Role of SDF-1α/CXCR4 signaling pathway in clinicopathological features and prognosis of patients with nasopharyngeal carcinoma

The present study aims to explore the role of stromal cell-derived factor-1α (SDF-1α)/stromal cell-derived factor receptor-4 (CXCR4) signaling pathway to the clinicopathological features and prognosis of patients with nasopharyngeal carcinoma (NPC). From January 2009 to December 2010, 102 patients with NPC and 80 patients with chronic nasopharyngitis were enrolled for the study. Immunohistochemical staining, quantitative real-time polymerase chain reaction (qRT-PCR), and Western blotting were employed to determine the expressions of SDF-1α and CXCR4 proteins in NPC tissues and chronic nasopharyngitis tissues. Chi-square test was conducted to analyze the associations of the expressions of SDF-1α and CXCR4 proteins with the clinicopathological features of NPC patients. Spearman rank correlation analysis was used to analyze the correlation between the SDF-1α protein expression and CXCR4 protein expression. The mRNA and protein expressions of SDF-1α and CXCR4 in NPC tissues were significantly higher than those in chronic nasopharyngitis tissues. The expressions of SDF-1α and CXCR4 proteins showed associations with T staging, N staging, tumor node metastasis (TNM) staging, skull base invasion, and cervical lymph node metastasis of NPC patients. Compared with NPC patients showing negative expressions of SDF-1α and CXCR4 proteins, those with positive expressions of SDF-1α and CXCR4 proteins had a significantly shorter survival time. SDF-1α protein, CXCR4 protein, EBV-IgG status, T staging, N staging, TNM staging, skull base invasion, and cervical lymph node metastasis were independent risk factors for the prognosis of NPC. The findings indicated that SDF-1α/CXCR4 signaling pathway might be associated with the clinicopathological features and prognosis of patients with NPC.


Introduction
Nasopharyngeal carcinoma (NPC), a malignant nose pharynx tumor, is commonly misdiagnosed and difficult to treat [1]. As a disease with racial and geographical disparities, NPC is very common in Southern China and Southeast Asia with an occurrence rate of 25-50 patients per 100,000 people each year, and the incidence rate is a hundred times higher than that in western countries [2]. The occurrence and development of NPC are mainly associated with the infection of Epstein-Barr virus (EBV), and other co-factors likewise play an important role in the pathogenesis of NPC, such as contaminated food containing dietary nitrosamines, long-term exposure to wood and wood dust, smoking, viral infections, and host genetic susceptibility [3]. Based on the classification provided by Union for International Cancer Control (UICC), the main stages of NPC include T staging (local carcinoma growth), N staging (spreading to regional lymph nodes), and M staging (developing to distant metastasis) [4]. In NPC, the nasopharynx lymphatic drainage mainly shifts to the cervical lymph nodes [5], and the skull base invasion is regarded as a crucial prognostic value in this disease [6]. Radiotherapy, as the main therapy for NPC, can effectively control early stage NPC, but for patients suffering from advanced local NPC, the 5-year survival rate after radiotherapy is approximately 50% because of the high recurrence and cancer cell metastasis [7]. Therefore, there is an urgent need to find a new effective target from the perspective of molecular biology in order to have a more comprehensive and deeper understanding of the pathogenesis of and treatment for NPC and to improve the survival rate of patients with advanced NPC.
Stromal cell-derived factor-1 (SDF-1), also known as CXCL12, is a homeostatic chemokine signaling via the receptor CXCR4, and shows relations with blood cell production, immune system development, tumor progression, and angiogenesis in various tumors [8]. Findings revealed that SDF-1α is expressed in diverse cancer cells and associated with the formation and metastasis of cancer cells [9]. Stromal cell-derived factor receptor-4 (CXCR4), being a prominent chemokine receptors, can solitarily be triggered by chemokine CXCL12 and furthermore promote metastasis, angiogenesis, tumor progression, or tumor survival [10]. Inactivation of CXCR4 will disturb the regulation of proteases in the primitive hematopoietic stem cells, and further inhibit the migration of hematopoietic stem cells and the direct ablation of CXCR4 signals [11]. The SDF-1/CXCR4 axis can not only promote actin filament polymerization of tumor cells to form pseudopod that is conducive to directional migration, but also control the formation of local new vessels and impact tumor cell growth, invasion, and metastasis to some extent [12,13]. At present, there are not many researches about the expression of SDF-1α/CXCR4 signaling pathways in NPC tissues. The present study aims to investigate the role of SDF-1α/CXCR4 signaling pathway in clinicopathological features and prognosis of patients with NPC.

Ethics statement
All experiments were carried out in strict accordance with the Declaration of Helsinki [14], and the ethical approval was issued by the Ethics Committee of Women and Children's Hospital of Linyi on December 19, 2008. Written informed consent was obtained from all patients or guardians.

Study subjects
The observation group population consisted of 102 NPC patients including 62 males and 40 females (mean age: 52.9 + − 16.2 years) undergoing treating in the Department of Oncology of Women and Children's Hospital of Linyi from January 2009 to December 2010. According to the tumor node metastasis (TNM) staging criteria jointly developed by the Union for International Cancer Control (UICC) and American Joint Committee on Cancer (AJCC) [15], there were 21 cases in T1, 24 cases in T2, 26 cases in T3, 31 cases in T4 categories, and furthermore 18 cases in N0, 16 cases in N1, 30 cases in N2, and 38 cases in N3 categories; as for the TNM staging, there were 14 cases at stage I, 29 cases at stage II, 32 cases at stage III, and 27 cases at stage IV. Inclusion criteria was as follows: (1) all patients were non-keratinizing NPC (type II/ III) confirmed by histocytology; (2) patients were infected with NPC for the first time and did not receive radiotherapy or immunobiological therapy before tissue clamping; (3) all patients had complete medical records and baseline information, including head and neck computer tomography (CT) or magnetic resonance imaging (MRI), chest X-ray, bone scan, and B-ultrasound examination records. Exclusion criteria was: (1) patients had implications with other organ tumors; (2) patients were found with inflammation during the first examination, but were only diagnosed with NPC during the second or third examination; (3) patients without complete clinicopathological data or unwilling to cooperate for treatment. Additionally, another 80 patients with chronic nasopharyngitis were selected as the control group, including 43 males and 37 females (mean age: 54.8 + − 6.5 years). The nasopharyngeal tissue samples of the observation group and the control group were obtained, fixed in 10% formalin solution, and embedded in paraffin wax to make sections for immunohistochemical staining.

Immunohistochemical staining
The immunohistochemical staining steps were as follows: the tissue sections were placed in an oven at 65 • C for duration of 2 h, dewaxed in dimethyl benzene for 15 min thrice, and washed using gradient alcohol solution. Tissue sections were washed with distilled water, and 3% hydrogen peroxide was used for the removal of the peroxide enzyme. Subsequently, all sections were immersed in citric acid under high temperature and pressure for antigen retrieval for 1 min, rinsed and cooled using distilled water, and buffered with phosphate-buffered saline (PBS) liquid for 3 min thrice. Following the addition of rabbit anti-human SDF-1 monoclonal antibody (1:100; Abcam Inc., Cambridge, MA, U.S.A.) and mouse anti-human CXCR4 monoclonal antibody (1:200; Abcam Inc., Cambridge, MA, U.S.A.), the sections were incubated for 1 h at 37 • C and washed with PBS buffer for 3 min thrice. Diaminobenzidine (DAB) (Beijing Jinqiao Technology Co., Ltd., Beijing, China) was used for coloration for 5-10 min before water washing. Hematoxylin, alcohol, dimethyl benzene, and neutral balsam were used for counterstaining, dehydration, transparency, and final section sealing respectively.
The results of IHC for SDF-1α and CXCR4 proteins were evaluated by two experienced pathologists with no access to the clinical and pathological data of patients using the methods developed by Kobayashi et al. [16] and Mukheriee D et al. [17] respectively. SDF-1α positive cell staining was localized to the cytoplasm and a slight expression in the cell membrane, specifying that cytoplasm staining was defined as positive expression [18]; CXCR4 was primarily expressed in the cytoplasm or nucleus, indicating that cytoplasm staining, nucleus staining, or co-staining was defined as positive expression [19].

Quantitative real-time polymerase chain reaction
The total RNA of the nasopharyngitis tissues frozen at −80 • C was extracted. According to the instructions of TaqMan . C t is regarded as the number of amplification cycles before the real-time fluorescence intensity in response reaches the set threshold. Once the set threshold is reached, the amplification presents a logarithmic growth. The experiment in each group was repeated thrice.

Western blotting
The protein was extracted from nasopharyngitis tissues frozen at −80 • C using protein extraction kit (Nanjing KeyGen Biotech. Inc., Jiangsu, China) and the protein concentration was measured using BCA kit (Wuhan Boster Biological Technology Ltd., Wuhan, Hubei, China). The extracted protein was added with loading buffer and boiled in water for 10 min. Each well was loaded with 40 μg samples, and 10% polyacrylamide gel electrophoresis (Wuhan Boster Biological Technology Ltd., Wuhan, Hubei, China) was conducted to separate the protein at a voltage from 80 to 120 V, followed by the wet transference at voltage of 100 mV for 45 to 70 min. With polyvinylidene fluoride (PVDF) transference, the samples were added with 5% skim milk and sealed at room temperature for 1 h. The tissues were added with the diluted primary antibodies SDF-1α (1μg/ml; ab18919, Abcam, Cambridge, MA, U.S.A.) and CXCR4 (1 μg/ml; ab13854, Abcam, Cambridge, MA, U.S.A.) and were incubated overnight and washed with phosphate buffer solution (PBS) at room temperature trice (5 min per time). The tissues were added with the second antibody (1:2000 dilution) and incubated at room temperature for 1 h. The membrane was washed trice (8 min per time). The PIERCE chemiluminescence assay was conducted for development (Shanghai Bioleaf Biotech Co., Ltd., Shanghai, China). GAPDH was regarded as an internal reference. Bio-rad Gel Dol EZ imager was adopted for development. The ImageJ software was employed for gray value analysis of the target band. The protein expression = (the grey value of target band)/(the grey value of the internal parameter band). The experiment in each group was repeated thrice.

Follow-up
The follow-up was conducted from January 1, 2010 to December 31, 2015, and the median follow-up time was 36 (11-60) months. During the follow-up, 42 patients (41.2%) died, of which five cases (4.9%) died of other causes like cardiovascular disease or accident. Till the end of follow-up, there were 60 survivors (58.8%). Patients died of other causes or lost to follow-up, as well as final survivors were counted as censored data.

Statistical analysis
The statistical analysis was performed using the SPSS 21.0 (SPSS Inc., Chicago, IL, U.S.A.). The Chi-square test was used to analyze the correlation of the expressions of proteins with the clinicopathological characteristics of NPC patients. Correlation of different protein expressions was analyzed using the Spearman rank correlation analysis. The Kaplan-Meier curve was adopted to calculate the survival time. The associations between the expression of SDF-1α and CXCR4 and the OS and DFS were analyzed by the multivariate Cox regression model and Wald test. P<0.05 is considered as statistically significant.

Positive expressions of SDF-1α and CXCR4 proteins in NPC tissues and chronic nasopharyngitis tissues
Following the semi-quantitative scoring method, five high power fields (×400) were randomly chosen for each case. The scoring of the percentage of positive cells was as follows: not more than 5% positive cells equals 0 point, 6-25% equals 1 point, 26-50% equals 2 points, 51-75% equals 3 points, and not less than 76% equals 4 points. The scoring of positive intensity was as follows: no coloring equates to 0 point, yellow equates to 1 point, brown yellow equates to 2 points, and dark brown equals to 3 points. Once the scoring of the percentage of positive cells was multiplied with the scoring of positive intensity, 0 point was negative (−), 1-2 points was weak positive (+), 3-5 points was moderate positive (++), 6-8 points was strong positive (+++). (−) was defined as negative expression, and (+), (++), and (+++) were all defined as positive expression [20].
The SDF-1α protein staining was primarily localized to the cytoplasm, showing brownish yellow granules, and partially seen in the cell membrane as seen in Figure 1. Among the 102 NPC tissues, 46 cases showed positive expression (positive rate: 45.1%), in which 11 cases were weak positive (+), 25 cases were positive (++), and 10 cases were strong positive (+++). In the 80 chronic nasopharyngitis tissues, 11 cases showed positive expression (positive rate: 13.8%), moreover, all were weak positive (+). The positive expression rate of SDF-1α protein in NPC tissues was significantly higher in comparison with the chronic nasopharyngitis tissues (P<0.05). The CXCR4 protein was mainly localized in the cytoplasm or nucleus of NPC tissues with brownish yellow coloration as seen in Figure 2. In the 102 NPC tissues, 75 cases showed positive expression (positive rate: 73.5%), in which 19 cases were weak positive (+), 32 cases were positive (++), and 24 cases were strong positive (+++). In the 80 chronic nasopharyngitis tissues, 29 cases had positive expression (positive rate: 36.3%), of which 17 cases were weak positive (+), 10 cases were positive (++), and 2 cases were strong positive (+++). The positive expression rate of CXCR4 protein in NPC tissues was significantly higher than that in chronic nasopharyngitis tissues (P<0.05) ( Table 1).

The mRNA and protein expressions of SDF-1α and CXCR4 in the NPC and chronic nasopharyngitis tissues
According to the results from qRT-PCR and Western blotting, the mRNA and protein expressions of SDF-1α and CXCR4 were appreciably elevated in the NPC tissues in comparison with the nasopharyngitis tissues. The difference was of statistical significance as seen in Figure 3 (P<0.05).

Positive expressions of SDF-1α and CXCR4 proteins in NPC tissues and their associations with clinicopathological features of NPC patients
As shown in Table 2  staging, N staging, TNM staging, skull base invasion, and cervical lymph node metastasis. The patients in the T3 + T4, N2 + N3, or TNM III + IV stages, or patients with skull base invasion and cervical lymph node metastasis showed appreciably higher positive expressions of SDF-1α and CXCR4 proteins than those in T1 + T2, N0 + N1, TNM I + II stages, or those without skull base invasion or cervical lymph node metastasis (all P<0.05). The study findings indicate that the expressions of SDF-1α and CXCR4 proteins had any correlation with other clinicopathological features (gender, age, histological type, and EBV-IgG status) of NPC patients (all P>0.05).

Correlation between the expression of SDF-1α protein and the expression of CXCR4 protein in NPC tissues
The Spearman rank correlation analysis was employed to analyze the correlation of the expression of SDF-1α protein and that of CXCR4 protein, the results shown in Table 3. In 102 NPC patients, 46 cases showed positive expressions of both SDF-1α protein and CXCR4 protein, while 27 cases had negative expressions of both SDF-1α protein and CXCR4 protein.
With the escalation of positive expression of SDF-1α protein, the positive expression of CXCR4 protein was also elevated, indicating that the expressions of CXCR4 and SDF-1α proteins were significantly positively correlated in NPC tissues (r=0.467, P<0.001). Note: -, negative; +, weak positive; ++, positive; +++, strong positive.

Associations of the expressions of SDF-1α and CXCR4 proteins with the prognosis of NPC patients
As shown in Figure 4, the Kaplan-Meier survival analysis was adopted to examine the association between the expression of SDF-1α and CXCR4 proteins and the survival of NPC patients. Patients having different expressions of SDF-1α and CXCR4 proteins showed difference in overall survival time. Patients with positive expression of SDF-1α had a 5-year survival rate of 39.1%, which was significantly lower than that (75.0%) of patients with negative expression of SDF-1α (P<0.05). Patients with positive expression of CXCR4 had a 5-year survival rate of 52.0% that was significantly lower than that (77.8%) of patients with negative expression of CXCR4 (P<0.05). Kaplan-Meier survival analysis also showed that, compared with NPC patients having negative expression of SDF-1α or CXCR4 protein, those with positive expression of SDF-1α or CXCR4 protein had a significantly shorter survival time (all P<0.05).

Multivariate analysis of the prognostic factors in NPC patients
As shown in Table 4, factors such as gender, age, histological type, EBV-IgG status, T staging, N staging, TNM staging, skull base invasion, and cervical lymph node metastasis that are likely to affect the prognosis of NPC patients were included into the multivariate Cox regression analysis. SDF-1α protein, CXCR4 protein, EBV-IgG status, T staging, N staging, TNM staging, skull base invasion, and cervical lymph node metastasis were independent risk factors for the incidence of NPC (all P<0.05). The present study did not find any correlation between other indicators (gender, age, and histological type) and the prognosis of NPC patients (all P>0.05).

Discussion
NPC has a high incidence rate in Southeast Asia and Southern China, and the primary therapy is non-surgical intervention treatment, and additionally, shows a high cervical lymph node metastasis incidence [21]. Using a marker for NPC, we can effectively understand the mechanism of its occurrence and development, determine the extent of tumor development in a timely manner, uncover an effective therapeutic target, and provide indicators for better prognosis detection. Aiming to explore the role of SDF-1α/CXCR4 signaling pathway in the clinicopathological features and prognosis of patients with NPC, the study conducted a series of experiments and finally concludes that the protein expressions of SDF-1α/CXCR4 might be associated with the clinicopathological features and and their high protein expressions were risk factors for the prognosis of NCP patients.
In the present study, we found that the up-regulation of the expressions of SDF-1α and CXCR4 proteins was related to the pathogenesis of NPC. CXCR4 is a specific G-protein-coupled 7-span transmembrane receptor for chemokine SDF-1α, and is necessary for the maintenance and renewal of stromal cells [22,23]. The SDF-1α/CXCR4 signaling pathway can promote multiple downstream signaling responses to regulate chemotaxis, intracellular calcium, cell survival and proliferation, and gene transcription [24]. Besides, the SDF-1α/CXCR4 axis plays a vital role in tumor cell proliferation, migration, angiogenesis, and immune surveillance of tissues [25]. Studies conducted by Wang et al. [26] have noted that the expression of SDF-1α/CXCR4 can be induced via the glycolytic enzyme phosphoglycerate kinase 1 (PGK1), which is secreted by tumor cells. Furthermore, mitogen-activated protein kinases (MAPK) pathway can regulate SDF-1α/CXCR4 in an alternative way, i.e. the activation of the MAPK/extracellular signal-regulated kinase (ERK) can induce the activation of ERKs and their signaling pathway that affects the nuclear transcription factors through the ERK signaling cascade, in order to regulate SDF-1α/CXCR4 expression [27,28]. Moreover, SDF-1α is released by cancer-associated fibroblasts and the high expression of SDF-1α can form a local gradient for chemokines in the tumor region further inducing the expression of CXCR4 [29]. Our study confirms that chemokine SDF-1α and its receptor CXCR4 constitute the SDF-1α/CXCR4 biological axis, playing an important role in the spread and specific organ metastasis of NPC and providing an important pointer for the clarification of NPC metastasis mechanism [27]. Luo et al. [30] also showed that the expression of CXCR4 was elevated in NPC cells, and the expression of SDF-1α was high in the deep cervical lymph nodes and distant target organs, such as bone marrow, lung, and liver, common NPC metastasis sites.
The present study also found that the expressions of SDF-1α and CXCR4 were associated with the clinicopathological features of NPC patients, and their positive expressions were risk factors for the prognosis of NPC patients. SDF-1α and CXCR4 can regulate a vast range of important physiological processes, such as brain development, embryogenesis, blood vessel formation, wound healing, and cell activation [31]. SDF-1α can induce, control, and regulate tumor vascular formation, additionally, it is directly involved in the endothelial cell angiogenesis process through cell proliferation, differentiation, sprouting, and tube formation, and play a synergistic role with vascular endothelial growth factor (VEGF) in promoting vascularization [32]. In the NPC cells, SDF-1α and CXCR4 can induce the expression of VEGF, which in turn can upsurge the expression of CXCR4 and promote the invasion and metastasis of cells, which is consistent with our results [33]. Futhermore, tumor cells secrete the PGK1, which restricts blood vessel formation, whilst in tumor metastasis foci, the high expression of CXCL12/CXCR4 pathway can down-regulate PGK1 expression and consequently overcome the vascular inhibition produced by PGK1 to promote tumor growth [34]. Another study showed that the positive expression rate of CXCR4 in NPC tissues was notably higher than that in non-NPC tissues, and it performed an important role in the clinical staging, M staging, lymph node metastasis, and development of NPC [30], which was consistent with the results in our study. Additionally, it has been confirmed that the high expression of CXCR4 was closely related to the poor prognosis of NPC and the expressions of SDF-1α and CXCR4 have been recognized as a potential target factor for the prevention of NPC metastasis and prognosis [35,36].
In conclusion, our study provides strong evidence that the high expressions of SDF-1α and CXCR4 proteins in SDF-1α/CXCR4 signaling pathway are related to the pathogenesis of NPC, and overexpression of SDF-1α and CXCR4 may be a risk factor for the prognosis of the NPC patients, indicating they can provide novel and effective strategies for the treatment of NPC. Nevertheless, cell assays were not conducted to analyze the cell changes such as cell migration, which is one of the limitations of the present study. Besides, the underlying mechanism of in what way SDF-1α and CXCR4 proteins affect the occurrence, development, and prognosis of NPC lacks clarity and needs further exploration.