Mining CK2 in Cancer

Cancer is a leading cause of death worldwide. Cancer cells proliferate uncontrollably and, many cases, spread to other parts of the body. A protein historically involved in cancer is protein kinase CK2. CK2 is a serine-threonine kinase that has been involved in cell growth, cell proliferation and cell apoptosis. CK2 functions as an oncogene when overexpressed in mouse tissues, and can synergize with known oncogenes, such as ras, to induce cell transformation in cells in culture. CK2, typically the CK2α protein, is found elevated in a number of human tumors. However, we have little information on CK2α' and CK2β proteins, and scarce information on CK2 gene transcript expression. Here, we explore the expression of CK2 transcripts in primary tumor tissues using the database Oncomine in the six cancers with the highest mortality in the U.S.A. In addition, we studied the correlation between CK2 expression and overall survival using the Kaplan-Meier Plotter database in breast, ovarian, and lung cancers. We found widespread upregulation in the expression of CK2 genes in primary tumor tissues. However, we found underexpression of CK2α' transcripts in some tumors, increased CK2β transcripts in some invasive tumors, and deregulation of CK2 transcripts in some tumor precursors. There was also correlation between CK2 expression levels and patient survival. These data provides additional evidence for CK2 as a biomarker for cancer studies and as a target for cancer therapy.


Cancer prevalence
Cancer remains a leading cause of morbidity and mortality worldwide, and it is predicted to be of more importance in the next few decades. According to the World Health Organization, if global cancer rates remain unchanged, by 2030, new cancer cases will reach 21.4 million-a rise from 12.7 million new cancer cases in 2008 [1]. According to the United States Cancer Statistics, the top ten cancers by incidence rate (Age-Adjusted Invasive Cancer Incidence Rates for the 10 Primary Sites with the Highest Rates) from 2005-2011 are cancers of the prostate, breast, lung and bronchus, colorectal, corpus and uterus, urinary bladder, melanomas of the skin, non-Hodgkin lymphomas, cancers of the kidney and renal pelvis, thyroid [2]. Among them, the cancers with the highest death rates in the USA from 2005-2010 include lung and bronchus, prostate, breast, colon and rectum, ovarian and pancreatic cancers [2]. These cancers were the focus of our analysis.

General properties and features of CK2
Protein kinase CK2 is a highly conserved serine/threonine kinase linked to diseases such as cancer [3], cardiac hypertrophy [4], multiple sclerosis [5], and inflammation [6]. Furthermore, it has been shown to be critical in embryonic development [7], and circadian rhythms [8].
CK2 is encoded in mammals by two separate genes (CK2a and CK2a'); CK2a has broader tissue expression and higher-level expression compared to the CK2a' [18]. CK2 proteins can function as monomeric kinases or within a tetrameric complex of two CK2 kinase units and two units of the regulatory protein CK2b. In this tetrameric form, CK2b confers different substrate specificity to CK2a [19].
CK2a is an oncogene in mouse transgenic models, as CK2a overexpression in mammary gland and lymphoid compartment leads to tumor formation [20,21]. In accordance, CK2 is overexpressed in proliferating cells [22]. In addition, there is a link between CK2 expression and environmental toxins. In animal models, CK2 is overexpressed in response to aromatic hydrocarbons [23], and is upregulated in chemically-induced liver carcinogenesis [24].
A limited number of studies testing CK2 protein and activity in human tumors found that CK2 activity and protein expression is upregulated in human cancers [25,26] (reviewed in [22]). Although we have some knowledge on the expression of CK2a protein, we have little information on the expression of CK2a' or CK2b. Compared to protein expression levels, only a few studies test the expression levels of CK2 gene transcripts [25]. In fact, it is thought that CK2 protein and not CK2 gene transcript are elevated in cancer [22,27]. In addition, only recently it has been shown that there is deregulated expression of the CK2a pseudogene [28], a non-protein coding CK2a sequence [29]. CK2 can serve as a prognostic marker [12], and can strengthen the multi-drug-resistant phenotype in adult cancer tissues [30], indicating the importance of studying the expression levels of CK2 genes.
Therefore, we extracted data on transcript expression for CK2a, CK2a pseudogene (CK2aP), CK2a' and CK2b from the database Oncomine for lung and bronchus, prostate, breast, colon and rectum, ovarian and pancreatic cancers, focusing on clinical specimens of cancer vs. normal patient datasets, and separated by subtype when possible (Fig. 1). We also tested for the effect of deregulation of CK2 expression on patient overall survival using Kaplan-Meier Plotter. Here we present a summary of the results we obtained. The tables associated with the data include statistical significance, fold change, patient number and the dataset from which the information was obtained.

Oncomine analysis
The expression level of CK2 genes in the selected cancers was analyzed using Oncomine [31]. For this, we compared clinical specimens of cancer vs. normal patient datasets. In order to reduce our false discovery rate, we selected p,0.01 as a threshold. We analyzed the results for their p-values, fold change, and cancer subtype. In many instances we found several significant correlations in different tumor types, but we showed only one to three of them.

Kaplan-Meier plotter analysis
The prognostic value of the CK2 genes in ovarian, breast and lung cancer was analyzed using Kaplan-Meier Plotter (http://kmplot.com/analysis/), a database that integrates gene expression data and clinical data [32]. To date, Kaplan-Meier Plotter contains information on 22,277 genes and their effect on survival in 2,977 breast, 1,464 ovarian and 1,715 lung cancer patients. We focused our analysis on overall survival patient information. The patient samples have been split into two groups. The two patient groups (higher and lower expression levels) were compared using a Kaplan-Meier survival plot. The hazard ratio with 95% confidence intervals and log rank p value was calculated. We analyzed the best specific probes (JetSet probes) that recognized CK2a, CK2aP, CK2a', and CK2b. In order to reduce our false discovery rate, we selected p,0.01 as a threshold.

Lung cancer
Lung cancer has the highest mortality of all cancers in both men and women. The NCI predicts that in 2014, approximately 27% of all cancer deaths would be due to lung cancer. The NCI (http://www.cancer.gov/) [33] predicts that there would be over 224,000 new cases of lung cancer (about 14% of cancer diagnoses) in 2014, albeit a decrease in lung cancer incidence over the past few years (particularly in men). Lung cancers are divided into three groups. Non-small cell lung cancer (NSCLC) is the most common type of lung cancer (85% of lung cancers) and includes subtypes: squamous cell carcinoma, adenocarcinoma, and large cell carcinoma (LCLC). Small Cell Lung Cancer (SCLC) (10%-15% of lung cancers) quickly divides and metastasizes. Lung Carcinoid Tumors (5% of lung cancers), a type of neuroendocrine tumor, grow slowly and rarely spread. Advances in surgical and combined therapeutical approaches have increased the one-year survival by 7% (to 44% by 2005-08); however, the five-year survival rate for the combination of all stages is still low (16%); with small cell lung cancer having the lowest five-year survival rate (6%) compared to non-small cell lung cancer (18%).

CK2 in lung and bronchus cancer
Oncomine analysis of neoplastic vs. normal tissue showed that CK2a, CK2a', CK2b and CK2aP were significantly overexpressed in different types of lung cancer in different datasets (Table 1, Fig. 2). Overexpression of CK2a and CK2a' was found in squamous cell carcinoma and adenocarcinoma. In contrast, LCLC and SCLC had CK2a overexpression but not CK2a'. CK2b was upregulated in adenocarcinoma and LCLC but was downregulated in lung carcinoid. These data are in line with published articles on CK2 expression. For example, CK2a (CSNK2A1) transcript, protein and/or activity are elevated in lung cancer neoplastic tissues such as those found in squamous cell carcinomas of the lung [34,35,36], and in NSCLC cell lines [28,34]. In addition, the CK2a pseudogene (CSNK2A1P) is amplified in non-small cell lung cancer cell lines and lung cancer tissues [28].
Kaplan-Meier plotter analysis in overall lung cancer showed correlation between overexpression of CK2b and overall lower survival rates, and the opposite for CK2a' and CK2aP, and no significance for CK2a expression (Fig. 3). When we restricted the analysis by tumor type, no significant differences in overall survival were observed for the expression of CK2 genes in squamous cell carcinoma samples. However, in adenocarcinoma there were significant differences for CK2aP, CK2a', and CK2b mirroring the data on overall lung cancer. This indicates that the usage of CK2 gene expression as a predictor of outcome may be cancer subtype dependent (Table 1). Even though our Kaplan-Meier analysis of squamous cell carcinoma showed no significance for CK2a (p50.012), a published study identified CK2a (CSNK2A1) overexpression as one of two independent predictors of outcome (prognostic markers) in 21 patients with squamous cell carcinoma of the lung [34]. In particular, this was one of 11 genes specific of the poor prognosis cluster group (cluster II) of non-small cell lung cancers that had corresponding chromosomal aberrations. The authors validated these results independently in an additional set of 45 patients.

Breast cancer
Excluding cancers of the skin, breast cancer is the most frequently diagnosed cancer in women. According to the NCI, an estimated 40,000 breast cancer deaths in both women and men are expected in 2014 (6.8% of all cancers) [33]. Breast cancer ranks second as a cause of cancer death in women (after lung cancer). An . Breast cancer can happen in the ducts, the lobules and in the stroma. The most common breast cancers are ductal carcinoma (70% of cases) and lobular carcinoma (10% of cases). These can be subdivided into non-invasive (in situ) and invasive. Mucinous breast carcinoma is a subtype of ductal carcinoma (2-5% incidence) Breast cancer treatment can involve surgery, radiation therapy, chemotherapy, hormone therapy and/or targeted therapy. The overall 5-year survival rate is 89.2% in localized or low-grade breast cancer. Early detection and novel treatments have increased the survival rate for localized breast cancer to 98.6% and for regional breast cancer to 84.4%.

CK2 in Breast Cancer
Overall, Oncomine analysis showed significant levels of both CK2a and CK2b overexpression in most breast cancer types, but strong CK2a' underexpression in all breast cancer types ( Table 2, Fig. 4). A difference was found between ductal and lobular breast cancers, where ductal, but not lobular, expressed high levels of CK2a. A further difference was found in between non-invasive and invasive breast cancers. Both showed CK2a' underexpression, but in invasive breast cancer, CK2b was also very significantly overexpressed. Since CK2b protein is stabilized posttranslationally by CK2a [9], increased CK2b transcript expression in CK2a overexpressing cancers, may contribute to further increase CK2b protein expression.
Kaplan-Meier analysis revealed that in overall breast cancer high levels of expression of CK2a and CK2b correlate with lower survival rates (Fig. 5). CK2a' expression had no significant influence in survival in all breast cancers (Fig. 5). This is in line with a published study, CK2a (CSNK2A1) was found as part of a 186-gene ''invasiveness'' gene signature in breast cancer. This ''invasiveness'' gene signature was associated with overall survival and metastasis-free survival in patients with breast cancer, and when combined with the NIH prognostic criteria, the ''invasiveness'' gene signature was used to stratify patients with high-risk early breast cancer. This ''invasiveness'' gene signature was associated with overall survival and metastasis-free survival in patients with medulloblastoma, lung cancer and prostate cancer [37]. In addition, it has been found that unbalanced expression of CK2 genes is able to promote epithelial to mesenchymal transition in cell lines [38]. These studies emphasize the importance of analysis of the expression of CK2 genes during cancer progression.

Colon cancer
Colon cancer is the third leading cause of cancer related deaths when men and women are considered separately, and the second leading cause when both considering both sexes. The NCI predicts that in 2014 there would be an estimated 50,310 deaths in men and women, which would be 8.6% of all cancer deaths [33]. They have estimated 136,830 new cases of colorectal cancer in 2014. When combined, colorectal cancers make up 8.2% of all new cancer cases. The most common type of colorectal cancers is adenocarcinoma (95%). Its precursor, adenomatous polyps (adenomas) are polyps that are considered to be a precancerous condition. Treatments of this cancer type include surgery, chemotherapy, biological therapy, and radiation therapy. Early detection of colorectal cancers can increase survival rate when comparing localized colorectal cancers (90.3%) and regional colorectal cancers (70.4%). The relative 1-year relative survival rate is 84% and the relative 5-year survival rate for colorectal cancer is 64.9%.

CK2 in Colon cancer
Oncomine analysis revealed many cases of overexpression for all three CK2 genes and the CK2a pseudogene in adenocarcinomas (Table 3, Fig. 6). We found a statistically significant increase in CK2a in colon adenocarcinoma. In agreement with this finding, a published study shows that expression levels of CK2a transcripts are elevated in 10 pairs of colorectal tumor tissues compared to nontumor tissues [39]. Oncomine analysis found that there is also overexpression of CK2 genes in adenomas (Table 4). This is relevant, because CK2 has been shown to regulate Wnt/b-catenin signaling [7]. Wnt/b-catenin signaling is major player in colorectal cancer progression as it is proposed to be the first hit to be deregulated in colon cancer progression [15].

Ovarian cancer
Ovarian cancer ranks fifth in cancer death among women (3% of all cancers among women), accounting for more deaths than any other cancer of the  80-85%, endometrioid, which accounts for ,10%, and clear cell, which accounts for ,5% of all epithelial ovarian cancers. The less-common types of ovarian cancers include germ cell tumors (less than 2% of ovarian cancers), and stromal tumors (1% of ovarian cancers). The treatment methods of ovarian cancer include surgery, chemotherapy and radiation. A woman's risk of developing ovarian cancer in her lifetime is 1 in 72, and the age at which she is diagnosed affects survival rate. 56% of women younger than 65 years survive 5 years following diagnosis, while 27% of women older than age 65 survive 5 years following diagnosis. Overall, the 1-year relative survival rate of all ovarian cancer patients is 75% and the 5-year relative survival rate is 44%. If the tumor is localized, there is a   91.9% 5-year relative survival rate, but if the tumor is regional, the survival rate drops to 72.0%.

CK2 in Ovarian cancer
Oncomine analysis of neoplastic vs. normal tissue showed that CK2a and CK2b were significantly upregulated in several ovarian cancer types. CK2a' was found significantly downregulated in two ovarian tumor types. There were no significant results for CK2aP (Table 5, Fig. 7). Kaplan-Meier plotter analysis revealed that high-levels of expression of CK2a correlated with lower patients survival rates. In contrast, although CK2a' was significantly downregulated in some adenocarcinomas and CK2b that was upregulated in a few subtypes, Kaplan-Meier plotter analysis did not show any significance of CK2a' and CK2b expression levels in overall survival (Fig. 8).

Prostate cancer
Prostate Cancer is the most common cancer among men, after skin cancer, according to the NCI. The NCI predicts that in 2014, there would be 233,000 new cases of prostate cancer, accounting for 14% of all new cancer cases [33]. The NCI predicts 29,480 deaths in 2014, which would account for 5% of all cancer deaths. The most common type of prostate cancer is adenocarcinoma, which accounts for about 90% of all prostate cancers. There are subtypes of adenocarcinoma including atrophic, foamy, colloid, and signet ring carcinoma. The remaining 10% of all prostate cancers include transitional cell cancer, squamous cell cancer, carcinoid, small cell cancer, and sarcomas and sarcomatoid cancers. Treatments for prostate cancer include surgery, and different therapies such as radiation, hormone, chemotherapy, chemoradiotherapy, and immunotherapy. There is an extremely high survival rate overall. When caught early and the tumor is either localized or regional, there is a 100.0% chance or a 5-year relative survival. Overall, the relative 5-year relative survival rate for localized or low-grade cancer is 99.2%.

CK2 in Prostate cancer
In prostate cancer, Oncomine analysis revealed significant overexpression of CK2a in three subtypes, while CK2a' and CK2b were found only upregulated in one cancer subtype each. There were no significant results for CK2aP (Table 6, Fig. 9). This data are in line with publications showing significant increase in CK2a expression in prostate tumor lesions (p50.0224) [40], CK2 activity correlating with Gleason grade [41], and nuclear localization of CK2a associating with poor prognosis [42].

Pancreatic cancer
Pancreatic Cancer is the fourth-leading cause of death from cancer in the United States, after lung cancer, colon cancer, and breast cancer, but it is the 12 th most common type of cancer in the United States. The NCI estimates that there would be an estimated 39,590 deaths due to pancreatic cancer in 2014, which accounts for 6.8% of all cancer deaths [33]. They estimated that there would be 46,420 new cases, making up 2.8% of all new cancer cases. There are two types of pancreatic cancer cells, pancreatic exocrine tumors and pancreatic neuroendocrine tumors. About 95% of pancreatic cancers are classified as exocrine tumors, which has many different subtypes such as adenocarcinoma, and acinar cell carcinoma.
Adenocarcinoma accounts for about 90% of all pancreatic cancers, and the other subtypes are considered to be rare. Pancreatic neuroendocrine tumors account for less than 5% of pancreatic tumors, and have much better survival than pancreatic adenocarcinoma. Treatments for people with adenocarcinomas include surgery, chemotherapy, targeted therapy, and radiation therapy. Despite these treatments, the relative survival rate is fairly low. Even when the tumor is localized or regional, patients have a relatively low 5-year relative survival rate, which is 24.1% and 9.0% respectively. Accounting for all types of pancreatic cancer, there is a 6.0% that a patient can survive 5 years.

CK2 in Pancreatic cancer
Oncomine analysis revealed that only CK2b was significantly upregulated in pancreatic adenocarcinoma. CK2b was downregulated in pancreatic adenocarcinoma while CK2a' was downregulated in pancreatic ductal adenocarcinoma (Table 7). CK2a was overexpressed in both types of pancreatic cancer but not significantly (p-values around 0.02 to 0.03). CK2aP was not significantly changed.

Conclusions
It was reported that CK2a protein but not transcripts are deregulated in cancer [22,27]. In our analysis of Oncomine we found that not only CK2a, but also  CK2a', CK2aP, and CK2b transcripts were deregulated in various cancer types (summarized in Table 8). This discrepancy may be due to the fact that these original publications reported data from a limited number of tumor samples (4-8 tumor samples) from two tumor types [22,27]. However, more recent publications using larger patient samples show that CK2a transcripts are upregulated in lung and colon cancers [34,39]. Together with our study, these data suggest that deregulated CK2 gene transcript expression may be a mechanism underlying the increase in CK2 activity and protein levels detected in human tumors. However, posttranscriptional and posttranslational mechanisms may also be involved. In general, there is little data on the regulation of CK2 gene expression, and presumably both genetic and epigenetic mechanism may be involved. A number of studies show that regulation of CK2 transcription happens at the promoter level for CK2a and CK2b [43,44,45]. However, there is no data on whether CK2 genes themselves are controlled by epigenetic mechanisms. Still, since CK2 controls epigenetic regulators [46,47,48], there is a possibility that CK2 could control itself though epigenetic regulation. Nonetheless, there is no data on whether altered CK2 expression in cancer is due to genetic or epigenetic mechanisms. Identifying the genetic and epigenetic mechanisms that control CK2 expression will help understand the role of CK2 in biological processes and in disease.
Our data indicates that deregulated CK2 expression is an important factor during tumorigenesis. In our Oncomine analysis, the relative fold change in the expression of CK2 genes differs between tumors. In addition, we found a strong correlation between particular CK2 gene expression and certain subtypes of cancer. For example, CK2a' was found underexpressed in breast, ovarian and pancreatic cancers. In addition, while both ductal and lobular breast carcinoma underexpressed CK2a', ductal but not lobular breast carcinoma had upregulation of CK2a. Likewise, the CK2b transcripts were upregulated in both ductal and lobular breast carcinoma when they became invasive. This suggests that CK2b may be required to progress into invasive cancer. Another example is the lack of overexpression of CK2aP in non-neoplastic colon adenomas compared to its overexpression in all three of the colorectal adenocarcinomas.
In contrast with this study on transcript levels, we know relatively little about the expression of CK2 proteins in human cancer and its role in tumorigenesis. A number of publications study CK2 activity and protein levels in a few samples from selected human tumors. As discussed by Tawfic [22], increases on CK2 in cancers based on activity levels alone have to be considered carefully while studies of both protein levels and activity are better indicators of CK2 increases in cancers. In line with our findings, different fold change on the expression of CK2a and CK2b genes is found in kidney tumors [49]. Importantly, in head and neck tumors the level of CK2 correlated with tumor grade and stage [50]. However, a thorough analysis of the specific localization of CK2 transcripts and/or protein is needed to understand the role of CK2 in cancer. In this regard, inmunological studies show CK2 located in the leading edge of the tumor in heterotransplanted human colon tumors in mice [51]. Intriguingly, expression of nuclear CK2a protein was higher in colorectal adenocarcinomas than in adenomas compared to normal tissues, which suggests that CK2a's subcellular localization maybe involved in progression from adenomas to malignant tumors [52].
Our analysis of Kaplan-Meier Plotter also found correlations between CK2 gene deregulation and survival rates (Table 9). Overall, having higher levels of CK2 gene expression led to lower survival amongst patients with lung, breast, and ovarian cancer. Similarly, in head and neck tumors the level of CK2 also correlated clinical outcome [50]. However, in lung adenocarcinoma, displaying a higher level of CK2a' and CK2aP correlates with higher survival rates.
CK2 is an emerging key target in cancer therapy by impacting not only cell growth and proliferation but also anti-apoptotic activity in malignant cells. Indeed, CK2 inhibitors are being tested for their effectiveness in cancer treatment [11,27,53,54]. In addition, CK2 inhibitors are also studied in conjunction with other antitumor agents such as Melphalan and Imatinib [53,55]. The database mining findings here greatly support the use of CK2-targeted therapies to treat CK2 upregulation in cancer. For example, CK2a,a' and b genes were found significantly deregulated in colorectal adenomas and adenocarcinomas (this study and [39,52]), therefore CK2 inhibitors could be used to prevent both the start and progression of colon cancer. Our data indicates that, for some types of cancers, a CK2b-dependent CK2 targeted therapy may be more effective. This is the case of breast cancers, where CK2b overexpression correlated with invasive types of breast carcinoma. This property may be exploited to design more effective CK2 inhibitors. In this regard, compounds such as the inhibitor MNA that inhibits more efficiently CK2 holoenzyme (IC5050.3 mM) than monomeric CK2a (IC5052.8 mm) could be an effective CK2b-dependent CK2 targeted therapy [56,57]. Our data also indicates that for some cancers such as lung adenocarcinoma, the protective effect in patient survival of CK2a' may be counteracted by CK2 inhibitory drugs. In this case, CK2 inhibitors that target specifically CK2a may not affect the protective role of CK2a'. The importance of our analysis on CK2 gene transcript expression on cancer development, and on cancer patient diagnosis, treatment and survival needs to be further evaluated. First, the correlation between transcript and protein upregulation for different tumor types needs to be researched, as it is protein activity that will be the target of therapy. Second, determining the specific localization of CK2 proteins and transcripts changes in tumor tissues. Third, additional research on the in vitro and in vivo tumorigenic potential the understudied CK2 genes (CK2a', CK2b and CK2aP). Fourth, study the effect on tumorigenesis of different fold changes in the expression of CK2 genes, in particular incorporating the knowledge on CK2 protein localization in tumors for in vivo studies. Fifth, further research of ovarian, prostate and pancreatic cancer as they had a comparatively small number of samples per dataset, and also analyzing other types of cancer. Sixth, additional research on the impact on patient survival of changes in the CK2 genes in additional cancer types. Seventh, adjusting CK2 therapy development to CK2 cancer-specific expression. All together, these studies could lead to cancer diagnostic and predictive tools, and help develop more effective and specific cancer treatments.