A novel oncogenic pathway by TLS–CHOP involving repression of MDA-7/IL-24 expression

Background: Translocated in liposarcoma-CCAAT/enhancer binding protein homologous protein (TLS–CHOP) (also known as FUS-DDIT3) chimeric oncoprotein is found in the majority of human myxoid liposarcoma (MLS), but its molecular function remains unclear. Methods: We knockdowned TLS–CHOP expression in MLS-derived cell lines by a specific small interfering RNA, and analysed the gene expression profiles with microarray. Results: TLS-CHOP knockdown inhibited growth of MLS cells, and induced an anticancer cytokine, melanoma differentiation-associated gene 7 (MDA-7)/interleukin-24 (IL-24) expression. However, double knockdown of TLS–CHOP and MDA-7/IL-24 did not inhibit MLS cell growth. Conclusion: Repression of MDA-7/IL-24 expression by TLS–CHOP is required for MLS tumour growth, and TLS–CHOP may become a promising therapeutic target for MLS treatment.

More than 90% of human myxoid liposarcoma (MLS) cases are associated with the chromosomal translocation, which creates a chimeric oncogene comprising part of the TLS (translocated in liposarcoma) gene (also known as FUS (fused in Ewing's sarcoma)) and part of the CHOP (CCAAT/enhancer binding protein (C/EBP) homologous protein) gene (also called DDIT3 (DNA damage-inducible transcript 3) and GADD153 (growth arrest-and DNA damage-inducible gene 153)) (Crozat et al, 1993;Rabbitts et al, 1993;Powers et al, 2010). The resultant fusion gene TLS-CHOP encodes the N-terminal half of TLS fused to complete sequence of CHOP (Powers et al, 2010; Figure 1A). TLS-CHOP protein is considered to function as an abnormal transcription factor (Kuroda et al, 1999;Pérez-Mancera et al, 2008;Andersson et al, 2010). The definitive TLS-CHOP function for MLS development, however, is unclear.
In this report, we have found a novel pathway of TLS-CHOP with MDA-7/IL-24 repression.

Cell culture
The MLS-derived cell lines, 1955/91 and 2645/94, were kindly provided from Professor David Ron (University of Cambridge), and were cultured in Dulbecco's modified Eagle's medium (Sigma-Aldrich Corporation, St Louis, MO, USA) supplemented with 10% foetal bovine serum. Cell quantification was performed as previously described (Oikawa et al, 2004).

Microarray analysis
Cells were transfected with the TLS-CHOP or negative control siRNAs, and were incubated for 72 h. Biotin-labelled complementary RNA (cRNA) was then generated from 1 mg of total RNA of the cells using CodeLink iExpress Expression Assay Reagent Kit (GE Healthcare UK Ltd, Buckinghamshire, UK), and was hybridised to CodeLink Human Whole Genome Bioarray (GE Healthcare) using iAmplify cRNA Preparation and Hybridisation Reagents Kit (GE Healthcare) according to Expression Bioarray System User Guide ver. 2.0. The array slides were incubated for 21 h at 37 1C with shaking, and were scanned with a DNA microarray scanner G2505A (Agilent Technologies, Inc., Santa Clara, CA, USA). The scanned images were analysed and median normalised using CodeLink Expression Analysis Version 4.1.0.29054 (GE Healthcare). The data have been deposited in NCBI's Gene Expression Omnibus and are accessible through GEO Series accession number GSE33616.

TLS-CHOP knockdown represses cell growth of MLS-derived cell lines
First, we examined the activity of the three newly designed effective siRNAs that target different positions of TLS-CHOP in a preliminary experiment (Supplementary Figure 1), and selected the most effective siRNA among them (hereafter termed TLS-CHOP siRNA) for use in subsequent experiments. The TLS-CHOP  Only the probes showing over twofold change in both two cell lines are listed. 'NULL' in Gene symbol column means that the probe sequence is not entried in NCBI database. siRNA targets exon 2 of the CHOP gene ( Figure 1A). Although types 4 and 11 of TLS-CHOP variants do not have the target region, TLS-CHOP in over 80% of MLS is type 1 or 2. We confirmed that the two MLS-derived cell lines, 1955/91 and 2645/ 94, carries type 1 and type 2, respectively ( Figure 1B). TLS-CHOP knockdown by the siRNA inhibited cell growth and induced cell death in both cell lines ( Figure 1C-F). On the other hand, a nontargeting negative control siRNA did not affect cell growth, indicating that the effects of TLS-CHOP siRNA are not by offtarget effects.

TLS-CHOP knockdown induces MDA-7/IL-24 expression in MLS cells
Next, we compared mRNA expression profiles of both 1955/91 and 2645/94 cells transfected with TLS-CHOP siRNA or negative control siRNA by microarray analysis (see Materials and Methods). We found that several dozen genes showed at least two-fold differential expression by TLS-CHOP siRNA (Table 1). Among the genes, we focused on the MDA-7/IL-24 gene because it encodes an anticancer cytokine . TLS-CHOP siRNA induced a significant increase in the expression of MDA-7/IL-24 in both cell lines (Table 1; Figure 2B). Thus, to confirm that MDA-7/IL-24 is important for growth arrest by TLS-CHOP knockdown, we prepared MDA-7/IL-24 siRNA and performed double transfection with both TLS-CHOP and MDA-7/IL-24 siRNAs into 1955/91 cells. MDA-7/IL-24 knockdown cancelled the growth inhibitory effects by TLS-CHOP siRNA alone (Figure 2A and B).

DISCUSSION
We have demonstrated that TLS-CHOP knockdown in MLS cells represses cell growth ( Figure 1C-E), suggesting that TLS-CHOP plays an essential role for growth of MLS cells. Furthermore, our results suggest that TLS-CHOP may become a promising molecular target for MLS treatment.
TLS-CHOP knockdown in MLS cells induced increased expression of an anticancer cytokine MDA-7/IL-24 (Table 1; Figure 2B). Thus, we consider that although the cancerous characteristics of MLS cells have potential to induce MDA-7/IL-24 expression, TLS-CHOP represses it and contributes to maintain the tumour growth.
In conclusion, we have revealed a novel pathway involving repression of MDA-7/IL-24 expression for tumourigenesis and/or growth of MLS. We believe that our results will contribute understanding of molecular function of the chimeric oncoprotein and development of a novel molecular therapy for cancers.