Litchi seed extracts diminish prostate cancer progression via induction of apoptosis and attenuation of EMT through Akt/GSK-3β signaling

Litchi (Litchi chinensisSonnnerat, Sapindaceae), known as Chinese Cherry, is a subtropical fruit tree originating from southern China. Litchi seed extracts have diverse pharmacological effects, including anticancer. However, its anticancer effects and mechanisms on prostate cancer have not been determined. In this study, we used n-butyl alcohol extract of Litchi seed (NLS) to treat prostate cancer PC3, DU145, RM1 and C4-2B cells. NLS induced a significant decrease in cell viability and clonogenic growth in a dose-dependent manner. NLS induced cell apoptosis and cell cycle G1/S phase arrest by inactivating Akt signaling pathway, which were associated with activation of mitochondrial caspase-dependent apoptotic cascades, up-regulation of cyclin-dependent kinase (CDK) inhibitors p21 and p27, and inhibition of correlated cyclin/CDK network. In addition, NLS treatment significantly decreased cell migration and invasion via phenotypic inversion of EMT, correlated with increased expression of E-cadherin and β-catenin, and decreased expression of vimentin and snail, which is partially attributed to inhibiting Akt/GSK-3β signaling pathway. Finally, PC3 xenograft nude mice treated with NLS in vivo showed a significant decrease in tumor size without toxicity. These findings suggest that NLS has potential for development into a safe and potent alternative therapy for prostate cancer patients.


Results
Phytochemical analysis of NLS. We analyzed the phytochemical ingredients of NLS using HPLC-MS. Figure 1 showed the base peak chromatogram of NLS by HPLC-MS. All compounds were characterized by the interpretation of their mass spectra obtained by the MS/MS and were further aligned with published data from literatures. The 13 compounds presented were summarized in Table 1, including flavonoids, phenols, proanthocyanidins, sesquiterpenes, fatty acids, and so on.

NLS inhibited growth and viability of PCa cells.
To investigate the effect of Litchi seed extract on proliferation of PCa cells, PC3, DU145, C4-2B, and RM1 cells were treated with four fractions of Litchi seed extract (0 to 250 μ g/ml) for 24, 48 and 72 h, and viability was determined by MTS assay, normal prostatic epithelial cell line RWPE-1 was used as controls. We observed that the n-butyl alcohol and ethyl acetate fractions inhibited cell growth in a dose-dependent manner. More importantly, n-butyl alcohol fraction (NLS) was more potent in inhibiting proliferation of PCa cells compared to the ethyl acetate fraction ( Fig. 2A and Supplementary Fig. S1). The IC 50 values of NLS for PC3 cell were 147.3, 128.14 and 111.97 μ g/ml; for DU145 cell, 114.31, 104.09 and 105.23 μ g/ml; for RM1 cell, 116.52, 107.10 and 101.41 μ g/ml; and for C4-2B cell, 119.97, 105.46 and 122.78 μ g/ml for 24 h, 48 h and 72 h respectively. However, the viability of normal prostatic epithelial cell line RWPE-1 was not inhibited by NLS and IC 50 values were over 250 μ g/ml ( Fig. 2A).
In addition, clonogenic assay was performed to further validate the inhibiting effect of NLS on PCa cells. As shown in Fig. 2B,C, PC3 and DU145 cells treated for 8 days with NLS at 15, 30 and 60 μ g/ml showed a significant dose-dependent inhibition of colony formation compared to dimethylsulfoxide (DMSO)-treated controls. These results demonstrated that NLS showed potent growth inhibition towards PCa cells.

NLS Induced the Apoptosis of Prostate Cancer Cells.
To determine the primary driver for the decreased cell viability, we analyzed the apoptosis in NLS treated prostate cancer cells with flow cytometry. We observed significant dose-and time-dependent increase in the population of apoptotic cells in both PC3 and DU145 cell lines when treated with NLS (Fig. 3A,B). The apoptotic cell populations in PC3 and DU145 cells treated with high-dose NLS for 72 h, were increased to 55.3% and 28.2%, respectively, compared to that in the DMSO-treated control cells (6.4% and 5.8%).
To further explore the mechanisms underlying NLS-induced apoptosis of prostate cancer cells, we detected the expression of a panel of proteins involved in cellular apoptosis using western blotting. NLS treated cells showed increased expression of activated Caspase-3, associated with PARP cleavage (Fig. 3C), further validating that apoptosis was the primary mechanism of NLS-induced growth inhibition in PCa cells. To investigate whether NLS induced apoptosis is mediated through activation of extrinsic or intrinsic pathways, we next examined the expression of Caspases-8 and Caspases-9 in NLS treated cells. We found NLS led to an increase of cleaved Caspase-9 in cells associated with enhanced expression of cleaved Caspase-7, yet failed to induce the changes of caspase-8, which is involved in the extrinsic apoptotic pathway (Fig. 3C). These results suggested that NLS activated the intrinsic apoptotic pathway.
Next, we assessed the effect of NLS on the Bcl-2 family of proteins involved in apoptosis. Both PC3 and DU145 cells treated with NLS showed an increase in the expression of pro-apoptotic Bax, Bim and Puma. Moreover, NLS decreased the expression of anti-apoptotic Bcl-2 protein in DU145 cells. Furthermore, NLS inhibited significantly the phosphorylation of Akt and GSK-3β in both PC3 and DU145 cell lines (Fig. 3D), which are consistent with the roles for Akt and GSK-3β in the regulation of apoptosis. These data demonstrated that NLS induced cell apoptosis in prostate cancer cells by inactivating Akt signaling pathway and activating mitochondrial caspase-dependent intrinsic apoptotic cascades.

NLS Induced Cell Cycle Arrest in Prostate Cancer Cells.
Interestingly, we observed that NLS induced cell cycle arrest in prostate cancer cells as well. Propidium iodide (PI) staining for PC3 cell cycle progression showed that the majority of the cells were arrested in the S phase (Fig. 4A,B) after being treated with NLS. The increase of S phase cell population in PC3 was accompanied by a simultaneous decrease in the G1 phase cell population. In the DU145 cells, NLS treatment led to a significant increase in the proportion of cells in the G1 phase accompanied by a simultaneous decrease in the G2 phase cell population (Fig. 4A,B). Therefore, the inhibitory effects of NLS on cell proliferation and clonogenicity might partially attribute to cell cycle arrest at G1/S phase.
To determine which protein molecules might be responsible for NLS-induced G1/S arrest, we examined the levels of several cell cycle regulators during the G1/S transition. Western blotting results showed that NLS treatment elevated levels of the CDK inhibitors p21 and p27, decreased levels of CDK2, 4, 6 and cyclinD1, E2 in PC3 and DU145 cells compared with those in control cells (Fig. 4C). Taken together, our data demonstrated that NLS induced cell cycle arrest at G1/S phase through downregulation of cyclin-dependent kinases (Cdks) and upregulation of CDK inhibitors.    NLS were not able to move toward the wounded area and left an obvious gap (Fig. 5A). We also noted that the NLS remarkably reduced the number of invaded cells stained with crystal violet compared to the control groups ( Fig. 5C,D). These results indicated that NLS had metastatic inhibitory effects on human PCa cells.

NLS Reduced migration and invasion capabilities in
To explore the mechanism of NLS on anti-metastasis, we detected the expression of β -catenin, E-cadherin, Vimentin and Snail in PC3 and DU145 cells, which are known factors involved in epithelial-mesenchymal transition (EMT) and cancer metastasis. Western blotting results showed that the expression of E-cadherin was significantly up-regulated by NLS, and the level of β -catenin was also increased after 24 h and 48 h of NLS treatment. On the contrary, the expression of Vimentin and Snail was remarkably down-regulated in the NLS-treated cells compared to control cells (Fig. 5E). These results indicated that NLS reduced migration and invasion capabilities in PCa cells via phenotypic inversion of EMT.

NLS inhibited growth of PC3 xenografts in nude mice.
To evaluate the anti-PCa activity of NLS in vivo, we established human PC3 xenograft in nude mice by subcutaneous injection of PC3 -luc cells. The tumor volume and body weight were examined to evaluate the effects of the treatment. After 24 consecutive days treatment with NLS, xenografts of both NLS-treated groups had smaller tumor sizes compared to those in control group (Fig. 6A,C). As shown in Fig. 6A, NLS shrunk tumor growth to approximately 52.99% at the dose of 52 mg/kg and to 28.01% at the dose of 26 mg/kg compared to the vehicle control on Day 29. In addition, there was no significant loss of body weight during the NLS treatment (Fig. 6B), suggesting that NLS was not overtly toxic. The bioluminescence imaging of PC3 xenograft tumors in different groups at the end of experiments showed consistent results (Fig. 6C). Our results demonstrated that NLS efficiently inhibited tumor growth in PC3 xenograft models without major toxicity.

Discussion
PCa is the most commonly diagnosed solid tumor and the second leading cause of cancer related death in men in western countries 3 . Despite remarkable advances in early diagnosis, it often escapes timely detection. Mortality remains high in advanced PCa due to relapse after androgen deprivation therapy and chemotherapy. Thus, effective treatment strategies are urgently needed. Alternative therapies using oriental herbal formulae, such as PC-SPES, KMKKT and TBS-101 as well as herbal extract components suggest new approaches to the management of anti-PCa 15,16 . Since carcinogenesis is a multistage process, it could be problematic by using a single agent. Therefore, the synergistic interactions among the active ingredients present in plant-based extracts may help to protect against cancer.
Litchi seeds have been used alternatively in traditional Chinese medicine for the treatment of urologic neoplasms such as prostate cancer, bladder cancer and renal carcinoma etc 9,10 . Previous studies have indicated Litchi seed extract possesses anti-cancer activities against various tumor cells, including lung, cervical, colorectal and liver cancer 6 , yet its anti-PCa potential has not been assessed. We employed HPLC-MS to detect the ingredients of NLS and found the presence of flavonoids, phenols, proanthocyanidins, sesquiterpenes and fatty acids in NLS (Table 1), which are consistent with published reports [17][18][19][20] . Xu et al. concluded that Litchi seeds are mainly composed of proanthocyanidins and lichitanin A 20 . One major structure of lichitanin is epicatechin that exhibited strong anti-cancer effect in vitro 21 , suggesting the potential role of NLS on cancer prevention and treatment. It is also well known that plant-based extracts rich in proanthocyanidins have been regarded as nutrients with cancer preventive characteristics 22 . Both flavonoids and phenols usually have a wide range of pharmacological activities, such as anti-cancer, anti-inflammatory, due to their antioxidant property 23,24 . These findings emphasized the therapeutic potential of NLS, likely attributable to the presence of proanthocyanidins, flavonoids and phenols.
In the present study, we demonstrated that NLS inhibited the growth of four PCa cell lines (PC3, DU145, RM1 and C4-2B) in a dose-dependent manner using MTS assays ( Fig. 2A). Moreover, clonogenic assay further validated the inhibiting effect of NLS on PCa cells (Fig. 2B,C). Meanwhile, NLS exhibited the minimal toxicity on the viability of normal prostatic epithelial cell RWPE-1 ( Fig. 2A), suggesting selective activity of NLS against malignant cells. Moreover, we recapitulated NLS's anti-tumor efficacy in vivo, and there was no significant decrease in body weight and other associated adverse effects in mice during its treatment (Fig. 6). These findings imply NLS could be developed into a safe and potent alternative therapy for PCa patients.
We then systematically evaluated the molecular mechanisms of the anti-cancer activity of the NLS against prostate cancer cells. Previous studies showed that Litchi seeds extract could induce significant apoptosis or cell cycle arrest in hepatoma cells 13 , colorectal carcinoma cells 7 and sarcoma cells 25 . Our data confirmed that NLS inhibited PC3 and DU145 via both inducing apoptotic cell death and arresting cell in G1/S phases, as demonstrated by flow cytometry analysis (Figs 3 and 4).
It is well known apoptosis may be executed via the extrinsic pathway or through an intracellular cascade of events that involves mitochondria. In this case, caspases, a family of cysteine proteases, are the central regulators of apoptosis. Activation of initiator caspases, such as caspase-8/9 cleave and activate downstream effector caspases (including caspase-3/7), which in turn execute apoptosis 26 . The members of the Bcl-2 family also play an essential role in regulation of apoptosis 27 . Some family members promote apoptosis (e.g., Bax and Bim), while others inhibit it (e.g., Bcl-2) 28,29 . In the present study, we found that treatment of NLS increases the activities of caspase-3/7/9, and PARP, enhances the expression levels of Bax, Bim and Puma, down-regulates the expression level of Bcl-2, yet fails to induce the changes of caspase-8 (Fig. 3C,D), which characterizes the extrinsic pathway of apoptosis. These results suggest that NLS induces cell death in PCa cells through the mitochondrial apoptotic pathway.
To determine which protein might be responsible for NLS-induced G1/S arrest, we first examined the levels of several cell cycle regulators during the G1/S transition. In the late G1 phase, cyclin D-CDK4/6 kinase complex initiates phosphorylation of Rb and its release from E2F results in the progression of G1/S transition. This process is enhanced by cyclin E-CDK2 complex, but inhibited by cyclin-dependent kinase inhibitors p21and p27 30   Strikingly, we observed that expression of cyclinD1, cyclin E2 and CDK 2/4/6 was decreased while expression of p21 and p27 was increased in NLS treated cells compared to control cells (Fig. 4C). Our results suggest that NLS induces G1/S arrest through increasing the levels of p21 and p27 and decreasing the levels of cyclin D1, cyclin E2, and CDK2/4/6.
EMT is a crucial step in tumor progression and plays vital roles in cancer invasion and metastasis. In general, increased migration and invasion are positively correlated with EMT, which is characterized by repression of epithelial markers (e.g., E-cadherin) and induction of mesenchymal markers (e.g., Vimentin and Snail) 32 . β -catenin, one of the components of the adherens junction, binds directly to E-cadherin to act in cell-cell adhesion. Loss of functional E-cadherin, and thus of adherens junction-mediated cell-cell contacts was reported to be the first step of tumor invasion and metastasis in various cancers 33 . Snail encoded by SNAI1/SNA gene binds to the promoter of E-cadherin gene and represses its transcription 34 . Knockdown of Snail inhibited tumor cell invasion and migration 35 . In this study, we observed that NLS obviously increases levels of E-cadherin and β -catenin, and decreases levels of Vimentin and Snail in both PCa cells (Fig. 5E). These findings suggest that NLS negatively regulates EMT, and significantly attenuates cell migration and invasion in vitro.
To further understand the cellular mechanisms underlying our findings, we examined the status of various oncogenic signaling cascades associated with cancer cell survival, proliferation and migration. Among the examined biochemical pathways, we found that inactivation of Akt signaling pathway is notably associated with NLS's anti-PCa effect. Akt pathway is a typical pathway regulating cell survival, proliferation and migration in human cancer 36 , and is associated with induction of EMT 37 . Akt executes its effects in cells by phosphorylating a variety of downstream substrates, including GSK-3β , Bad, Bim, Bax, Caspase9, p21, p27, and so on 38 . Hyperactivation of Akt pathway has previously been observed in human prostate cancer 39 . GSK3β is one of the major downstream targets of Akt and is involved in migration and proliferation of cancer. The active form of GSK3β is in the dephosphorylated state, when it is phosphorylated by Akt, GSK3β loses its activity 40 . More importantly, GSK-3β negatively regulates the transcription of Snail, a repressor of E-cadherin and an inducer of the EMT. Therefore, inactivation of AKT signaling pathway could promote GSK-3β activity and suppress the expression of Snail, which in turn-impedes the EMT 41 . In our study, we demonstrated that NLS significantly inhibits phosphorylation of Akt, phosphorylation of GSK-3β , and expression of Snail in PCa cells (Figs 3D and 5E). These suggest that NLS may inhibit EMT in PCa cells through attenuating Akt/GSK-3β signaling pathway.
In addition, Akt stabilizes cyclins D1 and E through phosphorylation and inactivation of active GSK-3β , which phosphorylates cyclins D1 and E to stimulate their translocation from nucleus to the cytoplasm for degradation 38 . Akt has also been shown to phosphorylate and directly inhibit p21 and p27, which then inhibit CDK2 and CDK4/6 complexes and promote further G1/S cell cycle transition 42 . The anti-apoptotic function of Akt has been linked to phosphorylation and inactivation of Bad, Bax, Bim and caspase-9, activation of NF-κ B, inhibition of cytochrome c release from mitochondria, overexpression of Bcl-2, and so on 43 . In this study, we observed inactivation of Akt and activation of GSK-3β in NLS treated PCa cells (Fig. 3D), which suggest that inhibition of Akt signaling and activation of GSK-3β partially contribute to the pro-apoptotic and G1/S phase arrest activity of NLS in PCa cells.
In summary, our results demonstrated the anti-cancer properties of NLS in PCa cells, both in vitro and in vivo. NLS inhibits PCa cell growth, migration and invasion through induction of apoptosis, G1/S phase cell cycle arrest and phenotypic inversion of EMT, which is realized by attenuating Akt signaling pathway. Therefore, NLS may serve as a potential adjuvant agent for the treatment of prostate cancer patients.

Methods and Materials
Reagents. RPMI 1640, DMEM, fetal calf serum, trypsin, and antibiotics were purchased from Gibco Litchi seed extracts. Litchi seeds were purchased from LBX pharmacy (Nanning, China) and processed at Guangxi Botanical Garden of Medicinal Plants (Nanning, China) to obtain Litchi seed extract. Briefly, dried Litchi seeds were ground to fine powder and then extracted with 75% ethanol at 1:10 (m/v) of solid-liquid ratio, the temperature was 60 °C and lasted for 1.5 h, repeated the procedure 3 times. The obtained crude ethanol extract was filtered through a vacuum Buchner funnel and Whatman No. 1 filter paper. The solvent was evaporated in vacuum to obtain the extract. Suspended the extract at 1:1(m/v) of raw material and deionized water and then extracted successively using petroleum ether (PE), ethyl acetate (EtOAc) and n-butyl alcohol (n-BuOH) on the basis of systematic solvent extraction method 44 . The supernatant and sediment were separated through decompress filtration, the residue was re-extracted. The polarity of extracts obtained was variable from small to large (PE, EtOAc, n-BuOH and Ethanol). Combined the obtained extracted solution and concentrated using a rotary evaporator under reduced pressure, then were freeze-dried. The yield rate of each fraction from crude starting materials was 1.33%, 1.06%, 2.35% and 3.51% (w/w) respectively. The extracts were stored in freezer under − 20 °C before used. and negative-ion ESI mode with the following flow injection parameters: ion spray voltage: +/− 4500 V; source temperature: 550 °C; curtain gas: 40 psi N2; GS1: 50 psi N2; GS2: 50 psi N2; heated interface. MS1 spectra were obtained using the enhanced mass spectrum (EMS) scan with a declustering potential (DP) of ±100 V and an entrance potential (EP) of ±10 V. All MS2 spectra were collected by enhanced product ion (EPI) scan with Q1 set at unit resolution and the triple-quadrupole (QqQ) scanning ranges were set based the target parent ions, with a step size of 0.12 Da (at a rate of 120 Da/s). Data acquisition was carried out with Analyst Software 1.6.2 software.
Cell proliferation assay. Cell proliferation/viability was determined by MTS assay. Briefly, PC3, DU145, RM1, C4-2B and RWPE-1 cells were seeded at 3,000 cells/well in 96-well plates and allowed to adhere overnight. Then cells were treated with various concentrations (0, 31.25, 46.875, 62.5, 93.75, 125, 187.5, 250 μ g/mL) of four fractions for 24, 48 and 72 hours, and DMSO was used as negative control. 20 μ L MTS solution was added to each well at the end points and incubated at 37 °C for 2 hours. Absorbance at 490 nm was detected using microplate reader. Cell viability rate (%) = treated group/control group × 100%. The half-maximal inhibitory concentration (IC 50 ) was calculated using the Bliss method. Data are the averages of three independent experiments.
Colony-forming assay. PC3 and DU145 cells were plated at a density of 400 cells per well in 6-well tissue culture plates. 24 hours later, the cultures were treated with NLS (15, 30 and 60 μ g/ml) or DMSO for 8 days and medium was replaced with fresh one containing NLS every 4 days. For determination of colony formation, cultures were fixed with 4.0% paraformaldehyde and stained with crystal violet. The number of colonies was counted after scanned with Canon CanoScan 9000F Color Image Scanner. The colony formation rate (%) = number of colonies/number of plated cells × 100%.
Migration assay. The inhibition of tumor cell migration by NLS was performed by wound-healing assay 45 .
Briefly, PC3 and DU145 cells were allowed to grow into full confluence in 6-well plates, and then a vertical wound was created with a 200 μ l pipette tip. The cell debris was removed and fresh complete medium with various concentrations of NLS (60 and 120 μ g/ml) was added. Cells were incubated at 37 °C and 5% CO 2 for 24 hours and photographed using a microscope at three time points (0, 12 and 24 hours). The area of migration was measured and analyzed with Image J software and the experiments were conducted thrice.
Invasion assay. Invasion assay was determined using transwell chambers (Corning New York, NY, USA) with 8 μ m pore membranes as described previously 46 . Briefly, after 24, 48 and 72 hours pre-treatment with NLS (30, 60 and 120 μ g/ml) in 6-well plate, tumor cells were collected and re-suspended in serum free cell culture media. A total of 5 × 10 4 cells in 100 μ l were added in the upper chamber, and 600 μ l of complete medium was added into the bottom of the lower chamber. The cells were allowed to invade for 22 hours at 37 °C and 5% CO 2 . Invaded cells were stained with 0.2% crystal violet and counted in five random microscopic fields. The percentage of inhibition was expressed using control well as 100%. Each assay was performed in triplicate.
Cell cycle and apoptosis analysis. Cell cycle and apoptosis were analyzed by flow cytometry (FACScan; BD Biosciences) as previously described 47 with some modifications. In brief, PC3 and DU145 cells were seeded at 2 × 10 5 cells in 12.5 cm 2 culture flasks and allowed to grow for 24 h before NLS exposure (60 and 120 μ g/ml) or DMSO exposure. Cells were harvested after 24, 48, and 72 h and washed twice with phosphate-buffered saline (PBS). For cell cycle analysis, the treated cells were incubated with propidium iodide (PI) according to the protocol of CycleTEST PLUS DNA Reagent Kit (BD PharMingen, San Diego, CA). Then the stained cells were analyzed by flow cytometry using FL-2A to score the DNA content of the cells. Approximately 30,000 cells were evaluated per sample. The population of each phase was calculated using Modfit software (Verity Software House, Topsham, ME). For the apoptosis analysis, Annexin V-FITC apoptosis detection kit I (BD PharMingen, San Diego, CA) was used to analyze cell apoptosis, following the manufacturer's protocols. Briefly, the treated cells were stained with Annexin V-FITC and PI for 15 min at room temperature in the dark before flow cytometry analysis. Approximately 50,000 cells were evaluated for each sample. The experiments were conducted thrice, and the results are reported as the mean of the three experiments.
Western blotting analysis. PC3 and DU145 cells were treated with NLS (120 μ g/ml) for 24, 48 and 72 hours, then cells were washed twice with ice-cold DPBS and incubated in lysis buffer in the presence of phenylmethylsulfonyl fluoride, protease inhibitor cocktail, phosphatase inhibitors cocktail 2 and 3 to obtain cell extracts. The protein concentration of the cell extracts was detected by BCA Protein Assay Reagent Kit (Rockford, IL, USA), and 50 μ g of protein was loaded to SDS-PAGE gels and were transferred to polyvinylidene fluoride membranes (PVDF). The blots were incubated with the appropriate concentration of specific primary antibody overnight at 4 °C. After washing, the blots were incubated with secondary antibody conjugated with peroxidase and an enhanced chemiluminescence detection system (Rockford, IL, USA) was used for detection. GAPDH or β -actin was used as an equal loading control.
Tumor xenografts in nude mice and drug treatments. In mouse xenografts, we utilized the luciferase-expressing prostate cancer cell line (PC3 -luc ) for in vivo bioluminescence imaging using the IVIS imaging systems (Caliper Life Sciences, Hopkinton, MA, USA). Two million PC3 -luc cells were injected subcutaneously into the right flank of each mouse (day 0). After the primary tumors had reached a mean volume of about 50 mm 3 , mice carrying tumor xenografts were arbitrarily divided into 4 groups (n = 10 per group) and treatment was started. Mice were orally treated with NLS (52 and 26 mg/kg) once daily for 24 consecutive days. Paclitaxel (20 mg/kg), used as reference drug for positive control, was administered intraperitoneally once per week. Body weight and tumor size of the mice were monitored every 3 days, and in vivo imaging was performed once a week. Tumor volumes (V) were calculated using the formula V = ½ (length × width 2 ). During the experiment, the animals did not show any discomfort. Statistical analysis. Data were presented as mean ± SDs for at least three repeated individual experiments for each group, and analyzed by one-way ANOVA followed by LSD and Dunnett's T3 using SPSS version 17.0 software. A value of p < 0.05 was considered statistically significant.