Expression levels of microRNA-145 and microRNA-10b are associated with metastasis in non-small cell lung cancer

ABSTRACT

Although metastasis remains the overwhelming cause of death for patients with non-small cell lung cancer (NSCLC), the underlying mechanisms of metastasis remain unknown. Accumulating evidence suggests that microRNAs (miRNAs) are key players in the regulation of tumor cell invasion and metastasis. Expression of miR-9, miR-10b, miR-145, and miR-155, 4 miRNAs previously shown to play roles in metastasis in other tumor types, was compared in lymph node (LN)-positive NSCLC versus LN-negative NSCLC. Expression of miR-145 was significantly lower in LN-positive NSCLC (P < 0.05), while expression of miR-10b was significantly higher (P < 0.05). Expression of both miR-145 and miR-10b was correlated with lymph node metastasis in NSCLC (both Ps < 0.001). In addition, miR-10b facilitated the migration and invasion of lung cancer cell line A549, while miR-145 suppressed the migration and invasion capacity of A549 in vitro. These results suggest that miR-10b and miR-145 may act as an oncogene or tumor suppressor gene, respectively, in NSCLC metastasis.

Introduction

Lung cancer is one of the leading causes of cancer-related mortality in the world, with non-small cell lung cancer (NSCLC) accounting for approximately 80% of all lung cancers.^1,2 Despite significant progress in the diagnosis and treatment of lung cancer over the last 20 years, the long-term survival rate remains low at 10–15%. The cause of death for 80–90% of lung cancer patients is the development of metastatic lesions at sites distant from that of the primary tumor.^3-5 While 50–60% of patients develop metastasis during the course of treatment, up to 30% of patients present with metastasis at the time of diagnosis. In order to improve the low survival rate, it is critical to better understand the process of metastasis, the most devastating and lethal attribute of lung cancer.

Metastasis is a well-regulated process, and numerous factors impact the potential of a tumor cell to metastasize. Accumulating evidence suggests that microRNAs (miRNAs) are key players in the regulation of tumor cell invasion and metastasis. MiRNAs are a family of small noncoding RNAs, approximately 21–25 nucleotides in length, which serve as negative regulators by interacting with the 3′-untranslated region (3′-UTR) of mRNAs, resulting in mRNA degradation or repression of translation.⁶ MiRNAs play a fundamental role in the regulation of diverse cellular functions, and deregulation of miRNA expression is often associated with a variety of disorders, including human malignancies.^7,8 For instance, Ma et al. revealed that over-expression of microRNA-10b (miR-10b) initiated invasion and metastasis in models of breast cancer, and its expression in primary breast carcinomas is correlated with clinical progression.⁹ In addition, Sachdeva et al. found that miR-145 inhibited tumor growth, invasion, and metastasis in human breast cancer and colon cancer cells.^10,11 While miR-9, miR-10b, miR-145, and miR-155 have been shown to regulate metastasis in other types of solid tumors, their role in NSCLC metastasis remains unknown.

Therefore, in the present study, we investigated the possible contributions of these 4 metastasis-related miRNAs to NSCLC. We compared their expression in lymph node-positive and -negative primary NSCLC tumors and assessed their role in metastatic behavior using invasion and migration assays.

Materials and methods

Patients and tissue specimens

Study participants included 58 patients who were diagnosis with NSCLC and underwent surgical resection of lung cancer between 2006 and 2010 at the Department of Lung Cancer Surgery, Tianjin Medical University General Hospital. Written informed consent was obtained, and the institutional ethics committee of Tianjin Medical University General Hospital approved the study. The cases were selected based on the following criteria: (1) diagnosis of primary lung cancer clinical stage I to IV (pTNM); (2) undergoing surgical resection without prior chemotherapy or TKI treatment; (3) availability of outcome and follow-up data. Pathologic diagnosis was based on WHO criteria. Lung cancer staging for each patient was performed according to the AJCC Cancer Staging Manual, 7th edition, and was based on findings from physical examination, surgical resection, and computed tomography of the chest, abdomen, pelvis, and brain. The following information was collected from the patients' medical records: age, gender, clinical stage, pathologic diagnosis, differentiation, lymph node status, metastasis, smoking status, and overall survival time. Survival was calculated from the day of resection until April 1, 2011. Resected lung and lymph node tissues were immediately immersed in liquid nitrogen until RNA extraction.

Cells and cell culture

The A549 cell line was obtained from American Type Culture Collection (ATCC, Manassas, VA, USA). The cells were cultured in RPMI 1640 medium (GIBCO, Gaithersburg, MD, USA) supplemented with 10% fetal bovine serum (FBS) (GIBCO, Grand Island, USA), penicillin (100 units/mL), and streptomycin (100 units/mL) at 37°C in an atmosphere of 5% CO₂ in a humidified incubator.

miRNA extraction

miRNAs were extracted from tissues and cell lines using the miRNeasy Mini Kit (QIAGEN, Cat. 217004) according the manufacturer's instructions. RNA was quantified using a spectrophotometer (Beckman, USA), and its quality was checked by agarose gel electrophoresis.

Reverse transcription and quantitative real-time RT-PCR for miRNA expression

Total RNA (50 ng) was reverse transcribed into cDNA (cDNA) using the ABI miRNA reverse transcription kit (Applied Biosystems, Foster City, CA). TaqMan miRNA assays were used to quantify the expression levels of mature miR-9, miR-10b, miR-145, and miR-155 according to the manufacturer's instructions (Applied Biosystems). Approximately 0.6 μl cDNA and 5 μl of Universal PCR Master Mix without AmpErase UNG was added to commercially available PCR primers and FAM-labeled TaqMan probes for each miRNA (Applied Biosystems). The snRNA U6 (Applied Biosystems) was used as a normalization control for all samples. Each measurement was performed in triplicate. Real-time RT-PCR was performed using an Applied Biosystems 7900HT Fast Real-Time PCR System instrument and software. PCR data were analyzed using the comparative Ct method (ΔCt), as previously described for determining relative gene expression.^12,13 Relative miRNA expression of a target gene within a specimen was calculated as 2^−ΔCt, where ΔCt=Ct (miRNA) - Ct(U6).

In vitro migration and invasion assays

For transwell migration assays, 5 × 10⁴ cells were placed in the top chamber of each insert with an uncoated membrane (BD Biosciences, NJ). For invasion assays, 1 × 10⁵ cells were placed in the upper chamber of each insert coated with 150 μg Matrigel (BD Biosciences, NJ). For both assays, cells were trypsinized and resuspended in RPMI 1640 medium, and 600 μl of medium supplemented with 10% fetal bovine serum was injected into the lower chambers. After incubation at 37°C for 16 hours for the migration assay and 48 hours for the invasion assays, the cells remaining in the top chambers or on the upper membrane of the inserts were carefully removed. After fixation and staining in a dye solution containing 0.1% crystal violet and 20% methanol, cells adhering to the lower membrane of the inserts were counted and imaged using a TE2000 inverted microscope (Nikon, Tokyo, Japan). Results were the average of 3 independent experiments.

For wound healing assay, 5 × 10⁵ cells were seeded into 6-well plates after transfection and were cultured for 24 h at full confluency. A sterile 200 μl pipette tip was used to scratch the cells to form a wound. Cell debris was removed by washing with PBS and then replace with normal cell culture medium. Then,at the indicated time points of 0 hour, 24 hours, and 42 hours, cells were observed and photographs were taken by NIKON TE2000 inverted microscope. The distance of the wound zone was measured with Image J software. The relative migration was calculated by following formula:( 0 h wound width −24 h or 48 h wound width)/0 h wound width × 100%.

Statistical analysis

Statistical analysis was performed using SPSS software version 21.0 (Chicago, IL). The difference in miRNA expression between primary tumor tissues and paired metastatic lymph node tissues was compared by Wilcoxon signed-rank test. Mann-Whitney U and Kruskal-Wallis tests were performed to determine the relationship between miRNA expression and clinicopathological parameters. Survival was estimated using the Kaplan-Meier method, and the differences in survival according to miRNA expression were compared using the log-rank test. Differences between control and miRNA transfected groups were analyzed using the Student's t-test. A P value < 0.05 was considered statistically significant. All tests of significance were 2-sided.

Results

Characteristics of the research cohort

The study cohort included 37 male and 21 female NSCLC patients aged 39 to 78 years old (median 62.5 years) (Table 1). Of these 58 patients, 26 had squamous cell carcinoma (SCC), 25 had adenocarcinomas (ADC), 4 were mixed SCC and ADC, 2 had large cell carcinomas, and 1 had a carcinoid tumor. Lymph node metastasis occurred in 44 cases. Finally, 2 patients were diagnosed with stage I NSCLC, 32 with stage II, 15 with stage III, and 9 with stage IV.

Table 1.

Demographic and Clinical Characteristics of Patients with NSCLC (n = 58).

		Metastasis status
			LN-Negative	LN-Positive
Characteristic		N (%)	N (%)	Total
Gender
	Male	9 (64.3%)	28 (63.6%)	37 (64%)
	Female	5 (35.7%)	16 (36.4%)	21 (36%)
Age
	>62.5	8 (57.1%)	20 (45.5%)	29 (50%)
	≤62.5	6 (42.9%)	24 (54.5%)	29 (50%)
Histology
	ADC	5 (35.7%)	20 (45.5%)	25 (43%)
	SCC	8 (57.1%)	18 (40.9%)	26 (45%)
	Others	1 (7.1%)	6 (13.6%)	7 (12%)
Smoking history
	Nonsmoker	4 (28.6%)	16 (36.4%)	20 (34%)
	Smoker	10 (71.4%)	28 (63.6%)	38 (66%)
Stage
	I	2 (14.3%)		2 (3%)
	II	9 (64.3%)	23 (52.3%)	32 (55%)
	III	1 (7.1%)	14 (31.8%)	15 (26%)
	IV	2 (14.3%)	7 (15.9%)	9 (16%)

ADC, adenocarcinomas; SCC, squamous cell carcinoma; LN, lymph node.

Correlation between miRNA expression and clinicopathological characteristics in NSCLC

We investigated the correlation between expression levels of miR-145, miR-155, miR-10b, and miR-9 and clinicopathological characteristics of NSCLC, as shown in Table 2. Expression of both miR-145 and miR-10b was correlated with lymph node metastasis in NSCLC (P < 0.001 for both; see Table 2). Expression of miR-9 was correlated with histology and TNM stage (P = 0.049 and P = 0.029, respectively). Interestingly, the expression of miR-10b was not correlated with gender, smoking status, or TNM stage, but was correlated with age and histology type (P = 0.048 and P = 0.021, respectively). The expression of miR-145 and miR-155 was not correlated with age, gender, smoking status, or TNM stage. To further investigate the correlation between expression of miR-145 and miR-10b and lymph node metastasis, NSCLC samples were divided into 4 groups according to miR-145 and miR-10b expression levels (Fig. 1). Tumors with low miR-145 or high miR-10b expression invariably showed an invasive phenotype. Specifically, in the group with the lowest miR-145 expression (0–0.09), 100% (6/6) of NSCLC tumors were LN-positive compared to the group with the highest miR-145 expression (2.00–100.00), which was 23.1% (3/13) LN-positive. In addition, NSCLC tumors in the highest miR-10b expression group (0.010–0.060) were 100% (11/11) LN-positive compared to 53% (8/15) in the lowest group (Fig. 1). The logistic regression of those 2 microRNAs were all statistically significant (P < 0.001 for miR-145 and p = 0.043 for miR-10b respectively); the odds ratio associated with miR-145 expression and metastasis was 0.029 (95% confidence interval = 0.06 to 0.152) and the odds ratio associated with miR-10b expression and metastasis was 3.819 (95% confidence interval = 1.046 to 13.943).

Table 2.

Median expression levels of 4 miRNAs and their associations with clinicopathological factors in NSCLC.

Variable	Number (percent)	miR-145	miR-155	miR-10b	miR-9
Age
>62.5	30(51.7)	0.3350	0.0123	0.0022	0.0009
<=62.5	28(48.3)	0.3988	0.0102	0.0008	0.0009
P value		0.803	0.565	*0.048	0.503
Gender
Male	37(63.8)	0.3213	0.0090	0.0021	0.0013
Female	21(36.2)	0.6242	0.0127	0.0008	0.0004
P value		0.289	0.414	0.777	0.071
Histology
AD	25(43.1)	0.4931	0.0104	0.0008	0.0004
SCC	26(44.8)	0.3570	0.0127	0.0026	0.0015
Others	7(12.1)	0.3802	0.0040	0.0005	0.0010
P value		0.898	0.089	*0.021	*0.049
Smoking status
Non-smoker	20(34.5)	0.6455	0.0085	0.0015	0.0003
Smoker	38(65.5)	0.3189	0.0125	0.0012	0.0011
P value		0.395	0.266	0.974	0.155
Lymph node status
No metastasis	14(24.1)	7.0030	0.0086	0.0009	0.0008
Metastasis	44(75.9)	0.6720	0.0122	0.0016	0.0009
P value		*<0.001	0.435	*<0.001	0.856
TNM Stage
I	2(3.4)	0.0924	0.0276	0.0220	0.0003
II	32(55.2)	0.3435	0.0099	0.0014	0.0006
III	15(25.9)	0.3802	0.0152	0.0011	0.0013
IV	9(15.5)	0.3605	0.0159	0.0008	0.0027
P value		0.179	0.601	0.386	0.082
Stage
I+II	34(58.6)	0.2350	0.0099	0.0015	0.0005
III+IV	24(43.4)	0.3771	0.0156	0.0011	0.0018
P value		0.190	0.312	0.377	*0.029

Expression of the 4 miRNAs was detected by qRT-PCR in primary tumor tissues from 58 NSCLC patients, and the median expression of each group is listed. The association between relative expression of each miRNA and clinicopathologic characters was analyzed using Mann-Whitney U and Kruskal-Wallis test as appropriate; p-values are given in italics. Total number of patients included in the analyses was 58, median age was 62.5 years.

Abbr: LN-negative: lymph node negative; LN-positive: lymph node positive.

Figure 1.

The expression levels of miR-145 and miR-10b are correlated with lymph node metastasis status. NSCLC samples were divided into 4 groups according to miR-145 and miR-10b expression levels. The proportion of patients with metastasis for each group is shown. Tumors with low miR-145 and high miR-10b expression invariably showed an invasive phenotype. The logistic regression analysis showed that the expression levels of miR-145 and miR-10b were all statistically significantly correlated with lymph node metastasis status (P < 0.001 for miR-145 and P = 0.043 for miR-10b, respectively).

Expression levels of miR-145 and miR-10b are associated with metastasis in primary NSCLC tissues

To further determine whether expression of miR-9, miR-10b, miR-145, and miR-155 is correlated with metastasis in NSCLC, we compared their expression in tumors from patients with lymph node (LN) metastasis (LN-positive, n = 44) to those from patients without metastasis (LN-negative, n = 14). We found that miR-145 expression was more than 10-fold lower (0.672 vs. 7.003, P < 0.001) (Fig. 2, Table 2). In contrast, expression of miR-10b was significantly higher in LN-positive tumors (0.0016 versus 0.0009, P < 0.001). There was no significant difference in expression of miR-155 or miR-9 (P = 0.435 and P = 0.856, respectively).

Figure 2.

MiR-145 (A), miR-155 (B), miR-10b (C), and miR-9 (D) expression in clinical NSCLC specimens with (n = 44) or without (n = 14) lymph node metastasis. The term -△Ct was used to describe the expression levels of miR-145, miR-155, miR-10b, and miR-9 in primary lymph node-negative and lymph node-positive NSCLC. Graphs represent the expression of miRNAs in primary tumor samples relative to the U6 control. The line indicates the mean value for each group; n is the number of clinical specimens. There was a significant difference in miR-145 (P < 0.0001) and miR-10b (P < 0.0001) expression between these 2 groups.

The expression of miR-145 and miR-10b were significantly different in paired primary lung cancer and metastatic lymph node tissues

Expression of the miRNAs was then compared in 44 cases of paired human NSCLC primary lung cancer and matched metastatic lymph node tissues. The expression in the primary lung cancer tissue was scaled to 1 to allow for comparison. Two of the miRNAs, miR-145 and miR-10b, were found to be differentially expressed between primary lung cancer tissues and matched metastatic lymph node tissues (Fig. 3). MiR-145 expression in metastatic lymph nodes was significantly down-regulated compared to the matched primary lung cancer tissues, with lower expression in 30 cases (30/44, 68.2%) (P = 0.033 Wilcoxon paired rank sum test). MiR-10b expression in metastatic lymph nodes was significantly up-regulated compared to the matched primary lung cancer tissues, with higher expression in 34 cases (34/44, 77.3%) (P < 0.001, Wilcoxon paired rank sum test). There was no significant association for miR-155 or miR-9 (P = 0.086 and P = 0.521, respectively).

Figure 3.

Comparison of miR-145 (A), miR-155 (B), miR-10b (C), and miR-9 (D) expression levels in paired primary lung cancer tissues and metastatic lymph node tissues (n = 44). Each bar represents the relative expression of normalized miRNA expression in 44 pairs of matched primary NSCLC and lymph node metastasis tumor samples. The expression of miRNAs in the primary sample of each pair was scaled to 1 to allow comparison. Expression of miR-145 and miR-10b was significantly different in paired primary lung cancer and metastatic lymph node tissues.

High expression of miR-10b tends to be associated with poor outcome

As of May 2011, only 55 of the 58 patients had complete follow-up data, including 16 deaths and 39 cases of survival. We conducted a survival analysis of these 55 patients based on expression levels for each miRNA of greater or less than the median. As shown in complementary Fig. 1, Kaplan-Meier survival analyses revealed that none were associated with survival. However, patients whose primary tumors displayed high expression of miR-10b tended to have a shorter median survival time (583 days vs. 810 days, P = 0.064); the sample size was small, which may have resulted in this not achieving statistical significance.

miRNA-10b and miR-145 affect NSCLC cell invasion and metastasis in vitro

To further investigate whether miR-10b and miR-145 affect NSCLC cell migration and invasion, we evaluated cancer cell invasion and migration using wound healing assay and transwell assay. A549 cells were infected with miR-10b, miR-145, or a non-target scrambled oligonucleotides. As shown in Fig. 4A, compared to the scrambled groups, Q-RT-PCR analysis revealed that both miR- 10b and miR-145 were efficiently overexpressed (389-fold higher in miR-10b mimic-transfected group and 686-fold higher in miR-145 mimic-transfected group while compared to the scrambled groups, Ps < 0.01, respectively). Next, the wound-healing assay was performed to detect the migration activity of A549 cells, as shown in Fig. 4B. Compared to the scrambled groups, the results showed that the migration ability was significantly increased with overexpression of miR-10b at 24 h and 48 h (62.12% versus 25.49%, P < 0.05; 93.39% vs. 60.78% P < 0.05, respectively), while the migration ability was significantly decreased with overexpression of miR-145 at 24 h and 48 h (20.68% versus 25.49%, P < 0.05; 29.31% vs. 60.78%, P < 0.01, respectively). Furthermore, as shown in Figs. 4C–D, overexpression of miR-10b resulted in significantly more cells penetrating the transwell membrane (79.4 versus 19.8, 2.9-fold increased, P < 0.05 for invasion, and 98.6 vs. 61.6, 2.6-fold increased, P < 0.01 for migration assays), while overexpression of miR-145 suppressed the average number of cells penetrating the transwell membrane in migration (8.4 versus 24.4, 2.9-fold decreased, P < 0.01) and invasion assays (23.8 vs. 63.4, 2.6-fold decreased, P < 0.01). These data indicate that miR-10b facilitates the migration and invasion capacity of A549 cells, while miR-145 restricts the migration and invasion capacity of A549 cells in vitro.

Figure 4.

MiR-10b and miR-145 affect NSCLC cell invasion and migration in vitro. (A) The A549 cells were transfected with miR-10b, miR-145 or control scrambled oligonucleotides respectively. Q-RT-PCR analyzes the miRNAs levels in different treated. (B) Wound healing assay analyzes the migration activity in different treated cells. The vertical axis of coordinate represent the relative migration activity of cells. Invasion assays (top) and transwell assays (bottom) of A549 cells transfected with (C) miR-10b or scrambled or (D) miR-145 or scrambled oligonucleotides. Representative images are shown on the left, and quantification of 5 randomly selected fields is shown on the right. Values are expressed as mean ± SEM. * indicates P < 0.05 and ** P < 0.01 when compared with scrambled controls. Data shown are representative of 3 independent experiments.

Discussion

Although metastasis is the overwhelming cause of death for lung cancer patients, the mechanism of lymphatic metastasis in lung cancer is currently unknown. Elucidation of the role of miRNAs in promoting or suppressing tumor metastasis in other tumor types has provided a new perspective on the metastatic process. In the present study, we found that miR-145 and miR-10b were dysregulated, which suggests that miRNAs may contribute to the metastasis of NSCLC. Furthermore, we revealed that the expression levels of miR-145 and miR-10b in primary lung cancer tissues were associated with lymph node metastasis, such that over-expression of miR-10b and down-regulation of miR-145 may be a common event in NSCLC metastasis. Our results suggest that miR-10b and miR-145 may act as an oncogene and tumor suppressor gene, respectively, in NSCLC metastasis.

Of the 4 metastasis-related miRNAs assessed in this study, miR-10b is the best characterized.^9,14 Recent studies have provided functional evidence that overexpression of miR-10b can contribute to the development of metastasis. Expression of miR-10b is induced by the transcription factor Twist in breast cancer and inhibits translation of the mRNA encoding homeobox D10 (HOXD10), resulting in increased expression of a well-characterized pro-metastatic gene, RHOC.¹⁴ In glioblastoma (GBM), the most common and malignant primary brain tumor, miR-10b promotes proliferation and prevents death of cancer cells by targeting cell cycle inhibitors and pro-apoptotic genes.¹⁵ KLF4, a known tumor suppressor gene that has been reported to suppress esophageal cancer cell migration and invasion, has been identified as a direct target of miR-10b in esophageal cancer.¹⁶ In our study, miR-10b was significantly upregulated in lymph node-positive NSCLC, and the expression level was correlated with lymph node metastasis status. In addition, the expression of miR-10b was up-regulated in metastatic lymph node tissues compared to primary lung cancer tissues, which suggests a crucial role of miR-10b in lung cancer cell metastasis. Further investigation is needed into the function of miR-10b in the metastasis of NSCLC.

The role of miR-145 in tumorigenesis has also been previously explored. MiR-145 functions in a cell-specific manner and is downregulated in tumor tissues. It is capable of inhibiting tumor cell growth and invasion by targeting several genes such as c-Myc and mucin 1.^10,11 In addition, miR-145 suppresses tumor growth by inhibition of multiple tumor survival effectors and is potentially useful in breast cancer therapy.¹⁷ In the present study, we demonstrated that miR-145 expression in primary NSCLC tissues was significantly correlated with lymph node metastasis status. In addition, the expression of miR-145 was lower in metastatic lymph node tissues than in the paired primary lung cancer tissues. These data along with our transwell assay results suggest that miR-145 plays a crucial role in negatively regulating lung cancer metastasis.

Human miR-9 has been reported to be involved in the metastasis of several malignancies, including brain and breast cancer.¹⁸ Ma et al. reported that miR-9 can increase cell motility and invasiveness through directly targeting CDH1 (the E-cadherin-encoding mRNA), and the expression level of miR-9 is correlated with MYCN amplification, tumor grade, and metastatic status in breast cancer.¹⁹ However, in our studies, miR-9 did not appear to be correlated with lymph node metastasis but instead with histology and stage in NSCLC. Further investigation is needed into the function of miR-9 in NSCLC.

MiR-155 is over-expressed in many types of cancer cells.^20-22 MiR-155 has been reported to be a candidate oncogenic miRNA that plays an important role in promoting HCC cell invasion. MiR-155 expression levels were high in tumor tissues in patients with post-OLT HCC recurrence compared with patients with non-recurrence. In addition, patients with higher miR-155 expression have significantly poorer recurrence-free survival and overall survival in hepatocellular carcinoma.²² However, we found no evidence for a role in NSCLC metastasis.

Although this research is limited by a small sample size and by being a single-institution study, our results suggest that the expression levels of miR-145 and miR-10b may have potential applications for more useful clinical stratification of NSCLC patients and enable selection of candidates for additional or alternative treatments.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Funding

This work was supported by grants from the National Natural Science Foundation of China (81172233,81372306), the Tianjin key project of Natural Science Foundation (12JCZDJC24400), Tianjin Natural Science Foundation (13JCYBJC22600), and Tianjin Science and Technology Support Program (12ZCDZSY16100). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.