High expression of NDRG3 associates with unfavorable overall survival in non-small cell lung cancer

Abstract

BACKGROUND AND OBJECTIVE:

N-myc downstream-regulated gene 3 (NDRG3) is one of the important members of the NDRG family which crucially take part in cell proliferation, differentiation and other biological processes.

METHODS:

In this present study, western-blotting analysis was performed to evaluate NDRG3 expression in NSCLC cell lines. One-step quantitative reverse transcription-polymerase chain reaction (qPCR) with 16 fresh-frozen NSCLC samples and immunohistochemistry (IHC) analysis in 100 NSCLC cases were conducted to explore the relationship between NDRG3 expression and the clinicopathological characteristics of NSCLC.

RESULTS:

NDRG3 expression levels were statistically higher in NSCLC cell lines and tissue samples, compared with that of in non-cancerous cell line and tissue samples ( 0.05). The IHC data demonstrated that the NDRG3 expression was significantly correlated with pathological grade ($p=$ 0.038), N ($p=$ 0.020) and TNM stage ($p=$ 0.002). Survival analysis and Kaplan-Meier curve indicated that NDRG3 expression ($p=$ 0.002) and T ($p=$ 0.047) were independently associated with the unfavorable overall survival of patients with NSCLC.

CONCLUSIONS:

The data implied that NDRG3 expression may be identified as a new predictor in NSCLC prognosis.

1. Introduction

Lung cancer (LC) is one of the most common cancer and remains the leading cause of cancer-related death worldwide [1, 2]. In USA, for instance, over 220,000 new cases of LC were diagnosed and approximately 160,000 patients were dead in 2014 [3, 4]. In China, the situation of LC is severely disturbing because the mortality of LC has been increasing by more than 400% over the past three decades [5, 6]. LC principally consists of small-cell (SC) and non-small-cell lung cancer (NSCLC), while the latter accounts for 80–85% of all LC cases. Histologically, the most common subtypes of NSCLC are squamous cell carcinoma (SCC) and adenocarcinoma (ADC) [7]. Currently, surgery is the most promising and radical therapeutic strategy for NSCLC treatment. However, 30% NSCLC patients with early stages encounter relapse after tumor resection, let alone 80% patients are diagnosed in advanced stages [8]. In spite of substantial advances of diagnostic and therapeutic methods on NSCLC in recent years, the prognosis of NSCLC is still disappointing, with 5-year survival rate remains about 17% [9]. To date, a number of studies have explored the important molecular alterations in NSCLC and provided important biomarkers that stratify patients according to the outcome [10, 11]. There continues an urgent need for the identification of novel molecular biomarkers that indicate the prognostic status of NSCLC patients and provide potential treatment options [12, 13].

N-myc downstream-regulated gene 3 (NDRG3) is one of the principal members of the NDRG family which dramatically contribute to the cell proliferation, differentiation and other biological processes [14, 15]. NDRG3 is highly expressed in many organs, including testis, ovary, prostate, and thymus rudiment. It may play a role in spermatogenesis as it was detected in the outer layers of the seminiferous epithelium [16]. Lately, several attentions have been drawn to the complicated roles of NDRG3 in cancer development. NDRG3 was found to act as a tumor promoter and noticeably associated with the malignant phenotype of prostate cancer (PCa) [17, 18]. NDRG3 could upregulate the expression of angiogenic chemokines (CXCL1, CXCL3 and CXCL5), enhance the angiogenesis and finally promote the tumor growth [17]. Fan et al. reported that NDRG3 was overexpressed in hepatocellular carcinoma (HCC) and NDRG3 represented a key diagnostic and therapeutic target for HBV-related HCC [19]. Ren et al. similarly stated that the aberrant expression of NDRG3 facilitated the tumor progression of PCa and the upregulation of NDRG3 may be an efficient indicator for PCa prognosis [20]. Because of the accumulation of evidence for the diverse functions of NDRG3 in tumorigenesis, the expanding research of NDRG3 activity in human cancer is of substantial significance.

In the present study, we detected NDRG3 expression by western-blotting analysis (NSCLC cell lines), one step quantitative reverse transcription-polymerase chain reaction (qPCR) test and immunohistochemistry (IHC) analysis (NSCLC clinical samples). Furthermore, we investigated the associations of NDRG3 expression with the clinicopathological parameters of NSCLC patients. Finally, the prognostic role of NDRG3 in NSCLC was illustrated.

2. Methods

2.1. Cell lines

Three NSCLC cell lines (SK-MES-1, A549 and H1299), and human bronchial epithelial cell line (16HBE) were obtained from the cell bank of the Chinese Academy of Science (Shanghai, China) and cultured routinely by our laboratory. SK-MES-1 is a lung squamous cell carcinoma cell line and A549 and H1299 are two lung adenocarcinoma cell lines. 16HBE is a non-cancerous cell line.

2.2. NSCLC patients and samples collection

Sixteen fresh frozen NSCLC samples and corresponding non-cancerous tissue samples were obtained from The First People’s Hospital of Kunshan Affiliated with Jiangsu University to perform one-step qPCR test. Simultaneously, 100 cases of NSCLC (50 SCC and 50 ADC) were enrolled to construct tissue microarrays (TMA) and the TMA were purchased from Outdo Biotech Co., Ltd (Shanghai, China) to execute immunohistochemistry (IHC) analysis. The important clinical parameters of each patient, including gender, age, tumor size, pathological grade, histological type, tumor status (T), lymph node metastasis (N), distant metastasis (M) and TNM stage, were also acquired from the TMA data that provided by Outdo Biotech Co., Ltd. None of the NSCLC patient received any forms of treatments (radiation therapy, chemotherapy, or immunotherapy) before surgery. Written informed consent was obtained from each patient included in this study. Ethical approval to perform this research was approved from the Human Research Ethics Committee of Nanjing Medical University and each local hospital.

2.3. Western-blotting analysis

Western-blotting analysis was performed to detect NDRG3 expression in NSCLC cell lines. Cells were washed and lysed with cell lysis buffer. Equal amounts of proteins were separated by 10% SDS-PAGE and transferred onto nitrocellulose membranes. The membranes were first incubated with a primary rabbit monoclonal anti-NDRG3 antibody (1:2000, ab133715, Abcam, Cambridge, MA, USA) and then a secondary antibody (Dako Cytomation, Carpinteria, CA), and finally detected with an ECL kit and autoradiography using X-ray film. $\beta$-actin was used as an internal control.

2.4. One-step qPCR test in fresh NSCLC samples

Sixteen fresh-frozen NSCLC samples as well as the corresponding non-cancerous tissue samples were collected. Total RNA was extracted from the fresh tissues using Trizol (Invitrogen, USA) following the manufacturer’s manual. The protocol of qPCR analysis was described previously [21]. The primers for NDRG3 were as follows: forward primer 5’-CCA GGA CTT TGA CTG TCA GGA-3’ and reverse primer 5’-AGT GCT GGG TGA TCT CTT GC-3’. The $\beta$-actin was employed as internal control, and the primers for $\beta$-actin were as follows: forward primer 5’-TTA ATC TTC GCC TTA ATA CTT-3’ and reverse primer 5’-AGC CTT CAT ACA TCT CAA-3’. The PCR conditions were as follows: denaturation at 95${}^{\circ }$C for 20 s, annealing at 56${}^{\circ }$C for 30 s and extension at 72${}^{\circ }$C for 30 s. Expression data were normalized to the geometric mean of the $\beta$-actin housekeeping gene and calculated by using the comparative Ct (2${}^{-\mathrm{\Delta }\mathrm{\Delta }\text{Ct}}$) method as previously described [22].

2.5. IHC analysis in NSCLC TMA

IHC analysis was performed as previously desc- ribed [23, 24]. TMA sections were incubated with rabbit monoclonal anti-NDRG3 antibody (1:200, ab133715, Abcam, Cambridge, MA, USA), rabbit monoclonal anti-EGFR antibody (1:150, ab52894, Abcam), rabbit monoclonal anti-ALK antibody (1:200, ab16670, Abcam) and rabbit monoclonal anti-PD-L1 antibody (1:200, ab205921, Abcam). The secondary antibody used was horseradish peroxidase-conjugated anti-rabbit antibody (Dako). Phosphate-buffered saline (PBS) was used as negative control. NDRG3 immunostaining was evaluated according to the intensity and percentage of NDRG3-positive cells. Staining intensity was marked as follows: 0 (negative staining), 1 (yellow staining), 2 (light brown staining), and 3 (dark brown staining). Staining percentage of NDRG3 was categorized as follows: 1 (0–10%), 2 (11–50%), 3 (51–80%), and 4 (81–100%). The product of the staining intensity and percentage gave rise to the IHC score [25]. The final results of IHC analysis were defined using a two level system following the IHC score: 4 indicates low expression, while 4–12 indicates high expression.

2.6. Statistical analysis

The results of qPCR test and IHC analysis concerning NDRG3 expression were analyzed with the Wilcoxon nonparametric signed-rank test. The associations between NDRG3 expression and clinicopathologic characteristics were calculated by chi-square tests. Univariate and multivariate Cox regression models was performed to identify prognostic factors that affected the overall survival. Survival curves were drawn using the Kaplan-Meier method. For all tests, a two-tailed p value of less than 0.05 was considered statistically significant. All statistical analyses were performed by utilizing STATA 14.0 (Stata Corporation, College Station, TX, USA) and SPSS 18.0 (SPSS Inc, Chicago, IL, USA).

3. Results

3.1. Clinical characteristics of 100 NSCLC patients

A patient sample of 75 males and 25 females, with a median age of 62.38 years (range 37–84 years), was collected in this study. There were 78 cases with tumor diameter $⩾$ 5 cm, 21 with tumor diameter 5 cm. In terms of the pathological grade of NSCLC classification, 15 patients were in grade 1, 61 patients were in grade 2 and 24 patients were in grade 3. The histologic type of 50 patients was SCC while the other 50 was ADC. The number of patients with T1, T2, T3, T4 were 25, 54, 17 and 3 respectively. Positive N was witnessed in 45 patients, while positive M was only detected in 1 patient. Based on TNM staging system, 58 patients were in stages I–II, whereas 31 patients were in stages III–IV (Table 1).

Table 1.

Correlation of high NDRG3 protein expression with clinicopathological characteristics in NSCLC

Groups	No.	NDRG3		${\chi }^2$	p value
		+	%
Gender
Male	75	36	48.0	1.091	0.296
Female	25	9	36.0
Age (years)
$⩾$ 60	65	30	46.2	0.005	0.947
60	32	15	46.9
Insufficient data	3	0
Tumor size (cm)
$⩾$ 3	78	38	48.7	1.580	0.209
3	21	7	33.3
Insufficient data	1	0
Pathological grade
Grade 1	15	7	46.7	6.536	0.038${}^{*}$
Grade 2	61	22	36.1
Grade 3	24	16	66.7
Histologic type
Squamous cell carcinoma	50	26	52.0	1.980	0.159
Adenocarcinoma	50	19	38.0
EGFR expression
Positive	57	29	50.9	1.850	0.174
Negative	43	16	37.2
ALK expression
Positive	9	3	33.3	0.544	0.461
Negative	91	42	46.2
PD-L1 expression
Positive	45	24	53.3	2.296	0.130
Negative	55	21	38.2
T
T1	25	9	36.0	1.846	0.605
T2	54	25	46.3
T3	17	9	52.9
T4	3	2	66.7
Insufficient data	1	0
N
Positive	45	26	57.8	5.405	0.020${}^{*}$
Negative	50	17	34.0
Insufficient data	5	2
M
Positive	1	1	100.0	1.235	0.267
Negative	99	44	44.4
TNM stage
Stage I-II	58	19	32.8	9.992	0.002${}^{*}$
Stage III-IV	31	21	67.7
Insufficient data	11	5

${}^{*}$ 0.05; SCC: Squamous cell carcinoma; ADC: Adenocarcinoma.

3.2. Detection of NDRG3 expression in NSCLC cell lines

As shown in Fig. 1, the result of western-blotting analysis revealed that NDRG3 expression was significantly elevated in NSCLC cell lines relative to that of the non-cancerous cell line.

Figure 1.

Western-blotting analysis was performed to examine NDRG3 expression in three non-small-cell lung cancer (NSCLC) cell lines and one non-cancerous cell line. NDRG3 protein expression in three NSCLC cell lines (SK-MES-1, A549 and H1299) are elevated relative to the non-cancerous cell line (16HBE). SK-MES-1 is a lung squamous cell carcinoma cell line and A549 and H1299 are two lung adenocarcinoma cell lines. 16HBE is a non-cancerous human bronchial epithelial cell line.

3.3. Detection of NDRG3 mRNA expression by qPCR test

Total RNA was extracted from 16 NSCLC tissues as well as the corresponding non-cancerous tissues, and then the one-step qPCR was performed to evaluate NDRG3 mRNA expression. When normalizing to $\beta$-actin, the means of NDRG3 mRNA in NSCLC tissues and in corresponding non-cancerous tissues were 5.18 $\pm$ 0.392 and 3.54 $\pm$ 0.283, respectively (t $=$ 3.39, df $=$ 28, $p=$ 0.002). NDRG3 expression in NSCLC tissues was nearly 1.5-fold higher than that in corresponding non-cancerous tissues (Fig. 2).

Figure 2.

One-step quantitative real-time polymerase chain reaction (qPCR) test was performed to detect the mRNA expression of NDRG3 in non-small-cell lung cancer (NSCLC) tissue samples and corresponding non-cancerous tissue samples. When normalized to $\beta$-actin, the expression of NDRG3 mRNA in NSCLC tissues (5.18 $\pm$ 0.392) was significantly higher than that in matched non-cancerous tissues (3.54 $\pm$ 0.283) (t $=$ 3.39, df $=$ 28, $p=$ 0.002).

3.4. Detection of NDRG3 protein expression by IHC analysis

High NDRG3 expression was detected in 45 (45.0%) of the 100 NSCLC tissue samples compared with 25 (25.0%) of corresponding non-cancerous tissue samples. The difference showed remarkable significance (${\chi }^2$$=$ 8.79, $p=$ 0.003). The results of IHC analysis agreed with the previous data of qPCR test which showed high level of NDRG3 mRNA expression. High expression of NDRG3 protein was mainly localized in the cytoplasm and nucleus of NSCLC cells (Fig. 3).

Figure 3.

Representative types of NDRG3 protein expression in non-small-cell lung cancer (NSCLC) tissue samples and corresponding non-cancerous tissue samples. A1, A2 and A3 High cytoplasmic expression of NDRG3 in squamous cell carcinoma (SCC) tissue sample. Red arrows show the positive staining in the cytoplasm of cancer cells. B1, B2 and B3 High nuclear expression of NDRG3 in adenocarcinoma (ADC) tissue sample. Green arrows show the positive staining in the nucleus of cancer cells. C1, C2 and C3 Low expression of NDRG3 in non-cancerous tissue sample. Original magnification: $\times$ 40 in A1-C1; $\times$ 200 in A2-C2; $\times$ 400 in A3-C3.

3.5. Detection of EGFR/ALK/PD-L1 expression by IHC analyses

As the significant markers for NSCLS, we detected EGFR/ALK/PD-L1 expression by IHC analyses. The results showed that the positive expression of EGFR/ALK/PD-L1 were 57, 9, 45 of 100 NSCLC cases, respectively (Table 1).

3.6. Correlation of NDRG3 expression with the clinicopathological items

To investigate the clinical significance of NDRG3 in NSCLC, the association between NDRG3 protein expression and clinicopathological items was examined. As is shown in Table 1, positive NDRG3 expression was significantly related to pathological grade ($p=$ 0.038), N ($p=$ 0.020) and TNM stage ($p=$ 0.002). In comparison, other clinical items, including gender, age, tumor size, histological type, EGFR/ALK/PD-L1 expression, T and M showed no statistical association with NDRG3 protein expression (Table 1).

3.7. Survival analysis

The univariate analysis demonstrated three factors that associated with overall survival of NSCLC, including NDRG3 expression ($p=$ 0.001), T ($p=$ 0.003) and TNM stage ($p=$ 0.006). Multivariate analysis further confirmed that NDRG3 expression ($p=$ 0.002) and T ($p=$ 0.047) were two independent prognostic predictors for 100 NSCLC patients in this present study (Table 2). Kaplan-Meier survival curves were constructed to indicate that NSCLC patients with high NDRG3 expression and advanced T suffered unfavorable survival time (Fig. 4).

Table 2.

Univariate and multivariate analysis of prognostic factors for overall survival in NSCLC patients

	Univariate analysis			Multivariate analysis
	HR	p value	95% CI	HR	p value	95% CI
NDRG3 expression
High versus low	2.84	0.001${}^{*}$	1.430–5.242	2.74	0.002${}^{*}$	1.430–5.242
Gender
Male versus female	1.41	0.353	0.680–2.945
Age
$⩾$ 60 years versus 60 years	1.59	0.178	0.810–3.117
Tumour diameter
$⩾$ 3 cm versus 3 cm	1.79	0.157	0.800–3.982
Pathological grade
Grade 1–2 versus grade 3	0.55	0.058	0.301–1.019
Histologic type
Squamous cell carcinoma versus adenocarcinoma	1.05	0.866	0.571–1.946
EGFR expression
Positive versus negative	1.52	0.159	0.847–2.756
ALK expression
Positive versus negative	1.12	0.805	0.443–2.851
PD-L1 expression
Positive versus negative	1.44	0.196	0.826–2.541
T
T1 versus T2 versus T3 versus T4	0.56	0.003${}^{*}$	0.383–0.817	0.62	0.047${}^{*}$	0.389–0.994
N
Positive versus negative	1.43	0.239	0.788–2.601
M
Positive versus negative	4.00	0.175	0.539–29.705
TNM stage
Stage I-II versus stage III-IV	0.44	0.006${}^{*}$	0.245–0.787	0.76	0.441	0.381–1.523

${}^{*}$ 0.05; HR: hazard ration; CI: confidence interval.

Figure 4.

Survival analysis of non-small-cell lung cancer (NSCLC) patients by Kaplan-Meier method. A Overall survival rate in patients with high NDRG3 protein expression (green line) was significantly lower than that in patients with low NDRG3 expression (blue line). B Overall survival rate in patients with advanced tumor status (T3 $+$ T4) (yellow line and purple line) was significantly lower than that in patients with early T (T1 $+$ T2) (blue line and green line).

4. Discussion

Myc, originally recognized as an onco-protein, regulates its target genes and takes great parts in cellular activities, such as proliferation, apoptosis, and pathogenesis [26]. The NDRG family, which is composed of NDRG1-4, is identified as a novel type of Myc-repressed genes and believed to play important roles in diverse biological processes of human cancer [14]. NDRG1 has been reported to exhibit multiple functions in different cancer types [27, 28, 29, 30]. NDRG2 has been shown to inhibit the activation of transcription 3 (STAT3) in breast cancer and induce the downregulation of TCF/$\beta$-catenin signaling in colorectal cancer [31, 32]. NDRG4 is believed to act as a tumor suppressor gene in colorectal cancer and glioma [33, 34]. Although information involvement of NDRG3 and cancer is not extensive, the characteristics of NDRG3 in several cancers have been studied as well. Up-regulation of NDRG3 was detected in both HCC tumor specimens and cell lines [19]. Overexpression of NDRG3 enhanced migration ability of cancer cells in vitro while stimulated tumor growth in nude mice in vivo, suggesting an oncogenic mode of activity [17]. NDRG3 was upregulated when lymphoma cells were treated with silurus asotus lectin (SAL), indicating that NDRG3 could be modulated by signal transduction through the binding of SAL to globotriaosylceramide [35]. NDRG3 contributed to the malignant progression in PCa while high expression of NDRG3 associated with poor survival for PCa patients [20]. Although there were few mechanism studies investigating the characteristics of NDRG3, it was identified to be activated by the ER-$\beta$ and as a part of an antiproliferative activity in the breast cancer [36]. NDRG3 expression could be modulated by CXCL3 in a negative feedback mechanism and finally affected the oncogenic potential of prostate cancer [37]. A latest study reported that microRNA31-NDRG3 axes were substantial HCC survival and drug resistance. NDRG3 anomaly could significantly affect the expression of p53, bcl-2 and caspase-3 [38]. Taking into account the significance of NDRG3, the profound understanding of the function of NDRG3 in NSCLC, deserve further exploration.

In this study, the mRNA expression of NDRG3 was firstly investigated. The data of western-blotting analysis illustrated that NDRG3 expression was elevated in NSCLC cell lines in comparison to that of in non-cancerous cell line. Consistently, the results of qPCR test showed that NDRG3 expression in small NSCLC samples was statistically increased than that in non-cancerous tissues. IHC analysis subsequently demonstrated that the protein expression of NDRG3 in NSCLC TMA was also higher than that in non-cancerous tissues. The data of qPCR and IHC analyses were consistent with the results of previous researches that pointed high expression of NDRG3 in several human cancers [17, 19, 20]. Moreover, high NDRG3 protein expression correlated with three important clinical parameters, including pathological grade, N and TNM stage. The above data also agreed with a previous study and supported the oncogenic behaviors of NDRG3 in tumorigenesis [17].

In survival analysis, NDRG3 expression, T and TNM stage were found to be significantly associated with the overall survival of NSCLC patients by using univariate mode. Multivariate analysis further identified that NDRG3 expression and T may serve as the independent prognostic factors for overall survival. Kaplan-Meier curve finally implied that the life span of patients with NDRG3 expression encountered a significantly worse prognosis than that of patients with low expression. These data are in accord with the previous report that high NDRG3 expression predicted poor survival in PCa [20].

However, Tsui et al. reported that high NDRG3 expression remarkably correlated with the antitumor characteristics of growth differentiation factor-15 (GDF15) in human bladder carcinoma cells [39]; Estiar et al. stated that BC patients with low NDRG3 expression encountered unfavorable event free survival rate [36]. In my view, the inconsistent data indicate that NDRG3 may have different properties in different cancer types or under different circumstances, just like NDRG2. The other issue of this present study is that we only calculate the clinical outcome by analyzing OS, not DFS (disease-free survival) or PFS (progression-free survival). The main reason is that the original TMAs only provide OS data. We will try to make the improvements in collecting clinical samples and corresponding data in our further researches.

In summary, this study reported on the differential expression of NDRG3 in NSCLC, with both mRNA and protein level simultaneously. In addition, the relationship between NDRG3 expression and clinical characteristics of NSCLC patients, especially prognosis status, was investigated for the first time. NDRG3 may be recognized as a novel prognostic biomarker in NSCLC and targeting NDRG3 may provide a therapeutic strategy for NSCLC patients.

Conflict of interest

None.