo in association with pathology or toxicity, this finding may provide an important link to underlying mechanism.
In conclusion our study has demonstrated a significant association between blood glutathione levels and HNSCC. This study also confirms the dysregulation of glutathione level in HNSCC when compared with normal tissues. There is increased dysregulation in more metastatic disease when compared with primary tumors. The results obtained from this study suggest that altered blood and tissue glutathione levels may be used as prognostic marker and ancillary screening tool for individual with risk of HNSCC.