Tumorigenic role of orphan nuclear receptor NR0B1 in lung adenocarcinoma

Abstract

Cancer stem cells are a limited population of tumor cells that are thought to reconstitute whole tumors. The Hoechst dye exclusion assay revealed that tumors are composed of both a main population and a side population of cells, which are rich in cancer stem cells. NR0B1 is an orphan nuclear receptor that is expressed to a greater extent in the side population, as compared with the main population, of a lung adenocarcinoma cell line. In this study, we investigated the role of NR0B1 in lung adenocarcinoma cells. Reduction of NR0B1 expression levels in lung adenocarcinoma cell lines resulted in vulnerability to anti-cancer drugs and decreased abilities for invasion, in vitro colony formation, and tumorigenicity in non-obese diabetic/severe compromised immunodeficient mice. When 193 cases of lung adenocarcinoma were immunohistochemically examined, higher levels of NR0B1 expression correlated with higher rates of lymph node metastasis and recurrence. Multivariate analysis revealed high NR0B1 expression levels, stage of the disease, and size of tumor to be independent unfavorable prognostic factors for overall and disease-free survival rates. In clinical samples, NR0B1 expression levels inversely correlated to the proportion of methylated CpG sequences around the transcription initiation site of the NR0B1 gene, suggesting the epigenetic control of NR0B1 transcription in lung adenocarcinoma. In conclusion, NR0B1 might play a role in the malignant potential of lung adenocarcinoma.

Figure 1

High NR0B1 expression in SP of A549. A: Dot blot analysis of A549 cells stained with Hoechst 33342 dye in the absence (left) or presence (right) of verapamil. SP and MP cells were boxed. B: Quantitative real-time RT-PCR was performed with mRNA obtained from SP and MP of A549. The amount of NR0B1 mRNA was normalized for the amount of GAPDH mRNA. The values represent the mean ± SE of three experiments. *P < 0.01 by the Student’s t-test.

Figure 2

Effect of NR0B1 expression on the proliferative activity and the SP formation. A: Immunoblot analysis for the amount of NR0B1 protein in the control and si-RNA knocked down cells. B: Comparison of cell proliferation of NR0B1 knocked down cells to that of control cells. The values represent the mean ± SE of three experiments. C: Dot blot analysis of control (upper) and knocked down (lower) cells stained with Hoechst 33342 dye in the absence (left) or presence (right) of verapamil. SP cells were boxed. D: Proportion of SP in the control and knocked down cells.

Figure 3

Effect of NR0B1 expression on the resistance against topotecan and the cell invasion activity. A: Viability in the presence of various amount of topotecan was compared. B: Matrigel invasion assay. Cells invading through Matrigel are shown. Original magnification, ×100. C: Invading cell number per mm². The values shown in (A) and (C) are the mean ± SE of three experiments. *P < 0.01 by the Student’s t-test.

Figure 4

Effect of NR0B1 on the in vitro colony formation and the in vivo tumorigenicity activities. A: Colonies derived from the control and the NR0B1 knocked down cells. Original magnification, ×40. B: Number of colonies per cm². C: Tumors in NOD/SCID mice at the injection sites of the control and the knocked down cells. D: The change of the volume of tumors. E: Histology of tumors derived from the control and the knocked down cells. Original magnification, ×400. The values in (B) and (D) represent the mean ± SE of three experiments. *P < 0.01 by the Student’s t-test.

Figure 5

Effect of NR0B1 in PC-14. A: Immunoblot analysis for the amount of NR0B1 protein in the control and si-RNA knocked down PC-14 cells. B: Matrigel invasion assay using PC-14 cells. Cells invading through Matrigel were shown. Original magnification, ×100. C: Invading cell number per mm². D: Viability of PC-14 cells in the presence of topotecan was compared. E: Colonies derived from the control and the NR0B1 knocked down PC-14 cells, and the number of colonies per cm². The values in (C), (D), and (E) represent the mean ± SE of three experiments. *P < 0.01 by the Student’s t-test.

Figure 6

Expression of apoptosis- and invasion-related genes in NR0B1-knocked down A549 cells. A: Semiquantitative RT-PCR. B: Real-time quantitative RT-PCR to examine the expression level of Bcl-2 and MMP-2. The value is shown as the mean ± SE. *P < 0.01 by the Student’s t-test.

Figure 7

Effect of methylation on NR0B1 promoter activity. A: CpG sequences in the NR0B1 promoter. The vertical line represents a single CpG site. Transcription initiation site (TIS) is shown as arrow. B: Digestion patterns with the methylation-sensitive restriction enzyme HpaII of methylated and mock-methylated reporter plasmids. The luciferase reporter plasmid containing NR0B1 promoter was methylated or mock-methylated in vitro. Subsequently, each construct was digested with HpaII and electrophoresed. C: Relative luciferase activities of the methylated and mock-methylated NR0B1 constructs. The values represent the mean ± SE of three experiments. *P < 0.01 by the Student’s t-test. The SE of the methylated construct was too small to be shown.

Figure 8

Inverse correlation of NR0B1 expression level to the proportion of methylated CpG sequences in clinical samples of lung adenocarcinoma. A: Quantitative real-time RT-PCR was performed with mRNA obtained from tumors. The amount of NR0B1 mRNA was normalized for the amount of GAPDH mRNA. The highest four and the lowest four cases were shown (cases 1 to 4, and cases 5 to 8, respectively). B: Methylation status of CpG sequences in NR0B1 promoter. Six clones for each population were analyzed by bisulfite sequencing. Open and closed squares denote unmethylated and methylated cytosines, respectively. C: Methylation status of CpG sequences in NR0B1 promoter of normal lung tissue from two cases.

Figure 9

NR0B1 expression in clinical samples of lung adenocarcinoma. A–E: Representative pictures of NR0B1 immunostaining in various cases. (×400) NR0B1 signals were detected in the nuclei of approximately 90% (A), 70% (B), 30% (C) of tumor cells. In some cases, the signals were detected not only in the nucleus but also in the cytoplasm (D). In other cases, few NR0B1-positive tumor cells were detected (E, <1%). F: Quantitative real-time RT-PCR was performed with mRNA obtained from 11 tumors as described in Figure 8A. The amount of NR0B1 mRNA was normalized for the amount of GAPDH mRNA. The amount of NR0B1 mRNA was higher in immunohistochemically defined NR0B1-hi cases than in NR0B1-lo cases. The bar shows mean value of NR0B1 amount.

Figure 10

Relation of NR0B1 expression to prognosis of clinical cases. A: OS in 193 lung adenocarcinoma cases. Five-year OS in all patients, patients with stage I, II, and III was 84.8%, 95.1%, 52.4%, and 53.8%, respectively. B: Five year DFS in all patients, patients with stage I, II, and III was 73.1%, 87.9%, 30.5%, and 23.4%, respectively. OS (C) and DFS (D) curves in NR0B1-hi and NR0B1-lo cases.