The expression and clinical significance of GADD45A in breast cancer patients

Background Growth arrest and DNA-damage-inducible protein 45 alpha (GADD45A) was previously found to be associated with risk of several kinds of human tumors. Here, we studied the expression and clinical significance of GADD45A in breast cancer. Methods We performed an immunohistochemical study of GADD45A protein from 419 breast cancer tissues and 116 adjacent non-neoplastic tissues. Results Significantly high GADD45A expression were observed in breast cancer tissues compared with adjacent non-neoplastic tissues (P < 0.001) and were independently correlative with estrogen receptor negative (P = 0.028) and high Ki-67 index (P < 0.001). Kaplan–Meier survival analysis revealed that patients with high GADD45A expression levels had a worse long-term prognosis in triple negative breast cancer (P = 0.041), but it was not an independent prognostic factor in multivariate analysis (P = 0.058). Conclusions GADD45A expression levels are significantly correlative with estrogen receptor status and Ki-67 index in human breast cancer. Patients with triple negative breast cancer might be stratified into high risk and low risk groups based on the GADD45A expression levels.

two of the most important genes maintaining the stability of the genome and suppressing the development of tumor through various mechanisms (Li et al., 2017b). As a member of the GADD45 family of genes that are known as stress sensors, GADD45A modulates the cellular response to a variety of stress conditions, including genotoxic and oncogenic stress (Cretu et al., 2009). Also, some research showed GADD45A can promote or suppress breast carcinoma development depending on different signaling pathways (Carrier et al., 1999;Miki et al., 1994).
Additionally, the correlation of GADD45A expression and clinicopathologic factors is not clear in breast cancer thus far. The aim of the current study is to investigate the significance of GADD45A in breast cancer, including the correlation with clinicopathologic factors and long-term survival situation.

Tissue samples and clinical data
The study retrospectively collected 419 cases of tissue specimens from breast cancer patients, undergoing the operation from February 2005 to October 2007, at Changhai Hospital, Shanghai, China. Patients with other primary tumor sites were excluded. Clinicopathologic data was recorded from the patients' medical records including age at diagnosis, molecular classification (ST Gallen), TNM stage (AJCC), pathology types, histological grade (Elston-Ellis grade) and clinical outcomes. Cutoff for positivity was set at ≥1% of cells staining positively for ER or PR. HER-2 status was assessed using the HercepTest TM according to the manufacturer's instructions. HER-2 positivity was defined as a 3+ score on IHC in >10% of invasive tumor cells. Equivocal IHC cases (2+ score or 3+ score in ≤10% of invasive tumor cells) were submitted to FISH analysis. A ratio of HER-2 signals to chromosome 17 signals >2.0 was used as the cutoff to define HER-2 amplification. Ki-67 index greater than 20% was classified as high. In short, Luminal A was defined as ER+ or PR+, HER2and low Ki67, Luminal B as ER+ or PR+ and HER2+ or high Ki67; HER2 positive as HER2+ and ER-and PR-; basal-like as HER2-/ER-/PR-. In addition, 116 cases of adjacent non-neoplastic tissue samples were also collected from mammoplasties as a control group.
GADD45A color reaction/positive staining were observed in 419 tumor tissues and 116 adjacent non-neoplastic tissues. The study was approved by the Ethics Committee of Biomedicine, Navy Medical University. The samples were used after obtaining informed written consents from all included patients.

Tissue microarray and immunohistochemistry
The breast cancer tissue microarray (TMA) was obtained from the surgically resected tissues. Biopsies were fixed in 10% neutral formalin. TMA blocks were made by tissue arraying instrument (Beecher Instruments, Sun Prairie, WI, USA). Cylinders cores (1.5 mm in diameter) of tissue blocks were extracted from the center of the tumor, and then arranged into blank recipient paraffin blocks. The TMA blocks were cut into 4-µm sections and treated for immunohistochemistry (IHC). Following deparaffinization and rehydration of the tissues sections, antigen retrieval was performed in the pressure cooker in citrate buffer, pH 6.0, for 30 min at a sub-boiling temperature. Endogenous peroxidase was blocked with 3% peroxide for 10 min. Primary GADD45A antibody (1:200; catno. Sc-792; Santa Cruz Biotechnology, Dallas, TX, USA) were used on tissue sections for immunohistochemical staining. For a negative control, we replaced the primary antibody rabbit immunoglobulin G or normal murine with the same dilution.

Statistical analyses
Statistical analyses were conducted using the IBM SPSS Statistics (for Windows, version 21.0). The Pearson χ2 test or Fisher's exact test was used to compare the GADD45A expressions between breast cancer tissues and adjacent non-neoplastic tissues and to confirm the correlations between clinicopathological factors. In the combined analyses, all the datasets were pooled and the odds ratios and P-values were estimated with logistic regression model. To assess the effect of GADD45A expression and other factors on overall survival (OS), different Kaplan-Meier curves were plotted with log-rank test. A multivariate Cox regression model was fitted to test the independent contribution of each variable to the patient's OS. For all data, differences were considered statistically significant when P < 0.05. Separate analysis was performed for the subtypes of the breast cancer.

Expression of GADD45A in breast carcinoma and normal tissues
The cytoplasmic staining for GADD45A was identified in the breast carcinoma tissues and adjacent non-neoplastic tissues. Overall, high expression of GADD45A was found in 72.3% (303/419) cancer tissues and 37.9% (44/116) adjacent noncancerous tissues. The percentage of high GADD45A expression of breast carcinoma tissues was obviously higher than that of adjacent noncancerous tissues and the difference was statistically significant (P < 0.001, Table 1).

Correlation between GADD45A expression and clinical pathological characteristics
The clinical pathological factors used in the current study include the following: age at diagnosis, TNM stage, pathology types, histological grade, molecular subtypes, ER (estrogen receptor), PR (progesterone receptor), HER2 (human epidermal growth factor receptor-2) and Ki-67 index. We found a significant correlation between GADD45A expression and ER status, Ki-67 index, histological grade and molecular subtypes but not between GADD45A expression and age at TNM stage, pathology types, PR and HER2. The high expression rate of GADD45A was 78.0% (131/168) in ER negative group which is 68.5% (172/251) in ER positive group and the difference was statistically significant (P = 0.034, Table 2). The high GADD45A expression rate of Ki-67 index high-level group (≥20%) (83.1%, 207/249) was apparently higher than that of Ki-67 index low-level group (<20%) (56.2%, 96/170), (P < 0.001). Meanwhile, a positive correlation was found between GADD45A expression and histological grade (P = 0.013). It was also found between GADD45A expression and molecular subtypes (P < 0.001). ER status, Ki-67 index and histological grade were selected into logistic regression model analysis (Molecular subtype was not included because its effect was similar with ER status). The results from multivariate analysis confirmed that ER negative and Ki-67 ≥20% were closely related to high GADD45A expression (Table 3).

Survival analysis
Kaplan-Meier curves and log-rank tests were used to perform survival analysis to identify the correlation between survival and GADD45A expression. The last follow-up was carried out in July 2017. Table 4 showed GADD45A expression was not related to survival when all  Fig. 2.
In the univariate analysis, the TNM stage and GADD45A were found to be significant independent predictors of OS, respectively (Table 5). However, in the multivariate analysis, only the TNM stage was identified the significant independent predictor of OS in the TN subtype patients (Table 5).

DISCUSSION
As GADD45A plays a role in genomic stability and tumorigenesis, some studies explored the role of GADD45A in various cancers, such as prostate cancer (Reis et al., 2015), ovarian cancer (Yuan et al., 2015), esophageal cancer (Ishiguro et al., 2016) and malignant gliomas (Cui et al., 2017), which revealed different GADD45A expressions in various carcinoma tissues. In research on esophageal squamous cell carcinoma (ESCC) (Wang et al., 2012) and another article on breast cancer (Tront et al., 2013), GADD45A protein levels were increased significantly in tumor tissues than those in normal tissues. In addition, a recent study showed that the expression of GADD45A was upregulated in human non-small-cell lung cancer (NSCLC) cells through a microRNA-138 related post-transcriptional pathway (Li et al., 2017a). Compared with the previous study for GADD45A in breast cancer (Tront et al., 2013), our study had a larger sample analysis for the correlations between clinicopathological factors and GADD45A expression. We also showed the correlation with PR and ER, respectively, while Tront's study only analyzed tissues of ER+/PR+. Moreover, our study was the first study estimating the prognostic significance of GADD45A in triple negative breast cancer. We found that GADD45A expression was upregulated in breast cancer tissues. DNA damage was a major source of potentially mutagenic lesions leading carcinogenesis (Gao, Herman & Guo, 2016). GADD45A was shown to participate actively in DNA repair mechanisms (Wingert & Rieger, 2016). Therefore, its upregulation may be due to extensive DNA repair in breast cancer tissues. Beyond that, a study showed that GADD45A promotes Myc-driven breast carcinogenesis by negatively regulating matrix metalloproteinase 10 (MMP10) through GSK-3β (Glycogen synthase kinase 3 beta)/βcatenin signaling, resulting in increased tumor vascularization and growth (Tront et al., 2010). In our opinions, this mechanism may result for the high GADD45A expression in breast cancer tissues.
In our study, ER negative and high Ki-67 index were demonstrated the significant association with GADD45A expression. ER-negative breast cancer, which had less treatment options such estrogen-blocking endocrine therapy, had more malignancy behaviors than ER-positive tumors. Ki-67 index was a cellular marker strictly associated with cell proliferation and emerged as a key discriminative feature of malignant phenotype in breast cancer (Pathmanathan et al., 2014). The above factors all suggested high malignant degree.
From this, we considered that GADD45A may be a biomarker for the malignancy grade of breast cancer tissue. Our results were opposite to the results presented by Tront et al. (2013) which showed that high GADD45A levels were associated with luminal A and luminal B subtypes but not in triple negative tissues. The discrepancy could be generated by the higher number of tissues analyzed in our study compared to the previous one (419 vs 45). In addition, we explored the correlation with ER and PR status separately while (Tront et al., 2013) analyzed only tissue where they were both positive, as mentioned above. As a result of these advantages, our study was a more credible and detailed exploration for the correlations between clinicopathological factors and GADD45A expression compared previous study.
For further evaluation, we assessed the effect of GADD45A expression on OS. We demonstrated that patients with higher GADD45A expression levels had a poor long-term prognosis for TN subtype based on Kaplan-Meier survival curves. However, GADD45A expression was not an independent prognosis factor in TN type breast cancer in multivariate analysis. The finding may suggest that GADD45A expression affect the prognosis through the effects of other factors but not the independent action, such as the effect of Ki-67 index, which was reported as an independent variable associated with prognosis in TN subtype patients (Adamo et al., 2017;Pan et al., 2017;Ricciardi et al., 2015). In addition, we found that TNM stage independently influenced the prognosis of TN breast cancer patients. The similar conclusion was determined in previous studies (Han et al., 2017;Yu et al., 2013;Yu et al., 2014).
The limitations of current study included the deficiency of follow-up in chemotherapy regimens of patients. It was difficult to analyze the influence of chemotherapy regimen on prognosis of breast cancer. The relationship between different chemotherapy treatments and prognosis in breast cancer should be the focus of the future study and further analyses in biological mechanisms about tumor progression are need. Moreover, the possible selective bias due to the small sample size and inclusion of only postoperative patients may exist in TN breast cancer group. Despite the limitations, the correlation between GADD45A expression and certain clinicopathological factors in breast cancer can be verified.

CONCLUSIONS
In present study, we explored the expression and clinical significance of GADD45A in breast carcinoma. High GADD45A was observed in breast cancer tissues compared with adjacent non-neoplastic tissues. ER negative and high Ki-67 index were correlative with high GADD45A expression. Patients with higher GADD45A expression levels had a poorer long-term prognosis in TN type breast cancer. However, only the TNM stage was the independent prognosis factor in TN type breast cancer. Our study suggested that GADD45A may play a role in breast cancer pathogenesis and may eventually help in understanding the biological mechanisms affecting tumor progression.

ADDITIONAL INFORMATION AND DECLARATIONS Funding
Funding for this study was provided by the National Natural Science Foundation of China [grant number 81572591]. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Grant Disclosures
The following grant information was disclosed by the authors: National Natural Science Foundation of China: 81572591.