High expression of nucleobindin 2 is associated with poor prognosis in gastric cancer

Abstract

Nucleobindin 2 has been reported that its high expression is associated with poor outcome and promotes cell migration and lymph node metastasis in breast cancer, colon cancer, and prostate cancer. However, we aimed to investigate the nucleobindin 2 expression in gastric cancer tissues and adjacent non-tumor tissues and its potential relevance to clinicopathological factors and prognosis using immunohistochemical analysis. In our study, nucleobindin 2 level in gastric cancer tissues was higher than in non-tumor tissues. A high expression of nucleobindin 2 is significantly associated with tumor depth, lymph node metastasis, lymphatic invasion, venous invasion, and clinical stage. Furthermore, the expression level of nucleobindin 2 protein was independent predictor of progression-free survival. In summary, nucleobindin 2 might play a crucial role in gastric cancer development and could serve as an independent predictor of prognosis of gastric cancer patients.

Introduction

Gastric cancer (GC) is the fifth most common cancer worldwide, with nearly 1 million diagnoses annually. ¹ Treatment with aggressive and adjuvant chemotherapy in advanced GC has led to improved survival rates, but the prognosis for patients remains poor. ² The tumor–node–metastasis (TNM) staging system is the main therapeutic reference and guidelines in current clinical practice, and there is the limitation to TNM staging system to introduce the necessity to identify more prognostic factors to tailor the treatment of individual patients. Regimens of cytotoxic chemotherapy commonly used as first-line therapy for advanced GC result in median survival ranging from 8 to 10 months. ³ Because of limited second-line treatment options, further research is required to identify new therapeutic targets for improving treatment outcomes. Moreover, there is no established biomarker allowing prediction of prognosis and response after treatments such as surgery or chemotherapy. The identification of prognostic factors does not overly determine therapeutic strategy, rather predictive factors instead of prognostic factors are required to determine the impact of individual factors upon therapeutic success or lack thereof. Therefore, a novel predictive biomarker might assist in the prolongation of survival of patients with GC.

Nucleobindin 2 (NUCB2) is a 396-amino acid protein, cleaved into the N-terminal nesfatin-1_1-82, nesfatin-2_85-163, and the C-terminal nesfatin-3_166-396. ⁴ NUCB2 contains a signal peptide, a leucine zipper structure, two Ca²⁺ binding EF-hand domains,^5,6 and has a wide variety of basic cellular functions.^4,7,8 NUCB2 is also a precursor protein of nesfatic-1, which was originally identified in hypothalamic nuclei, and which is a regulatory factor involved in the central control of food intake and energy balance.^5,9 There are several reports indicating that NUCB2 is also expressed in various human peripheral tissues. ⁹ Moreover, recent studies have reported that high levels of NUCB2 messenger RNA (mRNA) and protein are a potent prognostic factor for prostate cancer, endometrial carcinoma, and breast cancer.^10–13 Suzuki et al. ¹⁰ described high NUCB2 immunoreactivity in patients with breast cancer, and NUCB2 protein expression was associated significantly with lymph node metastasis and an increased risk of recurrence. Their in vitro approach demonstrated that NUCB2 expression was upregulated by estrogen. ¹⁰ They also found that NUCB2 significantly increased tumor cell proliferation, migration, and invasion, suggesting that NUCB2 plays a crucial role in the metastatic process of breast cancer. ¹⁰ Several researchers have reported that the expression level of NUCB2 was significantly higher in prostate cancer tissues than in non-cancerous tissues and was associated with lymph node metastasis and shorter survival.^11–13 An experimental study using colon cancer cells demonstrated that the inhibition of NUCB2 expression suppressed migration and invasion, as well as the expression of epithelial–mesenchymal transition (EMT)-related molecules, including N-cadherin, E-cadherin, and β-cadherin. ¹⁴ However, the clinicopathological significance of NUCB2 expression in GC has still not been elucidated. Therefore, we examined NUCB2 expression in a large number of GC patients, using immunohistochemistry, to explore its clinicopathological significance.

Materials and methods

Patients and clinical samples

A total of 328 consecutive patients with GC, who underwent radical section at our institution between January 2000 and December 2009, were included in the study. The clinicopathological details were obtained from pathological reports and medical records and included age, sex, tumor location, histology, tumor depth, lymph node metastasis, lymphatic invasion, venous invasion, and clinical stage. This study was approved by the institutional review board of Gunma University Hospital (Ethical Committee for Clinical Studies, Gunma University Faculty of Medicine). The age of the patients ranged from 28 to 90 years, and the median age was 69 years. None of the patients had received neoadjuvant chemotherapy. All surgical specimens were reviewed and classified according to the World Health Organization (WHO) classification by an experienced pathologist, who was unaware of clinical or imaging findings. Pathological TNM stages were established using the International System for Staging adopted by the American Joint Committee on Cancer and the Union Internationale Contre le Cancer. Histologically, all patients had adenocarcinoma (AC) histology, and 181, 66, 63, and 18 of the total patients had stage I, II, III, and IV tumors, respectively. The day of surgery was taken as the starting date for measurement of postoperative survival. The follow-up duration ranged from 72 to 5430 days (median: 2047 days). The approach to the evaluation and resection of these tumors has been described previously. ¹⁵

Immunohistochemistry

Immunohistochemical staining was performed with 2-µm-thick sections. All sections were incubated at 60°C for 60 min, deparaffinized in xylene, and rehydrated. Sections were then incubated with fresh 0.3% hydrogen peroxide in 100% methanol for 30 min at room temperature to block endogenous peroxidase activity. After rehydrating through a graded series of ethanol solutions, antigen retrieval was carried out in Immunosaver (NJ15T; NEM, Tokyo, Japan) at 98°C–100°C for 30 min, and then the sections were passively cooled to room temperature. After rinsing in 0.1 M phosphate-buffered saline (PBS, pH 7.4), the sections were incubated in Protein Block Serum-Free Reagent (Dako, Carpinteria, CA, USA) for 30 min to block non-specific binding sites. The sections were then incubated overnight at 4°C with rabbit monoclonal antibody against NUCB2 (US Biological Life Sciences, Salem, MA), at a dilution of 1:500 in PBS containing 0.1% bovine serum albumin, and then incubated at room temperature for 30 min. The reaction was visualized using the Histofine Simple Stain MAX-PO (Multi) Kit (Nichirei, Tokyo, Japan) according to the manufacturer’s instructions. ¹⁶ The chromogen 3,3′-diaminobenzidine tetrahydrochloride was applied as a 0.02% solution in 50 mM ammonium acetate–citrate acid buffer (pH 6.0) containing 0.005% hydrogen peroxide. The sections were lightly counterstained with hematoxylin and mounted. Negative controls were incubated without a primary antibody, and no detectable staining was evident.

Evaluation of immunostaining

Immunohistochemical slides were scanned and evaluated by two experienced researchers. The expression of NUCB2 was considered as positive only if nuclear staining was present. The NUCB2 expression scores were assessed by the extent of staining as follows: score = 1: ≤10% of tumor area stained, score = 2: 11%–25% stained, score = 3: 26–50% stained, and score = 4: ≥51% stained. The tumors in which stained tumor cells were scored as 2 or greater were categorized as high expression. The evaluation of this scoring system was defined according to previous investigation. ¹⁵

Statistical analysis

Statistically significant differences were identified with Student’s t test and χ² test for continuous and categorical variables, respectively. Survival curves were generated by Kaplan–Meier analysis, and differences were examined by log-rank testing. Overall survival (OS) was determined as the time from tumor resection to death from any cause. Progression-free survival (PFS) was defined as the time between tumor resection and the first disease progression or death. In addition, univariate and multivariate survival analyses were performed using Cox proportional hazards model for survival and a logistic regression model for radical surgery. Probability values of p < 0.05 were considered as statistically significant. All statistical analyses were performed using JMP 9.0 software (SAS Institute Inc., Cary, NC, USA).

Search of NUCB2 expression in GC in The Cancer Genome Atlas dataset of online RNA sequencing database

We downloaded an online The Cancer Genome Atlas (TCGA) dataset to validate the relevance of NUCB2 mRNA expression to OS in patients with GC. Kaplan–Meier survival curves of patients with GC with high or low NUCB2 expression were analyzed in the whole 36 patients. The cutoff value of NUCB2 mRNA expression was defined as its average value.

Results

NUCB2 expression in clinical GC samples

We used immunohistochemistry to investigate the expression of NUCB2 in the 358 primary lesions obtained from all patients with GC. NUCB2 was expressed more strongly in tumor tissues than adjacent relative normal tissues, and localized predominantly in the nuclei (Figure 1). Of GC sections from 328 patients, high expression of NUCB2 was observed in 40% (133/328). The percentage of sample scoring 1, 2, 3, or 4 for NUCB2 expression was 31% (102/328), 28% (93/328), 31% (103/328), and 9% (30/328), respectively. Figure 2 shows representative immunohistochemical staining for the categories of NUCB2 expression scoring.

Figure 1.

(a) Representative image of NUCB2 expression in one specimen contains cancer cells and relative normal cells. (b and c) High power view of the NUCB2 expression in non-cancer region and that of the cancer region.

Figure 2.

Representative immunohistochemical images to evaluate the NUCB2 expression level as (a) 1+, (b) 2+, (c) 3+, and (d) 4+ with scale bars.

Correlation between NUCB2 expression and clinicopathological factors

The patient’s demographics in relation to expression of NUCB2 are demonstrated in Table 1. The expression of NUCB2 was significantly associated with tumor depth, lymph node metastasis, lymphatic invasion, venous invasion, and clinical stage (Table 1).

Table 1.

Relationship between NUCB2 and clinical factors.

Clinicopathological variables	Total GC cohort (n = 328)
	Low (n = 195)	High (n = 133)	p value
Age
years	64.4 ± 9.5	63.5 ± 10	0.7804
Gender
Male	134 (59.3%)	92 (40.7%)	0.9906
Female	61 (59.2%)	42 (40.8%)
Histology type
Well differentiated, moderately differentiated	113 (62.1%)	69 (37.9%)	0.2775
Poorly differentiated, signet cell carcinoma	82 (56.2%)	64 (43.8%)
Tumor depth
M, SM, MP	135 (64.9%)	73 (35.1%)	0.0081*
SS, SE, SI	60 (50.0%)	60 (50.0%)
Lymph node metastasis
Absent	141 (65.3%)	75 (34.7%)	0.0028*
Present	54 (48.2%)	58 (51.8%)
Lymphatic invasion
Absent	96 (67.1%)	47 (32.9%)	0.0127*
Present	99 (53.5%)	86 (46.5%)
Venous invasion
Absent	140 (69.0%)	63 (31.0%)	<0.0001*
Present	55 (44.0%)	70 (56.0%)
Clinical stage
I, II	159 (64.4%)	88 (35.6%)	0.0015*
III, IV	36 (44.4%)	45 (55.6%)

NUCB2: nucleobindin 2; GC: gastric cancer; M: mucosa; SM: submucosa; MP: muscularis propria; SS: subserosa; SE: serosal evident; SI: serosal invasion.

*

Significant difference p < 0.05.

Prognostic significance of NUCB2 expression in GC

The 5-year survival rates of OS and PFS for all patients were 75% and 74%, respectively. Of 328 patients, 97 had died after initial surgery. Univariate analysis showed that tumor depth, lymph node metastasis, lymphatic permeation, vascular invasion, disease stage, and NUCB2 expression had a significant relationship with an unfavorable PFS and OS (Table 2). Multivariate analysis confirmed that tumor depth, vascular invasion, disease stage, and NUCB2 expression were independent prognostic factors for predicting poor PFS, and tumor depth and disease stage were identified as significant variables to predict OS for GC patients. Figure 3 shows the Kaplan–Meier survival curve in patients with high and low expression of NUCB2.

Table 2.

Univariate and multivariate analyses of survival rate.

Clinicopathological variables	Progression-free survival in GC patients						Overall survival in GC patients
	Univariate analysis			Multivariate analysis			Univariate analysis			Multivariate analysis
	HR	95% CI	p value	HR	95% CI	p value	HR	95% CI	p value	HR	95% CI	p value
Age (years)
≤65 vs >65	1.27	0.83–1.99	0.265	–	–	–	0.55	0.35–0.83	0.068	–	–	–
Gender
Male vs female	1.15	0.67–1.91	0.582	–	–	–	1.00	0.58–1.63	0.9996	–	–	–
Histology type
Well, moderate vs poor	1.48	0.96–2.25	0.0707	–	–	–	0.69	0.47–1.04	0.0778	–	–	–
Tumor depth
SS vs SE, SI	7.28	4.54–11.96	<0.0001*	1.95	1.46–2.59	<0.0001*	5.39	3.54–8.36	<0.0001*	1.30	0.86–1.97	0.0121*
Lymph node metastasis
Absent vs present	3.31	2.17–5.10	<0.0001*	1.08	0.77–1.37	0.8763	0.40	0.26–0.60	<0.0001*	1.34	0.81–2.12	0.2508
Lymphatic invasion
Absent vs present	4.96	2.92–8.99	<0.0001*	0.92	0.67–1.24	0.588	0.22	0.13–0.37	<0.0001*	1.79	0.91–3.55	0.895
Venous invasion
Absent vs present	5.09	3.27–8.12	<0.0001*	1.47	1.08–2.01	0.0130*	0.26	0.17–0.39	<0.0001*	1.65	0.99–2.82	0.0533
Disease stage
I, II vs III, IV	0.19	0.12–0.28	<0.0001	2.15	1.22–3.92	0.0082*	0.18	0.12–0.28	<0.0001*	2.15	1.22–3.92	0.0082*
NUCB2 expression
Low vs high	1.65	1.08–2.52	0.0214*	1.28	1.01–1.61	0.0383*	0.61	0.41–0.92	0.0211*	1.3	0.8–1.97	0.2104

GC: gastric cancer; HR: hazard ratio; CI: confidence interval; NUCB2: nucleobindin 2.

*

Significant difference p < 0.05.

Figure 3.

Progression-free survival and overall survival curves of 328 gastric cancer patients divided into two groups according to NUCB2 expression levels. The significant differences were given as p = 0.0214 for progression-free survival and p = 0.0211 for overall survival.

We next analyzed the prognostic significance of NUCB2 expression with respect to disease stage and tumor differentiation. In relation to disease stage, no statistically significant difference in the PFS and OS was observed between high and low NUCB2 expression in early (stage I/II) (Figure A1 and A2, online only) or advanced disease stage (stage III/IV; Figure A3 and A4, online only). Moreover, the analysis with respect to tumor differentiation showed that the prognosis for patients with poor differentiation was significantly worse in the high NUCB2 expression population than in the low NUCB2 expression population (Figure B1 and B2, online only). However, there was no significant difference between high and low NUCB2 expression in those patients with well-differentiated or moderately differentiated tumors (Figure B3 and B4, online only).

TCGA database search for NUCB2 expression in GC

Using Illumina GA in TCGA online database, we corrected 36 samples with GC and analyzed the clinicopathological factors and outcome of patients according to NUCB2 mRNA expression. Although this sample size is low, our analysis revealed that NUCB2 mRNA expression tended to be closely correlated with poor OS in GC patients (p = 0.0637; Figure 4). The patient’s clinicopathological details, in relation to NUCB2 mRNA expression, are listed in Table A1 (online only). High expression of NUCB2 mRNA was significantly associated with lymph node count (p = 0.0419). A correlation analysis between expression of NUCB2 and EMT-regulated genes showed a statistically significant correlation between NUCB2 and EMT markers (DSP, ITGAV, MMP3, TSPAN13, and CDH2; Tables A2, A3, and A4, online only).

Figure 4.

Kaplan–Meier curves of overall survival according to the expression levels of NUCB2 mRNA. NUCB2 mRNA expression tended to be closely correlated with poor OS in GC patients (p = 0.0637).

Discussion

We evaluated the prognostic significance of NUCB2 expression in GC using immunohistochemistry. Our results showed that NUCB2 expression in cancer tissues was higher than in non-cancerous tissues. Moreover, the high expression of NUCB2 in GC samples was significantly associated with tumor depth, lymph node metastasis, lymphatic invasion, venous invasion, and disease stage. We also found that NUCB2 protein expression was an independent prognostic factor for predicting unfavorable outcomes in patients with GC. In particular, a high expression level of NUCB2 was shown to be significantly associated with an unfavorable outcome in patients with poorly differentiated histology. To our knowledge, this is the first study to investigate the clinical significance of NUCB2 expression in GC.

Recently, it has been reported that NUCB2 mRNA levels were significantly higher in prostate cancer tissues than in adjacent non-cancerous tissues, ¹¹ and there was also a correlation with lower survival rates in several human neoplasms.^10,12,17 Our analysis of the online RNA sequencing TCGA dataset indicated that NUCB2 expression was significantly associated with lymph node count. More interestingly, a significant correlation between the level of NUCB2 mRNA and some EMT-regulated genes was observed using the TCGA online database. Kan et al. ¹⁴ reported that NUCB2 may enhance migration and invasion, upregulate EMT pathways in colon cancer, and promote tumor EMT via autocrine and paracrine pathways. In addition, they also study found that the mechanistic target of rapamycin (mTOR) pathway is essential in NUCB2-mediated migration, invasion, and the EMT pathway. ¹⁴ Some studies found that the mTOR pathway regulates cellular processes, such as cell proliferation and EMT, in cancer.^18,19 Our TCGA data analysis also showed a close correlation between NUCB2 mRNA expression and some EMT-regulated genes. This evidence supports the idea that the expression of NUCB2 may be linked to the EMT and mTOR pathways. However, the relationship between NUCB2 expression and EMT or mTOR pathways in GC patients remains unclear. Further investigations should focus upon the clinicopathological significance of NUCB2 expression in EMT and/or mTOR pathways. As the action of mechanism by NUCB2, recent study also has described that Calnuc/NUCB1 and NUCB2 serve as guanine nucleotide exchange factors (GEFs) for the inhibitory G-protein α-subunit, Giα, via an evolutionarily conserved motif. ¹⁷ Our survival analysis, based on the TCGA database, showed that NUCB2 mRNA expression was associated with poorer survival, but the difference was not statistically significant. Although a small sample size may bias the results of survival analysis, we nevertheless found prognostic significance for NUCB2 mRNA levels. The results of this independent cohort were consistent with other findings in this study; therefore, it seems that NUCB2 might represent a molecular prognostic marker for GC patients, identifying those who are more likely to have higher risk, and thus might be candidates for a more aggressive treatment.

In this study, we found that the expression of NUCB2 may play a crucial role in survival and tumor progression in the patients with poor tumor differentiation, compared to those with well-differentiated or moderately differentiated tumors. Although there was no statistically significant difference in the survival analysis according to disease stage (Tables A1, A2, A3, and A4, online only), GC patients with advanced disease exhibited a trend toward a poorer prognosis when NUCB2 expression was high. In all patients, high expression of NUCB2 was closely linked to progression and metastasis, suggesting potential use of NUCB2 as a biomarker. Our data showed that NUCB2 is useful to predict the outcome of patients with poorly differentiated histology after surgery. NUCB2 yielded a significant association with metastases, advanced disease, and poorly differentiated histology, thus, NUCB2 seemed to play an essential role in the malignant formation and tumor progression, although little is known why NUCB2 is highly expressed in poorly differentiated histology. We believe that the inhibition of NUCB2 may contribute to the suppression of GC cells in cases of advanced disease and poorly differentiated histology. Nowadays, systemic chemotherapy is suitable for GC patients with advanced disease, but the efficacy of chemotherapy is considered to be limited. The discovery of novel molecular targets is needed to facilitate improved prognoses of GC patients, especially in cases with advanced disease or poorly differentiated histology. There are no effective chemotherapeutic agents for GC patients with poorly differentiated histology, resulting in a dismal outcome, regardless of treatment. Further study is warranted to investigate the therapeutic significance of the suppression of NUCB2 within GC cells.

There are several limitations in our study. One limitation is that we could not analyze the mRNA expression of NUCB2 in the samples used. Since it remains unclear whether there is a close relationship between NUCB2 protein and mRNA expression levels in patients with GC, further study is warranted to investigate the clinicopathological significance of NUCB2 mRNA expression. Nevertheless, our survival analysis based on the TCGA database was consistent with other results in this study. However, the small sample size obtained from TCGA database may bias the results of the prognostic significance of NUCB2 mRNA expression. Another limitation is that several markers related to EMT were not actually assessed in our study. Since the role of NUCB2 expression within GC cells is unclear from the perspective of cancer biology, we tried to identify the clinical significance of NUCB2 as a single marker in our study. Further study is warranted to elucidate the relationship between NUCB2 expression and associated biomarkers. Finally, a previous study described that NUCB2 is ubiquitously expressed in normal tissues and downregulated in the gastric tumors compared to the adjacent relatively normal gastric tissues. ²⁰ Our investigation revealed the predominant positive staining of NUCB2 expression in the gastric tumor tissues. Although the antibody of NUCB2 is different from that of Kalnina’s study, the technical methods including immunohistochemistry may contribute to the opposite findings. ²⁰

In conclusion, our results provide the first convincing evidence that NUCB2 protein is upregulated in GC patients, especially in cases with poorly differentiated histology. Our study revealed that NUCB2 is an independent prognostic factor for PFS and OS of GC patients. The statistical analysis of clinical factors and NUCB2 expression suggest that NUCB2 is related to cancer malignancy. Our findings suggest that NUCB2 might be used as a new biomarker and a potential therapeutic target for GC.