More precise prediction in Chinese patients with penile squamous cell carcinoma: protein kinase CK2α catalytic subunit (CK2α) as a poor prognosticator

Purpose In this study, we assess the CK2α expression in human penile squamous cell carcinoma (SCC) and its clinical significance. Methods A total of 157 human penile SCC tissue samples were immunohistochemically analyzed. In addition, 12 human penile SCC and adjacent normal tissues were examined for CK2α protein and mRNA expression by Western blotting and real-time quantitative PCR, respectively. Survival was analyzed using the Kaplan-Meier test and the log-rank test. Multivariate Cox proportional hazard regression analysis was performed to determine the impacts of CK2α expression and the clinicopathological features on patient disease-specific survival (DSS). Likelihood ratios (LRs), Akaike information criterion (AIC) values, and concordance indexes (C-indexes) were investigated to evaluate the accuracies of the factors. Bootstrap-corrected C-indexes were used for internal validation (with sampling 1000 times). Results A significant difference in the distribution of CK2α was observed between the normal and penile carcinoma tissues (P<0.001). CK2α expression was associated with the pathological T and N stages in the penile cancer tissues (P<0.001). High CK2α expression was with significantly poorer DSS compared with low expression one (P<0.001). Western blotting and real-time quantitative PCR also confirmed that CK2α expression was increased in the penile cancer tissues. In multivariate Cox regression analysis, CK2α overexpression still was one of independent prognostic factors for penile SCC (P=0.005). The predictive accuracy of CK2α was verified by analysis of the C-indexes. Conclusion High protein kinase CK2α expression is associated with several prognostic factors and is thus a significant indicator of poor prognosis for penile cancer.


INTRODUCTION
Penile cancer is a rare tumors in the Europe and United States, with an incidence of less than 1.00 per 100,000 males [1]. However, it is still a major public health problem in non-Western countries, such as Uganda, Brazil, and India [2,3]. The main pathology type( ≥95%) of penile cancer are squamous cell carcinomas (SCCs), and the lymph node metastasis (LNM) are considered as the most important predictors of survival [2]. In contrast with other genitourinary cancers, penileSCC is typically characterized by stepwise regional lymph node (LN) spread, which is clearly related to survival [1,2,4,5]. Nevertheless, the outcomes vary among patients with pathologically proven nodal disease. In addition, the survival rates of these patients exhibit large variations according to the extent of nodal involvement [2,4,6]. Thus, the detection and treatment of these metastases as soon as possible and improved prediction of survival are important for clinicians managing these patients.
The stage, grade of differentiation and clinicopathological factors of a primary tumor are well known to be predictive factors for LNM and survival [2,6,7]. However, deciding to perform regional node dissection based solely on these characteristics leads to unacceptable false-negative and false-positive rates. Inguinal lymph node dissection (ILAD) alone is curative in many cases with minimal metastases [2-4, 6, 7]. Nevertheless, a considerable number of men with palpable nodes do not experience metastatic spread, and consequently, up to 80% may be overtreated [8,9], with resulting high morbidity [10,11]. However, following a wait-and-see policy in treatment of these patients poses the risk of metastasis development to a stage at which a cure is no longer possible [12]. Thus, increased understanding of the molecular factors and biological features is of paramount importance in penile SCC [13,14].
Protein kinase CK2 is a highly conserved and ubiquitous serine/threonine kinase composing with two catalytic subunits and two regulatory subunits [15]. CK2 is as a multifunctional protein kinase that is widely distributed in substrates [16,17]. Disruption of Saccharomyces cerevisiae genes encoding CK2 catalytic subunits is a lethal event, and it plays major roles in cell cycle progression, proliferation, and differentiation processes [18,19]. In addition, the clinical significance of CK2α subunits have been assessed in human cancers [20][21][22][23][24]. Thus, CK2α is now considered a potential target in the treatment of human malignancy [25]. As weʼve seen, the prognostic significance of nuclear CK2α in penile SCC is still unknown.
In this pilot study, we evaluated CK2α expression in primary penile SCC. In addition, the relationships between CK2α expression and clinicopathologic features were examined, as well as the clinical prognostic value of CK2α in penile SCC.

RESULTS
After a median time of 25 mo (range 1-133 mo), 43 patients died of penile cancer. The patients' clinical characteristics are shown in Table 1.
Immunohistochemical analysis revealed that CK2α was expressed in both the cell nucleus and cytoplasm ( Figure 1A, 1B), and cancer tissues were with high CK2α expression. Similar to the immunohistochemistry results, CK2α expression was significantly higher in the penile cancer tissues in 12 normal-tumor tissue pairs by Western blotting (Figure 2A, 2B) (P<0.001). Additionally, quantitative real-time PCR results also proved that CK2α expression was significantly increased in the tumor tissues over the non-tumor ones ( Figure 2C).
The results of univariate analyses related to DSS are shown in Table 1. They indicated that the T stage and N status (all P < 0.001), tumor grade (P=0.022) and CK2α overexpression (P<0.001) were significantly correlated with DSS. Further, Cox multivariate analysis revealed that CK2α overexpression was an independent prognostic predictor ( Table 2, HR=8.234, 95% CI=1.912-35.460, P=0.005). CK2α overexpression showed the best predictive accuracy for the LR, C-index and AIC among the basic factors. Accounting for CK2α expression, the predictive accuracy of the basic factors were with significantly increased (Table 3).

DISCUSSION
CK2α expression has been assessed in a part of turmors, including prostate, colorectal and lung cancers    Significance level: P<0.001. www.impactjournals.com/oncotarget [20][21][22][23][24]. However, with limited clinical data, the clinical significance of CK2α hasn't been studied. In the pilot study, we investigated the expression of CK2α in penile tissues from patients with penile cancer. To our knowledge, CK2α expression in human penile tissues is unknown, and it has not yet been measured in human penile cancer. In the present study, significant differences in CK2α expression were detected between penile cancer tissues and non-tumor tissues ( Figure 2). Several reports have consistently demonstrated that the CK2 level are increased in a variety of cancers [22,23]. A recent study has shown that nuclear CK2α expression in tumor colorectal tissues was significantly increased compared with that in non-tumor colorectal ones [22]. Another previous study involving cancer has demonstrated that CK2α accumulates in cancer cells and plays a key role in promoting cell survival [20]. Some studies considered that increasing CK2α expression might be related to the increased proliferative capacity of tumor cells [25][26][27][28].
For the first time, the results of our study showed that CK2α overexpression was closely correlated with the pathologic tumor status (P=0.001) and pathologic nodal status (P<0.001) ( Table 1). Although the mechanisms underlying these associations of CK2α are still unclear, these correlations are supported by several current lines of evidence. Mottet et al. have shown that blocking of the activation of hypoxia-inducible factor-1 (HIF-1) leads to CK2 overexpression and subsequent VEGF-C inhibition [29]. In addition, HIF-1 has been shown to stimulate   [31]. Constitutively, activation of STATs, which play the key role in the growth and differentiation, facilitate neoplastic behaviors of a number of cancers [32,33]. Yao et al. have showed that CK2 activity suppresses cell proliferation and even induces cell death in both LNCaP and PC-3/AR human prostate cancer cells compared with mock-transfected control cells [21]. As a CK2 catalytic subunit, CK2α may play an important role in the marked resistance of cancer cells to apoptotic signals. This array of cellular responses is broadly orchestrated by CK2-dependent regulation of key signaling molecules in the Hedgehog/Gli, NF K B, Wnt, PI3K/AKT/mTOR and hypoxia pathways that have been multifunction [34][35][36][37].
Our study also demonstrated that CK2α expression can function as an independent of prognostic markers for penile cancer (Tables 2 and 3). However, clinical data pertaining to the clinic value of CK2α are limited. Laramas et al. had shown that CK2α protein expression is elevated in prostate cancer by immunohistochemical analysis only [38]. In 2011, Lin et al. demonstrated that overexpression of nuclear CK2α in colorectal cancer tissues was correlated with several clinicopathologic factors and that it could be used as an independent prognostic marker for colorectal cancer [22]. Further, a study in which global gene expression profiling was performed on 190 human tumors of 14 cancer types resulted in the identification of 11 genes, including CK2, and the results suggested that the expression of CK2 was strongly correlated with that of MT1-MMP, in direct association with tumor prognosis [39]. However, no report of the prognostic significance of CK2α in penile SCC has done. Importantly, our results showed that CK2α overexpression could be used as an independent prognostic marker for penile cancer. In addition, accounting for CK2α expression in multivariate analysis of the basic factors had a positive impact on predictive accuracy (Table 3).
In summary, this pilot study revealed that high CK2α expression was significantly associated with prognosis in penile cancer treated with surgery. Thus, patients with penile cancer exhibiting CK2α overexpression should be carefully monitored.

Patients
After being approved by the institutional review board, patients with penile carcinoma treated at Sun Yat-sen University Cancer Center from March 2002 to March 2013 were retrospectively analyzed. A total of 43 patients presented with neoadjuvant chemotherapy and/ or radiotherapy, and 87 presented with metastatic disease. Patients were excluded if they had received palliative therapy or if their pathological data were incomplete. Ultimately, 157 patients with bilateral ILND were included in the study population.
The non-tumor tissues were obtained by separation of the grossly normal tissues from the tumor tissues in the resected specimens. After uropathological review, the pathologic features were determined anew according to the AJCC 7th TNM staging system. The histological grade was classified according to the Broder system . Long term follow -up: every 3 month for the first 2 yr after surgery, every 6 mo during the 3rd -4th years, and then once yearly thereafter. The follow-up deadline was December 2015.

Immunohistochemistry and interpretation
Mastadenoma and omission of the primary antibody were used as positive and negative control. 3% H 2 O 2 was using for blocking endogenous peroxidase for 10 min, and 10% normal goat serum was using for blocking nonspecific binding of immunoglobulins for 10 min. With appropriately diluted primary antibodies, sections were placed at 4°C for one night. The slides were subsequently incubated with a primary antibody, clonal anti-CK2α (Millipore, Massachusetts, USA 1:50 dilution) at 25°C for 30 minutes. After 5-8 minutes incubation with 3,3′-diaminobenzidine, the sections were counterstained with hematoxylin and microscopic analysis.

Quantitative real-time PCR
12 tumor and non-tumor pairs of colorectal tissues using for RNA with an RNA extraction kit (TAKARA, Dalian, China). RNA quality was analyzed using an Agilent Stratagene Bioanalyzer (Palo Alto, California, USA). The RNA integrity numbers (RINs) of all 24 samples indicated that the RNA quality was good, and the samples were then used for cDNA synthesis, after which they were stored at -20°C until further use. The amplification mixture (25 μl) contained 0.030 ng cDNA (3 μl), 1× TaqMan Master Mix, and 1× TaqMan Gene Expression Assay Primer/Probe Mix (TAKARA, Dalian, China). The thermal cycling parameters were as follows: 40 cycles at 95°C for 2 min, 95°C for 30 sec, 60°C for 30 sec, and 72°C for 30 s for pre-denaturation, denaturation, annealing and extension, respectively. The sequences of the primers used for PCR amplification designed according to GenBank mRNA sequences were TGTTCGTCATGGGTGTGAAC (forward) and ATGGCATGGACTGTGGTCAT (reverse).

Statistical analysis
Receiver operating characteristic (ROC) curves were generated to identify the best point for prediction of patient prognosis. The correlations between CK2α expression and the clinicopathologic parameters were examined using Pearson's correlation analysis. DSS was calculated using the Kaplan-Meier method, and assessed by log-rank tests. Univariate and multivariate Cox proportional hazard regression analyses were performed to determine the impacts of the clinical and pathological factors on patient DSS with removal of nonsignificant factors in a stepwise backward fashion. Likelihood ratios (LRs), Akaike information criterion (AIC) values, and concordance indexes (C-indexes) were investigated to evaluate the accuracies of the factors. Bootstrap-corrected c-indexes were used for internal validation to better gauge expected future predictive accuracy (with sampling 1000 times). All data were analyzed using R2.11.1 (http:// www.r-project. org), with statistical significance set at P<0.05.