High Expression of SOX2 and OCT4 Indicates Radiation Resistance and an Independent Negative Prognosis in Cervical Squamous Cell Carcinoma

Abstract

Radiotherapy (RT) as a preoperative or postoperative adjuvant or primary treatment is the most common management modality for locally advanced cervical cancer. Radioresistance of tumor cells remains a major therapeutic problem. Consequently, we aimed to explore if the stem cell biomarkers SOX2 and OCT4 protein could be used to predict radioresistance in patients with locally advanced cervical squamous cell carcinoma (LACSCC). These 132 patients were divided into two groups (radiation-resistant and radiation-sensitive groups) according to progress-free survival (PFS). Using pretreatment paraffin-embedded tissues, we evaluated SOX2 and OCT4 expression using immunohistochemical staining. The percentage of overexpression of SOX2 and OCT4 in the radiation-resistant group was much higher than that in the radiation-sensitive group (p<0.001 and p <0.001, respectively). The patients with high expression of SOX2 and OCT4 showed a shorter PFS than those with low expression. Our study suggests that the expression of SOX2 and OCT4 in tumor cells indicates resistance to radiotherapy and that these two factors were important predictors of poor survival in patients with LACSCC (hazard ratio [95% CI], 2.294 [1.013, 5.195] and 2.300 [1.050, 5.037], respectively; p=0.046 and p=0.037, respectively).

Keywords

cervical carcinoma radiation resistance SOX2 OCT4 cancer stem cell immunohistochemistry

Introduction

Cervical cancer is the third-most commonly diagnosed cancer and the fourth leading cause of cancer death in females worldwide (Jemal et al. 2011). Radiotherapy (RT) as preoperative or postoperative adjuvant or primary treatment can be utilized to treat all stages of cervical cancer. Although RT plays an important role in the treatment of locally advanced or inoperable cervical carcinoma, treatment outcomes remain poor when external beam RT (EBRT) and brachytherapy (BRT) are used. Tumor resistance to radiation is a major therapeutic problem (Kim et al. 2006). Since the response to RT and prognosis may differ between patients who have the same tumor diameter and FIGO stage, new markers need to be identified in order to more accurately predict the response of individual patients to RT and their prognosis.

Malignant tumors vary in their response to radiation as a consequence of resistance mechanisms at the molecular level. Radiosensitivity can be influenced by both intrinsic and extrinsic factors such as hypoxia (Rockwell et al. 2009; Dewhirst et al. 2008), the cell cycle (Shimura et al. 2010; He et al. 2011), DNA damage and repair (Bolderson et al. 2009; Beskow et al. 2009), apoptosis (Lehmann et al. 2007; Suzuki et al. 2007), growth factors and oncogenes (Mukherjee et al. 2009; Bergkvist et al. 2011), cancer stem cells (CSCs) and the epigenetic modification of genes (Baumann et al. 2008; Kawano et al. 2010). CSCs, which have been recognized as a source of local or distant cancer relapse, have the potential to dramatically affect clinical outcomes (Brunner et al. 2012). A CSC is defined as a cell within a tumor that possesses the capacity to self-renew and generate the heterogeneous lineages of cancer cells that comprise the tumor (Clarke et al. 2006). This definition directly implies that an anticancer therapy can eradicate a tumor if and only if all CSCs are killed. The radioresistance resulting from the presence of CSCs has been described in some cell lines and solid tumor types (Peitzsch et al. 2013). Essentially, these studies consist of combinations between available CSC biomarkers and radiobiological endpoints. Comprehensive lists of biomarkers used for the identification or isolation of putative CSCs in different tumor entities have been given in several recent reviews (Brunner et al. 2012).

Recently, the transcription factors SOX2 and OCT4 have been proposed as biomarkers for breast CSC-like cells in both cell lines and malignant tissues (Leis et al. 2012; Ezeh et al. 2005). Sex-determining region Y-box 2 (SOX2) is a high-mobility group (HMG) domain transcription factor and a member of the SRY-related HMG-box (SOX) family of transcription factors (Saigusa et al. 2009), and is known to bind in partnership with OCT4. OCT4 is an octamer-binding transcription factor that is a key regulator of pluripotency, differentiation and self-renewal in embryonic stem cells (ESCs) (Lengerke et al. 2011). SOX2 and OCT4, which are expressed in various tumor tissues but not in normal somatic tissues, indicate poor prognosis in some solid tumors, including prostate cancer, lung adenocarcinoma, gliomas, rectal cancer, gastric carcinoma and oral squamous cell carcinoma (Leis et al. 2012; Ezeh et al. 2005; Guo et al. 2011; Matsuoka et al. 2012; Luo et al. 2013; Saigusa et al. 2009; Wen et al. 2010; Atlasi et al. 2007; Chiou et al. 2008; Sholl et al. 2010). Knockdown of these genes can result in tumor cell apoptosis, decreased tumor sphere formation and inhibition of tumor formation in xenograft tumor models (Luo et al. 2013). Thus, accumulating evidence suggests that SOX2 and OCT4 may be strongly correlated with CSCs.

To clarify the association between the expression of stem cell markers and radioresistance in primary locally advanced cervical squamous cell carcinoma (LACSCC), we measured SOX2 and OCT4 expression before RT by means of immunohistochemistry. In primary LACSCC treated with primary RT, we evaluated the relationship between the expression of these two markers and the clinicopathological characteristics, RT sensitivity and survival of the samples.

Materials & Methods

Patients and Clinical Tissue Samples

From January 2005 to March 2012, 614 patients suffering from localized cervical squamous carcinoma (LSCC) were documented in the Radiation Oncology Department of Xiangya Hospital and Affiliated Tumor Hospital of Xiangya Medical School, Central South University, Changsha, P.R. China. Patients received no other anticancer treatment prior to primary radiotherapy (RT), or surgery after RT. Ninety-eight patients were diagnosed at other hospitals, where their tissue samples were saved, and these patients went on to receive RT in our hospitals. As these hospitals had their own management rules of tissue samples, we were unable to obtain the samples and the patients were thus excluded from our study. Among the remaining 516 patients whose follow-up period closed in May 2012, 205 patients had follow-up information but others were unreachable for various reasons (no phone, no phone number, no follow-up, among others). Seventy-three cases had clear PFS; however, they were excluded from our study because they showed no progression, the PFS was shorter than 36 months and therefore we could not assign these patients into one of the study groups. Finally, 132 cases were included in our study. The inclusion criteria were as follows: (a) pathologically proven squamous cell carcinoma (SCC) of the cervix; (b) no evidence of distant metastasis at diagnosis (FIGO stage Ib–IVa); (c) the existence of tissue blocks available for our research; (d) receipt of no other anticancer treatment prior to primary RT, or surgery after RT. The study was approved by the Research and Ethics Committee of our institution. The median follow-up time for survivors was 45 (range, 2–85.5) months. The median PFS was 43.5 (range, 0–85.5) months. The median patient age was 51 (range, 28–80) years. We divided the 132 patients into two groups: radiation-sensitive group (n=85) and radiation-resistant group (n=47) (Kim et al. 2006). The radiation-sensitive group included patients who showed no local recurrence and distant metastasis for ≥3 years after primary treatment (PFS ≥36 months). The radiation-resistant group included patients who had tumors that did not respond to RT at all, with local recurrence or distant metastasis at <3 years after the primary treatment (PFS <36 months). The radiation-resistant group was further divided into three subgroups: (1) a RT non-responsive subgroup comprising patients whose primary tumor persisted and did not shrink markedly after primary treatment until time of death; (2) a local recurrence subgroup comprising patients whose primary tumor had initially disappeared but subsequently showed local recurrence at <3 years after the primary treatment; and (3) a distant metastasis subgroup comprising patients whose primary tumor had disappeared after treatment but showed distant metastasis at <3 years after primary treatment. PFS was defined as the period from the end of therapy to the date of the first documented evidence of recurrence or metastatic disease. Evidence included clinical physical examination data, pathological biopsy or imaging studies. Each primary cervical tumor diameter was directly measured during clinical physical examination rather than by medical imaging. In our study, most patients with a shorter follow-up were classified into the resistant group. For example, one patient with FIGO stage IIIa LASCC had no cervical tumor after radiotherapy based upon the physical examination and pelvic CT scan. Two months later, however, this patient showed evidence of lung metastasis, as determined by a lung CT scan, which was used to diagnose the cause of a recurring cough. This patient chose to withdraw from treatment and we lost contact with her. Therefore, both follow-up and PFS for this patient were two months. Another patient experienced lung metastasis during RT and the cervical tumor did not respond to RT. This patient belonged not only to the RT non-responsive subgroup but also to the distant metastasis subgroup. Three patients experienced distant metastasis after RT, but their PFS was >36 months. Thus, we placed these patients in the radiation-sensitive group.

All patients were treated with EBRT and high-dose rate (HDR) intracavitary BRT after consultation with a radiation oncologist. HDR brachytherapy was initiated at 3–4 weeks after the commencement of EBRT. The median total dose at point A was 90 (range, 66–102) Gy. The median dose of EBRT at point A was 46 (range, 30–52) Gy. The median dose of HDR BRT at point A was 42 (range, 20–54) Gy.

Immunohistochemistry

For immunohistochemical detection of SOX2 and OCT4, 4-µm tissue sections were deparaffinized in xylene and subjected to microwave treatment (SOX2 for 20 min, OCT4 for 10 min at moderate intensity) in 0.01 M citrate buffer (pH 6.0). After cooling for 30 min and washing in PBS, endogenous peroxidase was blocked with 3% hydrogen peroxide for 30 min, followed by incubation with PBS containing 10% normal goat serum for 30 min. Specimens were incubated overnight at 4C with the anti-SOX2 (Santa Cruz Biotechnology Inc; Dallas, TX) (Ding et al. 2012) and anti-OCT4 (Cell Signaling Technology Inc.; Beverly, MA) (Pan and Thomson 2007) antibodies at a dilution of 1:100 and 1:200, respectively. Detection involving immunostaining was performed using the ChemMate Kit (Dako; Glostrup, Denmark) and 3,3-diaminobenzidine as the chromogen. In negative control staining, the primary antibody was replaced with nonimmune isotypic antibodies.

Evaluation of Staining

The stained tissue slides were viewed separately by two pathologists who were blinded to the clinical or clinicopathological status of the patients. The expression of SOX2 and OCT4 on the slides was evaluated by scanning the entire tissue specimen under low-power magnification (40×), and then confirmed under high-power magnification (400×). An immunoreactivity scoring system was used. The proportion of tumor cell staining was evaluated in terms of four grades: 0, no positive tumor cells; 1, <10% positive tumor cells; 2, 10−50% positive tumor cells; and 3, >50% positive tumor cells. The scoring criteria used for staining intensity were: 0, no staining; 1, weak staining; 2, modest staining; and 3, strong staining. The final score was calculated by multiplying the area of tumor staining by the intensity score (0, 1, 2, 3, 4, 6 and 9). According to this method of assessment, staining scores ≤4 and ≥6 were regarded as tumors with low and high expression, respectively (Chiou et al. 2008). We stained 132 blocks/tumors and, for each block, chose five random areas to obtain weighted scores according to the percentage of positive tumor cells and staining intensity at 400× magnification. Thus, we had five weighted scores for each block and the scores may be different for one patient. We selected the median weighted score from the five weighted scores as the typical score and these data were analyzed. Two pathologists scored all samples blindly, without knowledge of the clinical characteristics or prognosis.

Statistical Analysis

The relationships between SOX2 and OCT4 expression and the clinicopathological factors were analyzed using the Chi-square test or Fisher’s exact test. The radiation dose between the radiation-resistant and radiation-sensitive groups was estimated by the Student’s t-test. The relationships between SOX2 and OCT4 were analyzed using the Spearman-test. Survival curves were plotted using the Kaplan-Meier method, and differences between groups were assessed using the log-rank test. Univariate and multivariate survival analyses were performed according to the Cox proportional hazards model. SOX2 expression alone (high vs. low), OCT4 expression alone (high vs. low), both SOX2 and OCT4 high expression (yes vs. no), age (≥50 years vs. <50 years), FIGO stage (III + IVa vs. Ib +II), histopathological grade (middle + low vs. high), tumor diameter (>4 cm vs. ≤4 cm) and combined chemotherapy (platinum-based, no vs. yes) were included in the regression model. For all statistical tests, a p value ≤0.05 was considered statistically significant.

Results

Patient Clinical and Histopathological Characteristics

The clinical and histopathological characteristics of the patients enrolled in the study are detailed in Table 1. There were significant differences in tumor diameter, FIGO stage and histological grading between the radiation-resistant group and the radiation-sensitive group (p<0.001, p=0.004 and p<0.001, respectively). No significant differences were observed with respect to patient age, combined chemoradiotherapy (platinum-based), total dose at point A, EBRT dose at point A and BRT dose at point A between the two groups.

Table 1.

Patient Characteristics.

		Radiation Sensitivity
		Radiation-resistant Group	Radiation-sensitive Group
Parameters	Patients (n=132)	n=47	n=85	p value
Age				0.559^a
<50 years	49	19 (38.8%)	30
≥50 years	83	28 (33.7%)	55
FIGO stage				0.004^a
I + II	70	17 (24.3%)	53
III + IVa	62	30 (48.4%)	32
Histopathological grade				0.014^b
High	10	0 (0%)	10
Middle + Low	114+8	47 (38.5%)	75
Tumor diameter				<0.001^a
≤4 cm	79	16 (20.3%)	63
>4 cm	53	31 (58.5%)	22
Combined chemotherapy (platinum-based)				0.426^a
Yes	106	36 (34.0%)	70
No	26	11 (42.3%)	15
Histological type (SCC^d)	132	47 (35.6%)	58
Total dose at point A
median dose (range) (Gy)		92 (67–102)	89 (66–102)	0.585^c
EBRT dose of point A
median dose (range) (Gy)		48 (30–52)	46 (36–50)	0.518^c
Brachytherapy dose at point A
median dose (range) (Gy)		46 (21–54)	42 (20–54)	0.387^c

p value was estimated by Chi-square test.

p value was estimated by Fisher’s exact test.

p value was estimated by t-test.

SCC^d: Squamous Cell Carcinoma.

SOX2 and OCT4 Expression and their Relationship to Clinicopathological Parameters

SOX2 was located in the nuclei of cervical carcinoma cells and staining was much stronger in the radiation-resistant group than in the radiation-sensitive group (Fig. 1A and 1C). Table 2 depicts the proportion of low and high expression of SOX2. No significant relationship was observed between SOX2 expression and patient age (p=0.944), FIGO stage (p=0.714), histopathological grade (p=0.845), or tumor diameter (p=0.177).

Figure 1.

Representative examples of SOX2 and OCT4 staining of tumors in the radiation-resistant group and radiation-sensitive group (400×). (A) Strong positive staining of SOX2 in the radiation-resistant group. (B) Strong positive staining of OCT4 in the radiation-resistant group. (C) Weak positive staining of SOX2 in the radiation-sensitive group. (D) Weak positive staining of OCT4 in the radiation-sensitive group. The bar size is the same for all the figures. Scale bar, 90 µm.

Table 2.

Correlation between SOX2 and OCT4 Expression and Clinicopathological Parameters for Cervical Cell Squamous Carcinoma.

		SOX2 expression			OCT4 expression			SOX2 AND OCT4 dual high expression
		Low expression	High expression		Low expression	High expression		Yes	No
Parameters	Patients (n=132)	n=49	n=83 (%)	p value	n=56	n=76 (%)	p value	n=62 (%)	n=70	p value
Age				0.944^a			0.659^a			0.996^a
<50 years	49	18	31 (63.3)		22	27 (55.1)		23 (46.9)	26
≥50 years	83	31	52 (62.7)		34	49 (59.0)		39 (47.0)	44
FIGO stage				0.714^a			0.416^a			0.314^a
Ib + II	70	27	43 (61.4)		32	38 (54.3)		30 (42.9)	40
III + IVa	62	22	40 (64.5)		24	38 (61.3)		32 (51.6)	30
Histopathological grade				0.845^b			0.018^b			0.102^b
High	10	4	6 (60.0)		8	2 (20.0)		2 (20.0)	8
Middle + Low	122	45	77 (63.1)		48	74 (60.7)		60 (49.2)	62
Tumor diameter				0.177^a			0.107^a			0.158^a
≤4 cm	79	33	46 (58.2)		38	41 (51.9)		33 (41.8)	46
>4 cm	53	16	37 (69.8)		18	35 (66.0)		29 (54.7)	24

p value was estimated by Chi-square test.

p value was estimated by Fisher’s exact test.

OCT4 was also located in the nuclei of the cervical carcinoma cells, with staining much stronger in the radiation-resistant group than in the radiation-sensitive group (Fig. 1B and 1D). Table 2 depicts the proportion of patients with low and high expression of OCT4. A statistically significant relationship was observed between OCT4 expression and histopathological grade (p=0.018). However, no significant relationship was observed between OCT4 expression and patient age (p=0.659), FIGO stage (p=0.416) and tumor diameter (p=0.107).

In LACSCC patients, the proportion with low expression of SOX2 and/or OCT4 and high expression of both SOX2 and OCT4 is depicted in Table 2. No significant relationship was observed between SOX2 and OCT4 dual high expression and patient age (p=0.996), FIGO stage (p=0.314), histopathological grade (p=0.102) and tumor diameter (p=0.158).

Relationship between Expression of SOX2 and OCT4

In total, 132 qualified patients could be evaluated for SOX2 and OCT4 expression using immunostaining. Each marker was classified as low or high according to the degree of immunohistochemical staining (Table 3). A significant positive correlation was observed between SOX2 and OCT4 expression (r=0.483; p<0.001).

Table 3.

Expression of SOX2 and OCT4 in Cervical Squamous Cell Carcinoma Patients.

	OCT4
	Low expression, n=49	High expression, n=83 (%)
SOX2
Low expression n = 56	36	20 (35.7)
High expression n = 76	13	63 (82.9)
	r = 0.483
	p < 0.001^a

p value was estimated by Spearman-test.

SOX2 and OCT4 Expression and Response to Radiotherapy

Data from SOX2 expression based upon the immunohistochemistry of 132 LACSCC patients are summarized in Table 4. Comparing the proportion of patients with high expression of SOX2 revealed statistically significant differences between the radiation-resistant and radiation-sensitive groups (p<0.001). The radiation-resistant group was further subdivided into three subgroups in accordance with the patient clinical information. A significant relationship was observed between SOX2 expression and RT non-response, local recurrence, and distant metastasis.

Table 4.

Relationship between SOX2 and OCT4 Expression and Response to Radiotherapy.

		SOX2 expression			OCT4 expression			SOX2 AND OCT4 dual high expression
Parameters	Patients (n=132)	low expression	high expression (%)	p value	low expression	high expression (%)	p value	No	Yes (%)	p value
Radiation sensitivity				<0.001^a			<0.001^a			<0.001^a
radiation-resistant group	47	7	40 (85.1)		7	40 (85.1)		13	34 (72.3)
radiation-sensitive group	85	42	43 (50.6)		49	36 (42.4)		57	28 (32.9)
RT non-response				0.036^b			0.075^b			0.012^b
RT non-responsive subgroup	9	1	8 (88.9)		2	7 (77.8)		2	7 (77.8)
Radiation-sensitive group	85	42	43 (50.6)		49	36 (42.4)		57	28 (32.9)
Local recurrence				0.004^a			0.003^a			0.001^a
Local recurrence subgroup	17	2	15 (88.2)		3	14 (82.4)		4	13 (76.5)
Radiation-sensitive group	85	42	43 (50.6)		49	36 (42.4)		57	28 (32.9)
Distant metastasis				0.008^a			<0.001^a			0.003^a
Distant metastasis subgroup	22	4	18 (81.8)		2	20 (90.9)		7	15 (68.2)
Radiation-sensitive group	85	42	43 (50.6)		49	36 (42.4)		57	28 (32.9)

p value was estimated by Chi-square test.

p value was estimated by Fisher’s exact test.

Data from the immunohistochemical evaluation of OCT4 expression and SOX2 and OCT4 co-expression in LACSCC patients are also summarized in Table 4. There were also statistically significant differences among the groups.

SOX2 and OCT4 Expression and Patient Survival

As the patient cohort was stratified according to tumor expression of SOX2, the 5-year PFS rates in patients with low SOX2 expression (n=49) and high SOX2 expression (n=83) were 84.58% and 55.48 %, respectively; Kaplan-Meier analysis (log-rank test) revealed a significant difference between the two groups (p<0.001; Fig. 2A). With respect to OCT4 expression, the 5-year PFS rates in patients with low OCT4 expression (n=56) and high OCT4 expression (n=76) were 93.10% and 56.39%, respectively; Kaplan-Meier analysis (log-rank test) revealed a significant difference between the two groups (p<0.001; Fig. 2B).When the patient cohort was stratified according to the expression of both SOX2 and OCT4, the 5-year PFS rates were 80.01% and 50.91% in patients that showed low expression of SOX2 and/or OCT4 (n=70) and high expression of both SOX2 and OCT4 (n=62), respectively. It was noteworthy that patients who had tumors with high expression of both SOX2 and OCT4 had a worse prognosis than that of patients with low expression of SOX2 and/or OCT4 (Kaplan-Meier analysis; log-rank test; p<0.001; Fig. 2C). The 5-year PFS rate for all 132 patients was 66.79% (Fig. 2D).

Figure 2.

Kaplan-Meier survival curves for LACSCC patients. (A) The 5-year progression-free survival (PFS) rates were 84.58% and 55.48% in patients with low SOX2 expression (n=49) and high SOX2 expression (n=83), respectively. There was a significant difference in the overall survival rate between the two groups (p<0.001). (B) The 5-year PFS rates were 93.10% and 56.39% in patients with low OCT4 expression (n=56) and high OCT4 expression (n=76), respectively, with a significant difference in the overall survival rate between the two groups (p<0.001). (C) The 5-year PFS rates were 80.01% and 50.91% in patients that showed SOX2 and/or OCT4 low expression (n=70) and SOX2 and OCT4 dual high expression (n=62), respectively, with a significant difference in the overall survival rate between the two groups (p<0.001). (D) The 5-year PFS rate was 66.79% for all of the 132 patients with LACSCC.

The univariate Cox regression analysis data for hazard ratios are depicted in Table 5. The multivariate Cox regression analysis data indicated that SOX2 and OCT4 co-expression was not an informative independent prognostic factor regarding our cohort of patients with LACSCC (Table 5). However, SOX2 alone and OCT4 alone had significant independent predictive value concerning PFS (Table 5). In addition, FIGO stage and tumor diameter were also independent prognostic predictors (Table 5).

Table 5.

Univariate and Multivariate COX Regression Analysis of the Relationships between Clinicopathological Outcomes in 132 Patients with Cervical Squamous Cell Carcinoma.

Variable	Subset	Hazard radio (95% CI)	p value
Univariate analysis (n=132)
SOX2 expression alone	High vs. Low	3.471 (1.683, 7.157)	<0.001
OCT4 expression alone	High vs. Low	4.047 (2.013, 8.138)	<0.001
SOX2 AND OCT4 expression	SOX2 AND OCT4 dual high vs. SOX2 and/or OCT4 low	3.263 (1.791, 5.948)	<0.001
Age	≥50 years vs. <50 years		0.872
FIGO stage	III + IVa vs. Ib + II	2.610 (1.453, 4.689)	0.001
Histopathological grade	Low + Middle vs. High		0.067
Tumor diameter	>4 cm vs. ≤4 cm	3.366 (1.885, 6.012)	<0.001
Combined chemotherapy (platinum-based)	Yes vs. No		0.446
Multivariate analyses (n=132)
SOX2 expression alone	High vs. Low	2.294 (1.013, 5.195)	0.046
OCT4 expression alone	High vs. Low	2.300 (1.050, 5.037)	0.037
SOX2 AND OCT4 expression	SOX2 AND OCT4 dual high vs. SOX2 and/or OCT4 low		0.458
Age	≥50 years vs. <50 years		0.291
FIGO stage	III + IVa vs. Ib + II	2.461 (1.350, 4.486)	0.003
Histopathological grade	Low + Middle vs. High		0.051
Tumor diameter	>4 cm vs. ≥4 cm	2.887 (1.598, 5.218)	<0.001
Combined chemotherapy (platinum-based)	Yes vs. No		0.382

Discussion

By definition, one CSC has the potential to generate an entire tumor (Baumann et al. 2009), suggesting that all CSCs need to be inactivated for permanent local and distant tumor control after RT. RT that inactivates a high proportion of the non-CSC mass of tumor cells can lead to tumor regression and tumor growth delay without improving tumor control. However, RT that inactivates a high proportion of CSCs can improve tumor control. Increasing evidence suggests that CSCs have a higher radioresistance than non-stem cells from the same tumor (Saigusa et al. 2009). A number of mechanisms are involved in CSC-mediated tumor radioresistance (Krause et al. 2011), such as their intrinsic cellular radioresistance, especially regarding their ability to repair double-strand breaks (DSB) in DNA. Tumor radioresistance is also governed by CSC resistance to reactive oxygen species (ROS) (Diehn et al. 2009), the absolute number of CSCs before treatment (Yaromina et al. 2007), the repopulation capacity of CSCs between radiation fractions (Hessel et al. 2004) and microenvironmental factors such as hypoxia (Yaromina et al. 2010) and tumor lactate level (Sattler et al. 2010). These factors may interact and further increase tumor radioresistance mediated by CSCs.

SOX2 and OCT4 are essential transcription factors for the maintenance of pluripotent stem cells. The SOX2-OCT4 complex seems to be highly placed in the regulatory hierarchy of the pluripotent genetic network (Rodda et al. 2005). Furthermore, SOX2 and OCT4 are recognized CSC biomarkers in breast cancer. Thus, SOX2 and OCT4 play important roles in maintaining CSC biological behavior. For example, SOX2 silencing in glioblastoma tumor-initiating cells can stop proliferation and loss of tumorigenicity (Gangemi et al. 2009). Also, oncolytic adenovirus carrying OCT4 response elements in murine metastatic bladder cancer models can induce cytolysis (Wu et al. 2008).

Recent studies concerning cervical carcinomas indicate that CSCs confer a higher degree of tumor resistance to RT as compared with non-CSCs, and that SOX2 and OCT4 may be the CSC biomarkers. Lopez and coworkers (2012) showed that cancer-initiating cells derived from established cervical cell lines exhibited stem-cell markers and increased radioresistance. In another study, Feng and coworkers (2009) reported that primary cervical cancer contains a subpopulation of tumor-initiating cells with stem-like properties, and that SOX2 and OCT4 are the biomarkers for cervical carcinoma CSCs. The results from our study indicate that SOX2 and OCT4 were expressed significantly more strongly in the radioresistant group than in the radiosensitive group. Even when comparing the expression levels of SOX2 and OCT4 in the RT non-responsive subgroup, the local recurrence subgroup or the distant metastasis subgroup with those in the radiation-sensitive group, the statistical differences were also significant. Thus, we conclude that SOX2 and OCT4 play important roles with regard to LACSCC tumor radioresistance. In addition, high expression levels indicated a shorter PFS and an independent negative prognosis. Similar findings were reported in another study of rectal cancer (Saigusa et al. 2009), where SOX2 and OCT4 expression may be associated with tumor radioresistance and may predict distant recurrence and poor prognosis in rectal cancer patients treated with preoperative chemoradiotherapy.

Identification of the correlation between SOX2 and OCT4 may be an effective approach for elucidating the roles played by SOX2 and OCT4 in tumor radioresistance. In our study, a positive association between SOX2 and OCT4 was demonstrated. Although limited research has focused on the correlation between SOX2 and OCT4 in CSCs, some studies suggest that there is reciprocal transcriptional regulation of SOX2 and OCT4 via the SOX2/OCT4 complex in ESCs and that the function of SOX2 and OCT4 is dependent on this complex (Chew et al. 2005; Okumura-Nakanishi et al. 2005; Tomioka et al. 2002). Therefore, we assume that co-expression of SOX2 and OCT4 plays a more significant role than SOX2 alone or OCT4 alone. Although the co-expression of SOX2 and OCT4 was associated with a shorter PFS in our study, we found that this co-expression was not an informative independent prognostic factor regarding clinical outcomes, in contrast to the expression of SOX2 or OCT4 alone. It is possible that differences exist between the signaling pathways involved in CSCs and ESCs, or the numbers of the patients enrolled in our study were insufficient. Thus, further research is required here.

In summary, CSCs as a cause of local or distant relapse after RT are a potential therapeutic target. Treatment may be dependent on the presence of specific biomarkers because of the individualization treatment principle. SOX2 and OCT4 are suitable candidate biomarkers for patients with LACSCC, found in this study to be markers for tumor radioresistance and also independent negative prognostic factors. These findings require support from further research. First, retrospective clinical studies to evaluate SOX2 and OCT4 expression in pre-therapeutic biopsies across different centers will require sufficient samples for different cancers. Data from these studies will be important for predicting clinical outcomes and appropriate treatments. Second, in vitro studies are necessary to confirm the effect of radiotherapy in cell lines that highly express SOX2 and OCT4. SOX2 and OCT4 knockout models will also be useful here. Third, to address gaps in clinical observations and cell-based experimental data, in vivo radiobiological assays will be helpful in determining areas of tumor control after different radiation doses. These will be studied along with the frequency of tumorigenic CSCs as defined by specific phenotypic markers (SOX2 and OCT4) and functional features, including tumorigenicity, self-renewal and differentiation potential. If the biomarker profile is unfavorable, the therapeutic strategy for LACSCC may change. If favorable, we can detect inhibitors or antibodies targeted against these biomarkers, modulate the dose and frequency of fractionated radiotherapy, or even change the radioactive type (photon, proton, or carbon ion ray), according to the specific radiobiological characteristics of CSCs with high expression of these biomarkers. Certainly, we will consider the use of these markers as stratifiers in clinical trials.

Footnotes

Acknowledgements

We thank Professor Yang Lifang for advice and suggestions and Professor Yang Tubao for statistical guidance.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by National Natural Science Foundation of China [grant numbers 81372792, 81225013, 81101193]; Hunan Department of Science and Technology Foundation [grant numbers 2013SK2019, 2011FJ4184]; and the Freedom Explore Fund for the Doctoral Program of Central South University [grant number 2013zzts089].

References

Atlasi

Mowla

Ziaee

Bahrami

(2007). OCT-4, an embryonic stem cell marker, is highly expressed in bladder cancer. Int J Cancer 120:1598-1602.

Baumann

Krause

Hill

(2008). Exploring the role of cancer stem cells in radioresistance. Nat Rev Cancer 8:545-554.

Baumann

Krause

Thames

Trott

Zips

(2009). Cancer stem cells and radiotherapy. Int J Radiat Biol 85:391-402.

Bergkvist

Argyle

Pang

Muirhead

Yool

(2011). Studies on the inhibition of feline EGFR in squamous cell carcinoma: enhancement of radiosensitivity and rescue of resistance to small molecule inhibitors. Cancer Biol Ther 11:927-937.

Beskow

Skikuniene

Holgersson

Nilsson

Lewensohn

Kanter

Viktorsson

(2009). Radioresistant cervical cancer shows upregulation of the NHEJ proteins DNA-PKcs, Ku70 and Ku86. Br J Cancer 101:816-821.

Bolderson

Richard

Zhou

Khanna

(2009). Recent advances in cancer therapy targeting proteins involved in DNA double-strand break repair. Clin Cancer Res 15:6314-6320.

Brunner

Kunz-Schughart

Grosse-Gehling

Baumann

(2012). Cancer stem cells as a predictive factor in radiotherapy. Semin Radiat Oncol 22:151-174.

Chew

Loh

Zhang

Chen

Tam

Yeap

Ang

Lim

Robson

(2005). Reciprocal transcriptional regulation of Pou5f1 and Sox2 via the Oct4/Sox2 complex in embryonic stem cells. Mol Cell Biol 25:6031-6046.

Chiou

Huang

Lin

Liu

Tsai

Chou

Chien

(2008). Positive correlations of Oct-4 and Nanog in oral cancer stem-like cells and high-grade oral squamous cell carcinoma. Clin Cancer Res 14:4085-4095.

10.

Clarke

Dick

Dirks

Eaves

Jamieson

Jones

Visvader

Weissman

Wahl

(2006). Cancer stem cells-perspectives on current status and future directions: AACR Workshop on cancer stem cells. Cancer Res 66:9339-9344.

11.

Dewhirst

Cao

Moeller

(2008). Cycling hypoxia and free radicals regulate angiogenesis and radiotherapy response. Nat Rev Cancer 8:425-437.

12.

Diehn

Cho

Lobo

Kalisky

Dorie

Kulp

Qian

Lam

Ailles

Wong

Joshua

Kaplan

Wapnir

Dirbas

Somlo

Garberoglio

Paz

Shen

Lau

Quake

Brown

Weissman

Clarke

(2009). Association of reactive oxygen species levels and radioresistance in cancer stem cells. Nature 458:780-783.

13.

Ding

Liang

Zhao

Wang

(2012). Over-expression of Sox2 in C3H10T1/2 cells inhibits osteoblast differentiation through Wnt and MAPK signalling pathways. Int Orthop 36: 1087-94

14.

Ezeh

Turek

Reijo

Clark

(2005). Human embryonic stem cell genes OCT4, NANOG, STELLAR, and GDF3 are expressed in both seminoma and breast carcinoma. Cancer 104:2255-2265.

15.

Feng

Peng

Zhou

Ling

Wei

Tian

(2009). Identification and characterization of cancer stem-like cells from primary carcinoma of the cervix uteri. Oncol Rep 22:1129-1134.

16.

Gangemi

Griffero

Marubbi

Perera

Capra

Malatesta

Ravetti

Zona

Daga

Corte

(2009). SOX2 silencing in glioblastoma tumor-initiating cells causes stop of proliferation and loss of tumorigenicity. Stem Cells 27:40-48.

17.

Guo

Liu

Wang

Zhao

Wang

Bing

Zhang

Ling

Gao

Hao

(2011). Expression profile of embryonic stem cell-associated genes Oct4, Sox2 and Nanog in human gliomas. Histopathology 59:763-775.

18.

Zhou

Zhu

Wang

Mizota

Furusawa

Zhou

(2011). Cell cycle suspension: a novel process lurking in G(2) arrest. Cell Cycle 10:1468-1476.

19.

Hessel

Krause

Petersen

Horcsoki

Klinger

Zips

Thames

Baumann

(2004). Repopulation of moderately well-differentiated and keratinizing GL human squamous cell carcinomas growing in nude mice. Int J Radiat Oncol Biol Phys 58:510-518.

20.

Jemal

Bray

Center

Ferlay

Ward

Forman

(2011). Global cancer statistics. CA Cancer J Clin 61:69-90.

21.

Kawano

Akiyama

Agawa-Ohta

Mikami-Terao

Iwase

Yanagisawa

Ida

Agata

Yamada

(2010). Histone deacetylase inhibitors valproic acid and depsipeptide sensitize retinoblastoma cells to radiotherapy by increasing H2AX phosphorylation and p53 acetylation-phosphorylation. Int J Oncol 37:787-795.

22.

Kim

Lee

Song

Choi

Kim

Lee

Bae

(2006). Increased expression of pAKT is associated with radiation resistance in cervical cancer. Br J Cancer 94:1678-1682.

23.

Krause

Yaromina

Eicheler

Koch

Baumann

(2011). Cancer stem cells: targets and potential biomarkers for radiotherapy. Clin Cancer Res 17:7224-7229.

24.

Lehmann

McCubrey

Jefferson

Paine

Chappell

Terrian

(2007). A dominant role for p53-dependent cellular senescence in radiosensitization of human prostate cancer cells. Cell Cycle 6:595-605.

25.

Leis

Eguiara

Lopez-Arribillaga

Alberdi

Hernandez-Garcia

Elorriaga

Pandiella

Rezola

Martin

(2012). Sox2 expression in breast tumours and activation in breast cancer stem cells. Oncogene 31:1354-1365.

26.

Lengerke

Fehm

Kurth

Neubauer

Scheble

Muller

Schneider

Petersen

Wallwiener

Kanz

Fend

Perner

Bareiss

Staebler

(2011). Expression of the embryonic stem cell marker SOX2 in early-stage breast carcinoma. BMC Cancer 11:42.

27.

Lopez

Poitevin

Mendoza-Martinez

Perez-Plasencia

Garcia-Carranca

(2012). Cancer-initiating cells derived from established cervical cell lines exhibit stem-cell markers and increased radioresistance. BMC Cancer 12:48.

28.

Luo

Peng

Deng

Yao

(2013). Embryonic stem cells markers SOX2, OCT4 and Nanog expression and their correlations with epithelial-mesenchymal transition in nasopharyngeal carcinoma. PLoS One 8:e56324.

29.

Matsuoka

Yashiro

Sakurai

Kubo

Tanaka

Muguruma

Sawada

Ohira

Hirakawa

(2012). Role of the stemness factors sox2, oct3/4, and nanog in gastric carcinoma. J Surg Res 174:130-135.

30.

Mukherjee

McEllin

Camacho

Tomimatsu

Sirasanagandala

Nannepaga

Hatanpaa

Mickey

Madden

Maher

Boothman

Furnari

Cavenee

Bachoo

Burma

(2009). EGFRvIII and DNA double-strand break repair: a molecular mechanism for radioresistance in glioblastoma. Cancer Res 69:4252-4259.

31.

Okumura-Nakanishi

Saito

Niwa

Ishikawa

(2005). Oct-3/4 and Sox2 regulate Oct-3/4 gene in embryonic stem cells. J Biol Chem 280:5307-5317.

32.

Pan

Thomson

(2007). Nanog and transcriptional networks in embryonic stem cell pluripotency. Cell Research 17: 42-9

33.

Peitzsch

Kurth

Kunz-Schughart

Baumann

Dubrovska

(2013). Discovery of the cancer stem cell related determinants of radioresistance. Radiother Oncol 108:378-387.

34.

Rockwell

Dobrucki

Kim

Marrison

(2009). Hypoxia and radiation therapy: past history, ongoing research, and future promise. Curr Mol Med 9:442-458.

35.

Rodda

Chew

Lim

Loh

Wang

Robson

(2005). Transcriptional regulation of nanog by OCT4 and SOX2. J Biol Chem 280:24731-24737.

36.

Saigusa

Tanaka

Toiyama

Yokoe

Okugawa

Ioue

Miki

Kusunoki

(2009). Correlation of CD133, OCT4, and SOX2 in rectal cancer and their association with distant recurrence after chemoradiotherapy. Ann Surg Oncol 16:3488-3498.

37.

Sattler

Meyer

Quennet

Hoerner

Knoerzer

Fabian

Yaromina

Zips

Walenta

Baumann

Mueller-Klieser

(2010). Glycolytic metabolism and tumour response to fractionated irradiation. Radiother Oncol 94:102-109.

38.

Shimura

Kakuda

Ochiai

Nakagawa

Kuwahara

Takai

Kobayashi

Komatsu

Fukumoto

(2010). Acquired radioresistance of human tumor cells by DNA-PK/AKT/GSK3beta-mediated cyclin D1 overexpression. Oncogene 29:4826-4837.

39.

Sholl

Barletta

Yeap

Chirieac

Hornick

(2010). Sox2 protein expression is an independent poor prognostic indicator in stage I lung adenocarcinoma. Am J Surg Pathol 34:1193-1198.

40.

Suzuki

Oka

Yoshida

Shirai

Ohno

Kato

Tsujii

Nakano

(2007). Correlation between survivin expression and locoregional control in cervical squamous cell carcinomas treated with radiation therapy. Gynecol Oncol 104:642-646.

41.

Tomioka

Nishimoto

Miyagi

Katayanagi

Fukui

Niwa

Muramatsu

Okuda

(2002). Identification of Sox-2 regulatory region which is under the control of Oct-3/4-Sox-2 complex. Nucleic Acids Res 30:3202-3213.

42.

Wen

Park

Chung

Bang

Park

Song

(2010). Oct4 and Nanog expression is associated with early stages of pancreatic carcinogenesis. Pancreas 39:622-626.

43.

Shieh

Chang

Huang

Shiau

(2008). Tumor-selective replication of an oncolytic adenovirus carrying oct-3/4 response elements in murine metastatic bladder cancer models. Clin Cancer Res 14:1228-1238.

44.

Yaromina

Krause

Thames

Rosner

Krause

Hessel

Grenman

Zips

Baumann

(2007). Pre-treatment number of clonogenic cells and their radiosensitivity are major determinants of local tumour control after fractionated irradiation. Radiother Oncol 83:304-310.

45.

Yaromina

Thames

Zhou

Hering

Eicheler

Dorfler

Leichtner

Zips

Baumann

(2010). Radiobiological hypoxia, histological parameters of tumour microenvironment and local tumour control after fractionated irradiation. Radiother Oncol 96:116-122.