EIF3B correlates with advanced disease stages and poor prognosis,and it promotes proliferation and inhibits apoptosis in non-small cell lung cancer

Abstract

BACKGROUND:

Although up-regulation of EIF3B correlates with poor prognosis in carcinomas, the role of EIF3B in non-small cell lung cancer (NSCLC) is rarely known.

OBJECTIVE:

We aimed to investigate correlation of EIF3B with clinicopathological features and prognosis in NSCLC patients, and clarify its effect on cells proliferation and apoptosis.

METHODS:

Two hundred and eleven NSCLC patients underwent surgery were retrospectively reviewed. Tumor tissue and paired adjacent tissue were obtained. EIF3B expression was detected by immunohistochemistry, qPCR and western blot. EIF3B inhibitor, blank inhibitor, blank mimic and EIF3B mimic plasmids were transfected to A-549 cells. Cells proliferation and apoptosis were measured by CCK-8 and AV/PI. All processes were repeated for validation in PC9 cells.

RESULTS:

EIF3B expression increased in tumor tissue compared to paired adjacent tissue, and positively correlated with tumor size, lymph node metastasis and TNM stage. K-M curves revealed patients with EIF3B high expression had shorter DFS and OS, and its high expression independently predicted unfavorable DFS and OS. In vitro, EIF3B expression increased in NSCLC cells compared to normal cells. EIF3B increased cells proliferation but inhibited cells apoptosis.

CONCLUSIONS:

EIF3B overexpression correlates with advanced disease conditions and poor prognosis, and it promotes cells proliferation while inhibits apoptosis in NSCLC.

1. Introduction

Non-small cell lung cancer (NSCLC), accounting for over 85% of lung cancer, is one of the most frequently diagnosed malignancies and the leading cause of cancer related death worldwide [1, 2, 3]. Due to lack of obvious clinical symptoms, most patients are diagnosed as NSCLC at an advanced disease stage, contributing to worse prognosis, or even death [2, 4, 5]. Although imaging technology, targeted therapies, immunotherapies and individual treatment strategies have been greatly improved, the prognosis of NSCLC is still not satisfied since the 5-year survival rate is just nearly 15% [6]. Therefore, exploring new and convincing biomarkers for prognosis and clarifying novel underlying mechanisms of tumer progression of NSCLC are necessary.

Eukaryotic initiation factors (eIFs), a series of proteins or protein complexes, are essential in the rate-limiting initiation phase of eukaryotic translation, which play critical roles in the progress and prognosis of several carcinomas [7, 8, 9, 10]. As a member of the eIFs family, Eukaryotic translation initiation factor 3 subunit B (EIF3B) is highly expressed in several carcinomas, including bladder, breast, and prostate cancers [10, 11, 12]. Despite of the fact that up-regulation of EIF3B is correlated with poor prognosis in many carcinomas, the role of EIF3B in NSCLC is rarely known [11, 13, 14]. Therefore, our purpose was to investigate the correlation of EIF3B expression with clinicopathological features and prognosis in NSCLC patients, as well as clarify the effect of EIF3B on the cells proliferation and apoptosis in NSCLC cells.

2. Methods

2.1 Participants

Two hundred and eleven NSCLC patients underwent surgery at The Central Hospital of Wuhan from Jan 2009 to Dec 2012 were retrospectively reviewed in this study. The inclusion criteria were as followed: (1) Diagnosed as primary NSCLC based on clinical and pathological confirmation. (2) Age above 18 years. (3) Completed information of tumor features was available to retrieve. (4) Regularly follow-up patients. (5) Formaldehyde fixed, paraffin embedded tumor tissue sample and paired adjacent tissue sample were available to obtain. The exclusion criteria were as follows: (1) Received neoadjuvant therapy before surgery. (2) History of other solid tumors or hematologic malignancy.

2.2 Ethics approval

This present study was approved by the Ethics Committee of The Central Hospital of Wuhan, and was conducted under the Statement of Helsinki. All the participants or their guardians signed the informed consents or orally agreed with the informed consents by telephone with tape recording.

2.3 Data collection, follow up and survival calculation

Baseline information including age, gender, pathological grade, tumor size, lymph node metastasis and TNM stage were reviewed and collected by Electronic Medical Record System of The Central Hospital of Wuhan in this study. TNM stage was classified according to 7 ${}^{\text{th}}$ edition of the American Joint Committee on Cancer (ATCC) cancer staging manual. The neoplastic grading is a measure of cell anaplasia (reversion of differentiation) in the sampled tumor and is based on the resemblance of the tumor to the tissue of origin. Pathological grade are classified as 1, 2, 3, or 4 depending on the amount of abnormality, among which grade G1 meant well differentiated (low grade), G2 meant moderately differentiated (high grade), G3 meant poorly differentiated (high grade) and G4 meant undifferentiated (high grade) [15]. Patients were followed up regularly and the median follow duration was 49 months, last follow up date was 2017/12/31. Disease-free survival (DFS) was calculated from the time of surgery to the date of documented local or distant recurrence, or death from any cause. overall survival (OS) was calculated from the time of surgery to the date of death from any cause.

2.4 Immunohistochemistry (IHC) assay for EIF3B expression in tissues

Formaldehyde fixed, paraffin embedded tumor tissue sample and paired adjacent tissue sample were obtained from the Specimen Repository of The Central Hospital of Wuhan. After being deparaffinized and rehydrated, antigen retrieval was conducted in tissue sections in Tris-Ethylene Diamine Tetraacetic Acid (EDTA). Then endogenous peroxidase was block by H ${}_2$ O ${}_2$ , and sections were immersed with 4% bovine serum albumin for blocking nonspecific endogenous antigens. Subsequently, sections were incubated with anti-EIF3B antibody (Abcam, USA) as primary antibody at 4 ${}^{\circ }$ C overnight. After washing, the sections were incubated with horseradish peroxidase (HRP)-conjugated anti-IgG (Abcam, USA) as secondary antibody at room temperature for 30 minutes. Afterwards, tissue sections were washed and treated with diaminobenzidine (DAB), followed by counterstaining with hematoxylin. Then, sections were dehydrated and mounted.

As to the expression of EIF3B assessment, the positive cells were observed and counted under a light microscopy by two specialists without knowing any information about patients. In each slide, 100 cells in 5 high-power fields (HPF, $\times$ 400) were counted to evaluate the intensity of positive cells. Staining intensity was scored as 0 (negative), 1 (weak), 2 (moderate), to 3 (strong), while the grading scale for labeling frequency ranged from 0 (0%), 1 (1%–25%), 2 (26%–50%), 3 (51%–75%), and 4 (76%–100%) on the basis of the percentage of positively stained cells. Finally, Multiplying the score of staining intensity by the labeling frequency score was used to divide sections into two groups: EIF3B low expression group with final score $⩽$ 3; and EIF3B high expression group with final score $>$ 3.

2.5 Cells culture

NSCLC cell lines including A-549, NCI-H1650, NCI-H1299 and PC9 as well as normal lung epithelial cells (DEAS-2B) were purchased from Shanghai Institutes for Biological Science (Shanghai, China) or obtained from Shanghai QeeJen Bio-tech Institution (Shanghai, China). A-549 cells were cultured in RPMI F12K medium (Gibco, USA) supplemented with 10% (v/v) fetal bovine serum (FBS) (Gibco, USA) and antibiotics (100 units/ml penicillin and 100 $\mu$ g/ml streptomycin) (Corning, USA). NCI-H1650 cells and NCI-H1299 cells were cultured in RPMI-1640 medium (Gibco, USA) supplemented with 10% (v/v) FBS and antibiotics (100 units/ml penicillin and 100 $\mu$ g/ml streptomycin). PC9 cells were cultured in Dulbecco’s modified Eagles medium (DMEM) (Gibco, USA) supplemented with 10% (FBS), 100 units/ml penicillin and 100 $\mu$ g/ml streptomycin. DEAS-2B cells were cultured in MEM medium (Gibco, USA) supplemented with 10% (FBS), 100 units/ml penicillin and 100 $\mu$ g/ml streptomycin. All cells were incubated at 37 ${}^{\circ }$ C with 5% CO ${}_2$ .

2.6 Detection of EIF3B expression in NSCLC cell lines

Total RNA was extracted from NSCLC cell lines (A-549, NCI-H1650, NCI-H1299 and PC9) and normal lung epithelial cells (DEAS-2B) by TRIzol Reagents (Invitrogen, USA). Quantitative polymerase chain reaction (qPCR) assay and Western Blot assay were performed to detect the mRNA and protein expressions of EIF3B respectively.

2.7 EIF3B plasmids transfection and subsequent assays in A-549 cells

EIF3B inhibitor plasmid, blank inhibitor plasmid, blank mimic plasmid and EIF3B mimic plasmid were designed and constructed by a Shanghai Gemma company (Shanghai, China). These plasmids were transfected into A-549 cells as EIF3B (-), NC (-), NC (+) and EIF3B (+) groups. qPCR assay and Western Blot assay were performed to detect the mRNA and protein expressions of EIF3B at 24 h respectively. counting kit-8 (CCK-8) (Abcam, USA) assay was performed to determine cells proliferation at 0 h, 24 h, 48 h and 72 h. Annexin V (AV) apoptosis detection kit with propidium iodide (PI) (Invitrogen, USA) was used to detect cells apoptosis rate at 72 h.

2.8 Validation in PC9 cells

In order to further validate the effect of EIF3B on cells proliferation and apoptosis in NSCLC, additional experiments in PC9 cells were carried out. EIF3B inhibitor plasmid, blank inhibitor plasmid, blank mimic plasmid and EIF3B mimic plasmid were transfected into PC5 cells, which were classified as EIF3B (-), NC (-), NC (+) and EIF3B (+) groups respectively. After transfection at 0 h, 24 h, 48 h and 72 h, CCK-8 assay (Abcam, USA) for evaluation of cells proliferation was carried out. At transfection at 72 h, AV-PI (Invitrogen, USA) for cells apoptosis rate was performed.

2.9 qPCR assay

The mRNA level of EIF3B was detected by qPCR as follows: According to the manufacturer’s instructions, total RNA samples were extracted with TRIzol reagent (Invitrogen, USA). Total RNA was quantified by OD 260 (Takara, Japan). Then, 1 $\mu$ g of RNA per group was applied to cDNA synthesis separately. Following, those cDNA were subjected to qPCR by SYBR Green kit (Takara, Japan). The PCR profile was 95 ${}^{\circ }$ C for 5 min, and 40 cycles of 95 ${}^{\circ }$ C for 5 s, 61 ${}^{\circ }$ C for 30 s. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as the internal reference. EIF3B expression was calculated by the 2 ${}^{-\mathrm{\Delta }\mathrm{\Delta }\text{ct}}$ method. Primer sequences used in this study were as follows: EIF3B Forward primer: 5’-CGGTGCCTTAGCGTTT GTG-3’; EIF3B Re-verse primer: 5’-CGGTCCTTGT TGTTCTTCTGC-3’; GAPDH Forward primer: 5’-TG ACTTCAACAGCGACACCCA-3’; GAPDH Reverse pri-mer: 5’-CACCCTGTTGCTGTAGCCAAA-3’.

2.10 Western blot assay

Cells were lysed with radioimmunoprecipitation lysis buffer (1% protease inhibitor cocktail, 1% phenylmethanesulfonyl fluoride), and clarified by centrifugation. Subsequently, the protein concentration was detected by the BCA ${}^{\text{TM}}$ protein assay kit (Pierce, USA). Proteins were then separated on sodium dodecyl sulfate-polyacrylamide (SDS) gel electrophoresis and transferred to PVDF membrane. After blocked with 5% non-fat dried milk in PBST at room temperature for 1 h, PVDF membrane was incubated with EIF3B or GAPDH antibodies, and finally incubated with HRP-conjugated secondary antibody (Cell Signaling Technology, USA). The specific band was visualized by an ECL advanced Western blot analysis detection kit (BD, USA).

2.11 CCK-8 assay

Ten $\mu$ l CCK-8 and 90 $\mu$ l medium were added into A-549 cells, and the A-549 cells were then incubated under 5% CO ${}_2$ at 37 ${}^{\circ }$ C. Optical density (OD) value was determined by microplate reader (Biotech, USA) and represented the cells proliferation ability.

2.12 AV/PI assay

After being digested by pancreatin and washed by phosphate buffer solution (PBS), cells were suspended in 100 $\mu$ l Blinding Buffer. Subsequently, 2 $\mu$ l AV (Invitrogen, USA) was subsequently added into the cells and the cells were placed in the darkness for 15 min on the ice. Subsequently, 1 $\mu$ l PI (Invitrogen, USA) was added just before flow cytometry (FCM) assay and the results were analyzed by FCM (Becton Dickinson, USA).

2.13 Statistics

Statistical analysis was performed using SPSS Software 21.0 (IBM, USA) and GraphPad Prism 5.01 (GraphPad, USA). Data were mainly presented as mean $\pm$ standard deviation or count (percentage). Comparison between two groups was determined by t test, Chi-square test or McNemar test. DFS and OS were analyzed by Kaplan-Meier (K-M) curves and log-rank test. Univariate and multivariate Cox’s proportional hazards regression analysis was performed to determine the baseline factors affecting DFS and OS. The predicting values of baseline factors for DFS and OS. $P<$ 0.05 was considered as significant.

Table 1

Characteristics of NSCLC patients

Items	NSCLC patients (N $=$ 211)
Age (years)	61.31 $\pm$ 10.35
Gender (male/female)	143/68
Pathological grade
G1 (n/%)	30 (14.2)
G2 (n/%)	134 (63.5)
G3 (n/%)	47 (22.3)
Tumor size
$⩽$ 5 cm (n/%)	122 (57.8)
$>$ 5 cm (n/%)	89 (42.2)
Lymph node metastasis
Negative (n/%)	133 (63.0)
Positive (n/%)	78 (37.0)
TNM stage
I	71 (33.6)
II	72 (34.1)
III	68 (32.2)

Data were presented as mean $\pm$ standard deviation or count (percentage). NSCLC, non-small-cell lung cancer.

Table 2

Comparison of EIF3B expression between tumor tissue and paired adjacent tissue

Items	EIF3B expression		$P$ value
	Low	High
	expression	expression
Tumor tissue	101 (47.9)	110 (52.1)	$<$ 0.001
Paired adjacent tissue	162 (76.8)	49 (23.2)

Data were presented as count (percentage). Comparison was determined by McNemar test. $P<$ 0.05 was considered significant. EIF3B, eukaryotic translation initiation factors 3B.

Figure 1.

EIF3B expression in tumor tissue and paired adjacent tissue. A: Example of EIF3B high expression in tumor tissue. B: Example of EIF3B low expression in paired adjacent tissue. EIF3B, eukaryotic translation initiation factor 3 subunit B.

3. Results

3.1 Baseline characteristics

As shown in Table 1, the mean age of 211 NSCLC patients was 61.31 $\pm$ 10.35 years. There were 143 males and 68 females. According to the pathological grade, the numbers of patients in G1 grade, G2 grade, and G3 grade were 30 (14.2%), 134 (63.5%), and 47 (22.3%) respectively. There were 122 (57.8%) patients with tumor size $⩽$ 5 cm, while 89 (42.2%) patients with tumor size $>$ 5 cm. 133 (63.0%) patients presented with negative lymph node metastasis, while 78 (37.0%) patients presented with positive lymph node metastasis. As for the TNM stage, the numbers of patients in stage I, stage II and stage III were 71 (33.6%), 72 (34.1%) and 68 (32.2%) respectively.

3.2 EIF3B expression in tumor tissue and paired adjacent tissue

EIF3B expression was detected by IHC assay, which disclosed that EIF3B expression was higher in tumor tissue (Fig. 1A) compared to paired adjacent tissue (Fig. 1B) ( $P<$ 0.001, Table 2).

Table 3

Correlation of EIF3B expression with clinicopathological features of NSCLC patients

Items	EIF3B Low	EIF3B High	$P$ value
	expression	expression
	( $n=$ 101)	( $n=$ 110)
Age			0.168
$⩽$ 60 years	49 (53.5)	43 (46.7)
$>$ 60 years	52 (43.7)	67 (56.3)
Gender			0.189
Male	64 (44.8)	79 (55.2)
Female	37 (54.4)	31 (45.6)
Pathological grade			0.252
G1	18 (60.0)	12 (40.0)
G2	59 (44.0)	75 (56.0)
G3	24 (51.1)	23 (48.9)
Tumor size			$<$ 0.001
$⩽$ 5 cm	44 (60.7)	48 (39.3)
$>$ 5 cm	27 (30.3)	62 (69.7)
Lymph node metastasis			$<$ 0.001
Negative	79 (59.4)	54 (40.6)
Positive	22 (28.2)	56 (71.8)
TNM stage			0.002
I	45 (63.4)	26 (36.6)
II	33 (45.8)	29 (54.2)
III	23 (33.8)	45 (66.2)

Data were presented as count (percentage). Comparison was determined by Chi-square test. $P<$ 0.05 was considered significant. EIF3B, eukaryotic translation initiation factors 3B; NSCLC, non-small-cell lung cancer.

3.3 Correlations of EIF3B expression with clinicopathological features in NSCLC patients

EIF3B expression was positively correlated with tumor size ( $P<$ 0.001), lymph node metastasis ( $P<$ 0.001) and TNM stage ( $P=$ 0.002), while no correlation of EIF3B expression with age ( $P=$ 0.168), gender ( $P=$ 0.189), or pathological grade ( $P=$ 0.252) in NSCLC patients was observed (Table 3).

3.4 Correlations of EIF3B expression with DFS and OS in NSCLC patients

K-M curves revealed that NSCLC patients with EIF3B high expression presented with shorter DFS ( $P<$ 0.001, Fig. 2A) and OS ( $P<$ 0.001, Fig. 2B) than that in patients with EIF3B low expression.

3.5 Factors affecting DFS in NSCLC patients by Cox’s analysis

Univariate Cox’s proportional hazards regression was employed to assess factors influencing DFS in NSCLC patients (Table 4), which disclosed that EIF3B high expression ( $P<$ 0.001), poor pathological grade ( $P<$ 0.001), and higher TNM stage ( $P=$ 0.024) were associated with shorter DFS. Multivariate Cox’s proportional hazards regression disclosed that EIF3B high expression ( $P<$ 0.001), age $>$ 60 years ( $P=$ 0.044) and poor pathological grade ( $P<$ 0.001) could independently predict shorter DFS in NSCLC patients (Table 4).

Table 4

Cox’s analysis of factors affecting DFS

Items	Univariate Cox’s proportional hazard regression				Multivariate Cox’s proportional hazard regression
	$P$ value	HR	95% CI		$P$ value	HR	95% CI
			Lower	Higher			Lower	Higher
EIF3B high expression	$<$ 0.001	1.889	1.408	2.534	$<$ 0.001	1.816	1.337	2.467
Age $>$ 60 years	0.167	1.230	0.917	1.650	0.044	1.368	1.008	1.857
Gender (male)	0.321	1.174	0.860	1.604	0.751	0.949	0.689	1.309
Pathological grade	$<$ 0.001	1.506	1.196	1.895	$<$ 0.001	1.577	1.223	2.033
TNM stage	0.024	1.222	1.026	1.454	0.453	1.074	0.892	1.293

Data were presented as P value, HR and 95%CI. Baseline factors affecting DFS were determined by univariate and multivariate Cox’s proportional hazards regression model analysis. $P<$ 0.05 was considered significant. Pathological grade was scored as 1-G1, 2-G2, 3-G3. TNM stage was scored as 1-I, 2-II, 3-III. DFS, disease free survival; HR, hazard ratio; CI, confidence interval; EIF3B, eukaryotic translation initiation factors 3B.

Figure 2.

Correlation of EIF3B expression with DFS and OS in NSCLC patients. A: Shorter DFS was observed in EIF3B high expression patients. B: Worse OS was presented in patients with EIF3B high expression. Kaplan-Meier curves and log-rank test were performed for comparison of DFS and OS. $P<$ 0.05 was considered as significant. EIF3B, eukaryotic translation initiation factor 3 subunit B. DFS, disease free survival. OS, overall survival. NSCLC, non-small cell lung cancer.

Figure 3.

EIF3B expression in NSCLC cell lines. A: mRNA level of elF3B was increased in NSCLC cell lines (including A-549, NCI-H1299, NCI-H1650 and PC9 cells) compared to normal lung epithelial cell line DEAS-2B, and the most significant elevation was in A-549 cells. B: Protein level of elF3B was upregulated in NSCLC cell lines including A-549, NCI-H1299, NCI-H1650 and PC9 cells compared to normal lung epithelial cell line DEAS-2B, and most significantly increased in A549 cells. Comparison between two groups was determined by t test. $P<$ 0.05 was considered as significant. ${}^{*}P<$ 0.05, ${}^{**}P<$ 0.01. EIF3B, eukaryotic translation initiation factor 3 subunit B. NSCLC, non-small cell lung cancer.

Table 5

Cox’s analysis of factors affecting OS

Items	Univariate Cox’s proportional hazard regression				Multivariate Cox’s proportional hazard regression
	$P$ value	HR	95% CI		$P$ value	HR	95% CI
			Lower	Higher			Lower	Higher
EIF3B high expression	$<$ 0.001	2.018	1.454	2.802	$<$ 0.001	1.929	1.373	2.709
Age $>$ 60 years	0.501	1.117	0.809	1.543	0.186	1.258	0.895	1.769
Gender (male)	0.784	0.954	0.681	1.336	0.122	0.761	0.539	1.076
Pathological grade	$<$ 0.001	1.626	1.258	2.102	$<$ 0.001	1.674	1.253	2.236
TNM stage	0.003	1.351	1.110	1.644	0.153	1.167	0.944	1.441

Data were presented as $P$ value, HR and 95% CI. Baseline factors affecting OS were determined by univariate and multivariate Cox’s proportional hazards regression model analysis. $P<$ 0.05 was considered significant. Pathological grade was scored as 1-G1, 2-G2, 3-G3. TNM stage was scored as 1-I, 2-II, 3-III. OS, overall survival; HR, hazard ratio; CI, confidence interval; EIF3B, eukaryotic translation initiation factors 3B.

Figure 4.

EIF3B promoted proliferation and inhibited apoptosis of A-549 cells. A: mRNA level of EIF3B was downregulated in EIF3B (-) group and upregulated in EIF3B (+) group compared to corresponding NC groups. B: Protein level of EIF3B was lower in EIF3B (-) group and higher in EIF3B (+) group than corresponding NC groups. C: EIF3B promoted cells proliferation of A549 cells. D and E: EIF3B inhibited cells apoptosis of A549 cells. Comparison between two groups was determined by t test. $P<$ 0.05 was considered as significant. ${}^{*}P<$ 0.05, ${}^{**}P<$ 0.01, ${}^{***}P<$ 0.001. NC, negative control. EIF3B, eukaryotic translation initiation factor 3 subunit B.

3.6 Factors affecting OS in NSCLC patients by Cox’s analysis

As shown in Table 5, EIF3B high expression ( $P<$ 0.001), poor pathological grade ( $P<$ 0.001) and higher TNM stage ( $P=$ 0.003) were correlated with poor OS. Multivariate Cox’s proportional hazards regression revealed that EIF3B high expression ( $P<$ 0.001) and poor pathological grade ( $P<$ 0.001) independently predict unfavorable OS in NSCLC patients (Table 5).

3.7 EIF3B expression in NSCLC cell lines

In order to investigate the effect of ElF3B on NSCLC cells, in vitro experiments were further performed. Compared to normal lung epithelial cells DEAS-2B, both mRNA (Fig. 3A) and protein expressions (Fig. 3B) of EIF3B were increased in NSCLC cell lines including A-549, NCI-H1299, NCI-H1650 and PC9, and was most significantly elevated in A-549 cells. Therefore, A-549 cell line was chosen for the following in vitro experiments.

Figure 5.

EIF3B induced cells proliferation and suppressed cells apoptosis of PC9 cells. A: mRNA level of EIF3B was decreased in EIF3B (-) group and increased in EIF3B (+) group compared to corresponding NC groups. B: Protein level of EIF3B was decreased in EIF3B (-) group and increased in EIF3B (+) group than corresponding NC groups. C: EIF3B promoted cells proliferation of PC9 cells. D-E: EIF3B inhibited cells apoptosis of PC9 cells. Comparison between two groups was determined by t test. $P<$ 0.05 was considered as significant. ${}^{*}P<$ 0.05, ${}^{**}P<$ 0.01, ${}^{***}P<$ 0.001. NC, negative control. EIF3B, eukaryotic translation initiation factor 3 subunit B.

3.8 EIF3B promoted proliferation and inhibited apoptosis in A-549 cells

Both mRNA and protein expressions of EIF3B were decreased in EIF3B (-) group compared to NC (-) group and increased in EIF3B (+) group compared to NC (+) group, indicating the transfection was successful (Fig. 4A and B). At 72 h, cells proliferation was promoted in EIF3B (+) group compared to NC (+) group, while it was suppressed in EIF3B (-) group compared to NC (-) group (Fig. 4C), suggesting that EIF3B promoted proliferation in A-549 cells. Cells apoptosis rate was increased in EIF3B (-) group compared to NC (-) group, while it was decreased in EIF3B (+) group compared to NC (+) group at 72 h after transfection (Fig. 4D and E), suggesting that EIF3B inhibited apoptosis in A-549 cells.

3.9 Validation for the effect of EIF3B on cells proliferation and cells apoptosis in PC9 cells

To further verify the efficacy of EIF3B on cells proliferation and apoptosis in NSCLC, additional experiments were repeated in another NSCLC cells line PC9 cells. As shown in Fig. 5, mRNA and protein expressions of EIF3B were evaluated in EIF3B (+) group compared to NC (+) group and reduced in EIF3B (-) group compared to NC (-) group, suggesting the transfection was successful (Fig. 5A and B). Cells proliferation was increased in EIF3B (+) group compared to NC (+) group and was decreased in EIF3B (-) group compared to NC (-) group at 72 h (Fig. 5C), indicating that EIF3B promoted proliferation in PC9 cells. For cells apoptosis rate, it was decreased in EIF3B (+) group compared to NC (+) group but increased in EIF3B (-) group compared to NC (-) group at 72 h after transfection (Fig. 5D and E), which indicated that EIF3B suppressed cells apoptosis in PC9 cells.

4. Discussion

In this study, we observed that: (1) EIF3B expression was higher in tumor tissue than that in adjacent tissue, and it was positively associated with tumor size, lymph node metastasis and TNM stage. Moreover, EIF3B overexpression correlated with worse DFS and OS independently in NSCLC patients. (2) EIF3B promoted proliferation and inhibited apoptosis in NSCLC cells.

EIFs families, which participate in the eukaryotic translation via regulating interaction between ribosome and mRNA, are propo15sed as new therapeutic targets for carcinomas due to their functions in promoting tumorigenesis [14, 16, 17]. An interesting experiment indicates that down-regulation of eukaryotic translation initiation factor 5A2 (EIF5A2) suppresses cells growth in ovarian cancer cell line UACC-1598 [18]. Also, eukaryotic translation initiation factor 3 subunit D (EIF3D) plays a crucial role in promoting cells proliferation via affecting G2/M phase arrest in NSCLC [19]. Additionally, eukaryotic initiation translation factor 2 $\alpha$ (EIF2 $\alpha$ ) phosphorylation reduces cisplatin-induced apoptosis of NSCLC A-549 cells via decreasing p38 expression [20]. Hence, those previous data suggest that EIFs may be oncogenes in carcinomas, including NSCLC.

EIF3B, a common member of EIFs family, has been observed to be overexpressed in many cancers including esophageal squamous cell carcinoma, clear cell renal cell carcinoma, and prostate cancers [11, 14, 17]. In addition, up-regulation of EIF3B is found to be positively correlated with tumor depth, lymph node metastasis and TNM stage in patients with esophageal squamous cell carcinoma [17]. Also, EIF3B high expression is associated with an aggressive tumor phenotype in patients with clear cell renal cell carcinoma [14]. Furthermore, EIF3B expression positively correlates with tumor grade in patients with bladder and prostate cancers [11]. Therefore, these previous studies suggest that EIF3B overexpression associates with advanced disease conditions in patients with several cancers. In line with those previous data, we found that EIF3B expression was higher in tumor tissue than that in adjacent tissue, and it was positively correlated with tumor size, lymph node metastasis and TNM stage in NSCLC patients. The possible reasons are as follows: EIF3B has effect on promotion of cells proliferation, facilitation of cells migration or inhibition of cells apoptosis via modulating various genes expressions or pathways activations (including decreasing integrin $\alpha$ 5 expression, inhibiting tumor necrosis factor receptor superfamily member 21 expression or activating $\beta$ -catenin signaling pathway), thereby contributing to advanced disease conditions in NSCLC patients [11, 13, 17].

As to the predictive value of EIF3B on the prognosis of cancers, some studies have been performed. For instance, esophageal squamous cell carcinoma patients with high EIF3B expression present with poor DFS and OS [17]. For clear cell renal cell carcinoma, patients with elF3B overexpression also illuminate shorter OS [14]. Additionally, up-regulation of EIF3B correlates with poor disease-specific survival in patients with bladder cancer [11]. Even though those previous studies indicate that EIF3B acts as a novel biomarker for worse prognosis in some cancers, the effect of EIF3B on the prognosis of NSCLC is still unclear. In the present study, we observed that EIF3B overexpression independently correlated with worse DFS and OS in NSCLC patients. The possible reasons were as follows: EIF3B could influence expressions of some proteins or activation of pathways (such as reducing apoptosis inhibitory protein BCL2 expression or activating the pro-survival Akt pathway) to induce cells proliferation and inhibit cells apoptosis, thereby devoting to advanced disease conditions and finally leading to poor prognosis in NSCLC patients [14, 21, 22]. Moreover, elF3B might promote drug resistance via mediating some pathways, including activating the Akt pathway, thereby decreasing treatment efficacy and resulting in poor prognosis in NSCLC patients [21].

In order to investigate the function of EIF3B in cancers and clarify the underlying mechanisms, some in vitro experiments have been performed. For example, eIF3b contributes to osteosarcoma cells proliferation by regulating tumor necrosis factor receptor superfamily member 21 expression [13]. Also, depletion of EIF3B inhibits cells growth, prevents cells migration, hinders cells cycle, as well as disrupts actin cytoskeleton and focal adhesions by decreasing the expression of integrin $\alpha$ 5 in bladder cancer cells [11]. Additionally, knockdown of EIF3B inhibits cells proliferation by cell cycle arrest in U87 cells [23]. Therefore, these previous data indicate that EIF3B promotes tumorigenesis through facilitating cells proliferation and inhibiting cells apoptosis in several carcinomas, while the function of EIF3B in NSCLC cells is still unclear. To clarify the role of EIF3B in cells proliferation and apoptosis of NSCLC, a series of in vitro experiments were performed in this study, which illuminated that EIF3B promoted proliferation and inhibited apoptosis in A-549 cells, indicating that EIF3B might be involve in NSCLC pathology via increasing cells proliferation and decreasing apoptosis.

In conclusion, EIF3B overexpression correlates with advanced disease conditions and poor prognosis, and it promotes cells proliferation while inhibits cells apoptosis in NSCLC.