ZAR1 is a novel epigenetically inactivated tumour suppressor in lung cancer

Abstract

Background: Lung cancer is the leading cause of cancer-related deaths with 1.8 million new cases each year and poor 5-year prognosis. Promoter hypermethylation of tumour suppressors leads to their inactivation and thereby can promote cancer development and progression.

Results: In this study, we analysed ZAR1 (zygote arrest 1), which has been said to be a maternal-effect gene and its expression mostly limited to certain reproductive tissues. Our study shows that ZAR1 is expressed in normal lung but inactivated by promoter methylation in lung cancer. ZAR1 is hypermethylated in primary lung cancer samples (22% small cell lung carcinoma (SCLC) and 76% non-small cell lung carcinoma (NSCLC), p < 0.001) vs. normal control lung tissue (11%). In lung cancer cell lines, ZAR1 was significantly methylated in 75% of SCLC and 83% of NSCLC vs. normal tissue (p < 0.005/0.05). In matching tumours and control tissues, we observed that NSCLC primary tumour samples exhibited a tumour-specific promoter methylation of ZAR1 in comparison to the normal control lung tissue. Demethylation treatment of various lung cancer cell lines reversed ZAR1 promoter hypermethylation and subsequently re-established ZAR1 expression. In addition, we could show the growth inhibitory potential of ZAR1 in lung cancer cell lines and cancer cell lines. Exogenous expression of ZAR1 not only inhibited colony formation but also blocked cell cycle progression of cancer cell lines.

Conclusions: Our study shows for the first time the lung tumour-specific epigenetic inactivation of ZAR1 due to DNA methylation of its CpG island promoter. Furthermore, ZAR1 was characterised by the ability to block tumour growth through the inhibition of cell cycle progression in cancer cell lines. We propose that ZAR1 could serve as an epigenetically inactivated biomarker in lung cancer.

Keywords: DNA methylation; Epigenetics; Lung cancer; Tumour suppressor; ZAR1 (zygote arrest 1).

Fig. 1

ZAR1 expression pattern, missing expression in lung cancer and promoter hypermethylation in lung cancer cell lines. ZAR1 expression in lung cancer cell lines A427, A549, HTB171, H322, H358 and HeLa (cervix carcinoma) versus normal tissue panel (breast, kidney, liver, lung and heart) by RT-PCR and semi-/quantitative PCR were determined. a, b Expression levels were normalised to ß-ACTIN (ACTB) and to normal lung tissue (=100%). Promoter hypermethylation of lung cancer cell lines, HEK and HeLa cells by COBRA analysis c and pyrosequencing for quantification d are shown. PCR and digestion products were separated on 2% TBE agarose gel with 100 bp marker M (m methylated, pm partially methylated, pos. positive control, neg. negative control; + TaqI digested, − mock digested)

Fig. 2

ZAR1 promoter hypermethylation in lung cancer cell lines (SCLC). ZAR1 promoter methylation status was analysed by COBRA assay in various SCLC lung cancer cell lines. CpG island region from the ZAR1 promoter was amplified from bisulfite-treated DNA and digested with Taq1. Digestion products indicate ZAR1 promoter methylation. Digestion products were separated on 2% TBE agarose gel with 100 bp marker (m methylated, um unmethylated, pos. positive control, + TaqI-digested, − mock digested)

Fig. 3

ZAR1 promoter hypermethylation in primary lung tumours and summary of ZAR1 methylation. ZAR1 promoter methylation was studied by COBRA assay in primary lung tumours of SCLC and NSCLC samples and in normal control tissue. SCLC tumour samples were abbreviated SC and control tissue N (a). NSCLC tumour samples were abbreviated LT (lung tumours) and control tissue LN. LN1, LN3 and LN39 were matching samples to tumour samples LT1, LT3 and LT39 (b). CpG island region from the ZAR1 promoter was amplified from bisulfite-treated DNA and digested with TaqI. Digestion products indicate ZAR1 promoter methylation. Digestion products were separated on 2% TBE agarose gel with 100 bp marker (m methylated, um unmethylated, pos. positive control, + TaqI-digested, − mock digested). c Summary of ZAR1 promoter methylation in lung cancer cell lines and primary lung tumours is shown (Fisher exact test)

Fig. 4

Reversal of ZAR1 epigenetic inactivation in cancer cells by demethylation treatment. Various lung cancer cell lines as well as HEK and HeLa cells were treated with the demethylating agent 5-Aza-2’-deoxycytidine (0, 5 or 10 μM Aza). COBRA assay under Aza treatment shows demethylation of the ZAR1 promoter (a) and correlating reexpression of ZAR1 levels by semi-quantitative RT-PCR (b). PCR and digestion products were separated on 2% TBE agarose gel with 100 bp marker (m methylated, pm partially methylated, um unmethylated, pos. positive control, + TaqI-digested, − mock digested). c Real-time PCR quantification of ZAR1 reexpression after 5 μM Aza treatment, which was normalised to GAPDH. d Luciferase reporter assay was performed to determine the ZAR1 promoter activity. The artificial ZAR1 promoter was cloned into pRLnull and subsequently in vitro methylated. The promoters were transfected in HeLa cells and the ivm ZAR1 promoter showed full inactivation in contrast to its unmethylated construct and empty control construct. Student’s t test was used for statistical analysis

Fig. 5

ZAR1 inhibits colony formation of lung cancer cell lines. ZAR1 was overexpressed in lung cancer cell lines A549, A427 and HTB171 and ZAR1 inhibited colony formation. Cells were transfected with ZAR1-EGFP or EGFP-vector control. Colonies were selected for 3 weeks with G418 and then Giemsa stained (a, upper panel). The according ZAR1 overexpression was verified by semi-quantitative RT-PCR (a, lower panel). Inhibition of colony formation by ZAR1 was quantified and showed significant (Student’s t test) reduction in colony numbers (b)

Fig. 6

ZAR1 functions as a tumour suppressor in cancer cell lines through cell cycle arrest. a, b Localisation of overexpressed ZAR1 is exemplarily shown in A549 and HeLa with DAPI nuclear stain (×63 magnification). c–e Flow cytometry analysis found that overexpressed ZAR1 arrests the cell cycle in HeLa and HCT116 cancer cells. ZAR1 was overexpressed in cancer cells, cells were ethanol-fixed after 24 h (HeLa) and 48 h (HCT116) and propidium iodide staining was used to determine the DNA content for flow cytometry. The cell cycle distribution of HeLa and HCT116 cells upon ZAR1 reexpression is quantified (c, e), and gating is exemplarily depicted in HCT116 (d)