GAS5 long non-coding RNA in malignant pleural mesothelioma

Abstract

Background: Malignant pleural mesothelioma (MPM) is an aggressive cancer with short overall survival. Long non-coding RNAs (lncRNA) are a class of RNAs more than 200 nucleotides long that do not code for protein and are part of the 90% of the human genome that is transcribed. Earlier experimental studies in mice showed GAS5 (growth arrest specific transcript 5) gene deletion in asbestos driven mesothelioma. GAS5 encodes for a lncRNA whose function is not well known, but it has been shown to act as glucocorticoid receptor decoy and microRNA "sponge". Our aim was to investigate the possible role of the GAS5 in the growth of MPM.

Methods: Primary MPM cultures grown in serum-free condition in 3% oxygen or MPM cell lines grown in serum-containing medium were used to investigate the modulation of GAS5 by growth arrest after inhibition of Hedgehog or PI3K/mTOR signalling. Cell cycle length was determined by EdU incorporation assay in doxycycline inducible short hairpinGAS5 clones generated from ZL55SPT cells. Gene expression was quantified by quantitative PCR. To investigate the GAS5 promoter, a 0.77 kb sequence was inserted into a pGL3 reporter vector and luciferase activity was determined after transfection into MPM cells. Localization of GAS5 lncRNA was identified by in situ hybridization. To characterize cells expressing GAS5, expression of podoplanin and Ki-67 was assessed by immunohistochemistry.

Results: GAS5 expression was lower in MPM cell lines compared to normal mesothelial cells. GAS5 was upregulated upon growth arrest induced by inhibition of Hedgehog and PI3K/mTOR signalling in in vitro MPM models. The increase in GAS5 lncRNA was accompanied by increased promoter activity. Silencing of GAS5 increased the expression of glucocorticoid responsive genes glucocorticoid inducible leucine-zipper and serum/glucocorticoid-regulated kinase-1 and shortened the length of the cell cycle. Drug induced growth arrest was associated with GAS5 accumulation in the nuclei. GAS5 was abundant in tumoral quiescent cells and it was correlated to podoplanin expression.

Conclusions: The observations that GAS5 levels modify cell proliferation in vitro, and that GAS5 expression in MPM tissue is associated with cell quiescence and podoplanin expression support a role of GAS5 in MPM biology.

Figure 1

GAS5 expression is lower in MPM and it is increased upon drug induced growth arrest. A. GAS5 is lower in MPM cell lines compared to normal mesothelial cell lines. GAS5 lncRNA expression was analysed by qRT-PCR in 22 MPM cell lines and in 7 normal mesothelial cell lines. Expression of GAS5 was normalized to internal control histones relative to the mean expression of GAS5 in normal mesothelial cells in culture according to –∆∆Ct method. B and C. Drug induced growth arrest increases GAS5 expression in MPM primary cells (ZL55SPT and SDM103T2) and cell lines (ZL55 and ACC-Meso4). D. Expression of GAS5 promoter reporter gene was analysed in control versus HhAntag and NVP-BEZ235 treated samples. Promoter activity, expressed as relative light unit (RLU) normalized to control set at 100%, is significantly increased by treatment with HhAntag and NVP-BEZ235. Values are expressed as mean ± SD from three independent experiments *; p < 0.05.

Figure 2

Silencing of GAS5 increases glucocorticoid responsive genes. A. GAS5 lncRNA silencing increases the expression of glucocorticoid responsive genes GILZ (B) and SGK1 (C). Values are normalized with histones and shown as mean ± SD from three to four independent experiments. *; p < 0.05.

Figure 3

Silencing of GAS5 shortens the cell cycle length as determined by EdU incorporation assay. A. Representative micrographs of shGAS5 cells pre-treated with or without doxycycline during 72 h then incubated with EdU for 12 hrs. Scale bar: 100 μm. B. Cell cycle length was measured from a time course of EdU incorporation. Five different fields were imaged for each group in each of three independent experiments under fluorescence microscope and number of EdU positive cells was counted in approximately 1200 cells (DAPI). The graph plots the percentage of EdU positive cells vs. time in axes. The time taken for 100% of EdU positive cells was determined from the linear regression and slopes in presence or absence of doxycycline were compared p < 0.0001. C. Time in hours needed for 30, 50, 75 and 100% of cells to become EdU positive calculated according to the linear regression method.

Figure 4

Nuclear enrichment of GAS5 lncRNA after induction of growth arrest by HhAntag. A. Left panel: Representative micrographs of fluorescence in situ hybridization showing GAS5 (yellow), GAPDH (green) for cytosolic control, RNU (red) for nuclear mRNA control, and nuclei stained with DAPI (blue). Right panel: same image as left panel showing only RNU and GAS5. Scale bar: 10 μm. B. Quantification of GAS5 in ZL55SPT cells and nuclei was determined by calculating intensity of fluorescent spots by Imaris image analysis using spot detection method. The graph plots the fluorescent spots per cell vs. intensity. *; p < 0.05 compared to the non-treated control. C. Integration of area under the curve to evaluate the proportion of nuclear GAS5 in control and HhAntag-induced growth arrest. Representative of three independent experiments.

Figure 5

GAS5 expression is higher in podoplanin positive MPM tumor cells. A. Fluorescence in situ hybridization and immunohistochemical analysis of GAS5, Ki-67 and podoplanin in a representative patient. Upper panel shows fluorescence in situ hybridization of control and GAS5, immunostaining of Ki-67 (ab15580, Abcam). Lower panel shows the immunostaining of control and podoplanin (D2-40, DAKOCytomation) and hematoxylin and eosin stained tumor tissue. Scale bar: 50 μm. B. Matrix of relative gene expression in epitheliod, biphasic and sarcomatoid mesothelioma shown as heat map. Heat map is a grid of rectangles with colors that indicate the value of the matrix elements, where high expression is blue and low expression is red. Rows of each heat map correspond to genes, whilst columns correspond to MPM tumor samples. C. GAS5 expression is higher in MPM tissue expressing higher levels of podoplanin *p < 0.05, Mann–Whitney test.