Despite the use of combination therapies including surgery and other systemic treatments (radiotherapy, endocrine therapy, and anticancer agents), many breast cancer patients will ultimately develop metastatic disease, which remains to be essentially incurable. Therefore, identification of novel prognostic and predictive biomarkers for breast tumors is needed and remains a long awaited priority to enhance individualized treatment. Recently, we identified molecular gene signatures associated with aggressive breast cancer . Here, we present a validation of the prognostic role of five selected candidate biomarkers (CCNB2, ASPM, KIAA0101, CDCA7, and SLC27A2) included in these gene signatures using an independent breast cancer cohort.
In the present study, elevated CCNB2 protein levels were significantly associated with shorter DSS. Prominently, the HR of patients whose tumors were CCNB2-positive was 6.1 corresponding to a dramatic increase in probability of mortality compared with patients whose tumors had very little or no expression.
It is well-known that different cyclins regulate the eukaryotic cell cycle in association with cyclin dependant kinases (cdks) at specific points of the mitotic cycle. Cell cycle progression follows periodic alternations in the protein levels of cyclins, D, E, A and B. Deregulated levels of cyclins have been frequently observed in breast cancer . Elevated levels of cyclin E was significantly correlated with DSS in patients with breast cancer . Amplification and/or overexpression of cyclin D1 have been correlated to poor prognosis in breast cancer patients . The Cyclin A upregulation was reported to be associated with increased risk of recurrence among breast cancer patients with node negative tumors . The CCNB2 gene is a member of the B-type cyclin family, including cyclin B1 and B2. It is involved in the G2-M transition in eukaryotes by activating CDC2 kinase and its inhibition induces cell cycle arrest [9, 22, 23]. In agreement with a crucial role in cell growth, numerous studies detected overexpression of CCNB2 in human tumors, including lung, colorectal adenocarcinoma, and pituitary adenomas [24–27]. Serum circulating CCNB2 mRNA levels were found to be higher in lung and digestive tract cancer patients compared to normal controls and were correlated with cancer stage and metastasis status . Furthermore, the CCNB2 gene was included in the set of genes detected in node-negative breast tumors associated with poor prognosis . Our results suggest that CCNB2 has an oncogenic potential and its overexpression may give some proliferative advantage. We have shown a significant association of CCNB2 protein expression with breast tumor type, indicating that the CCNB2 protein levels were unevenly distributed among the histological types. Notably, no statistically significant differences were identified between CCNB2 protein levels in relation to patient age, tumor size, tumor grade, ER/PR status, HER2/neu, stage, and axillary lymph node status. However, a dissimilar trend has been reported in serum circulating CCNB2 mRNA, since it was found to correlate with both grade and metastasis status . The multivariate analysis including CCNB2 and several clinicopathological parameters further verified that CCNB2 remained an independent prognostic indicator for DSS. These data indicate that tumors with histological grade (I, II and III), axillary lymph node status (positive, negative), tumor size (0–2, 2–5, and >5), ER/PR status (positive, negative) and HER2/neu status, exhibiting CCNB2 protein expression, have a more unfavorable prognosis, with an increased risk of overall shorter survival rates. Moreover, the predictive power of CCNB2 in addition to the clinicopathological parameters model was slightly higher compared to the lower C-index of the model including all clinicopathological parameters alone. Thus, the accuracy in patient prognosis may be improved by measuring CCNB2 expression in cases of breast cancer.
The ASPM protein was reported to participate in spindle organization, spindle orientation, mitotic progression, and cytokinesis [10, 11, 30, 31]. Furthermore, the ASPM protein is over-expressed in various cancers [10, 32–35] and its knockdown inhibits tumor proliferation . There was no association between elevated ASPM protein levels and DSS or any other clinical parameters, suggesting that ASPM may be involved only in cancer initiation. A significant correlation between CCNB2 and ASPM was observed. The induction of ASPM and CCNB2 has previously been reported to regulate the G2/M cell cycle progression . Thus, elevated levels of CCNB2 may reflect a functional correlation with ASPM overexpression, which could play a role in the progression of breast carcinoma. Furthermore, up-regulation of CCNB2 and ASPM was detected in glioblastoma multiforme xenograft tumors and de novo glioblastoma multiforme tumors . Activation of CCNB2 and ASPM genes induces tumorigenic phenotypes in a number of cancers, whereas their inhibition abrogates cellular proliferation in mice and induces genomic instability [23, 38].
The CDCA7 gene has been implicated in neoplastic transformation and it is one of the downstream targets of the Myc oncogene . Interestingly, high nuclear expression of CDCA7 was seen in all the analyzed tumors in the present study. It is known that deregulation of cell cycle control is a fundamental feature of cancer pathogenesis, therefore it was not unexpected that CDCA7 protein has been observed to be expressed at high levels in almost all selected tumors. KIAA0101 is predominantly expressed in mitochondria and partially in nuclei, playing an essential role in the regulation of DNA repair, cell cycle progression, and cell proliferation . Moreover, the KIAA0101 gene is over-expressed in tumors of the esophagus , colon , lungs [41, 42], and breast . KIAA0101-positivity was observed in 79% of the immunostained tumors. Up-regulation of KIAA0101 was confirmed by real time qRT-PCR, which showed the over-expression of the gene at the mRNA level in 92% of the studied tumors. The possibility of tissue heterogeneity, accounting for the discordance between mRNA and protein expression cannot be excluded. In addition, there was no association between elevated CDCA7 and KIAA0101 protein levels and DSS or any other clinical parameters. The CDCA7, and KIAA0101 may therefore only have role in tumor initiation. Discordant results were detected between mRNA and protein expression of SLC27A2. High mRNA expression was detected in 83% of the analyzed tumors, but protein expression was only seen in 25%, possibly owing to posttranscriptional regulation and differences in mRNA, and protein turnover rates or poor specificity of the antibody used for IHC [44, 45]. These findings suggest that down-regulation of SLC27A2 in the selected tissues at the protein level may contribute to disease progression. Indeed, this gene was reported to regulate the tumor suppressor gene PARP and decreased SLC27A2 expression levels were found in the metastatic compared to the non-metastatic neuroendocrine tumors . However, no significant difference could be seen on the effect of SLC27A2 protein expression on DSS in breast cancer, nor could any association between the protein expression of SLC27A2 and the conventional clinical characteristics be observed.
To our knowledge, this is the first study in breast cancer patients reporting CCNB2 as a prognostic marker for unfavorable patient prognosis. Since several publications revealed that altered CCNB2 expression is seen in many cancer types, therefore further investigation to elucidate the mechanism by which CCNB2 exerts its effects may prove useful in the development of novel anticancer agents.