FUS-DDIT3 prevents the development of adipocytic precursors in liposarcoma by repressing PPARgamma and C/EBPalpha and activating eIF4E

Abstract

Background: FUS-DDIT3 is a chimeric protein generated by the most common chromosomal translocation t(12;16)(q13;p11) linked to liposarcomas, which are characterized by the accumulation of early adipocytic precursors. Current studies indicate that FUS-DDIT3- liposarcoma develops from uncommitted progenitors. However, the precise mechanism whereby FUS-DDIT3 contributes to the differentiation arrest remains to be elucidated.

Methodology/principal findings: Here we have characterized the adipocyte regulatory protein network in liposarcomas of FUS-DITT3 transgenic mice and showed that PPARgamma2 and C/EBPalpha expression was altered. Consistent with in vivo data, FUS-DDIT3 MEFs and human liposarcoma cell lines showed a similar downregulation of both PPARgamma2 and C/EBPalpha expression. Complementation studies with PPARgamma but not C/EBPalpha rescued the differentiation block in committed adipocytic precursors expressing FUS-DDIT3. Our results further show that FUS-DDIT3 interferes with the control of initiation of translation by upregulation of the eukaryotic translation initiation factors eIF2 and eIF4E both in FUS-DDIT3 mice and human liposarcomas cell lines, explaining the shift towards the truncated p30 isoform of C/EBPalpha in liposarcomas. Suppression of the FUS-DDIT3 transgene did rescue this adipocyte differentiation block. Moreover, eIF4E was also strongly upregulated in normal adipose tissue of FUS-DDIT3 transgenic mice, suggesting that overexpression of eIF4E may be a primary event in the initiation of liposarcomas. Reporter assays showed FUS-DDIT3 is involved in the upregulation of eIF4E in liposarcomas and that both domains of the fusion protein are required for affecting eIF4E expression.

Conclusions/significance: Taken together, this study provides evidence of the molecular mechanisms involve in the disruption of normal adipocyte differentiation program in liposarcoma harbouring the chimeric gene FUS-DDIT3.

Figure 1. Adipogenic gene expression in liposarcomas of FUS-DDIT3 transgenic mice and in human liposarcoma cell lines carrying the translocation t(12;16)(q13;p11).

(A) Hematoxylin/eosin stained sections showing the presence of lipoblasts with round nuclei and accumulation of intracellular lipid in a liposarcoma arisen in the chest region of FUS-DDIT3 mouse (10× and 40× magnifications are shown). (B) Western blot analyses of regulators of adipocyte function in white adipose tissue (WAT), liposarcoma arisen in FUS-DDIT3 transgenic mice and human liposarcomas cell lines expressing FUS-DDIT3 (LIS-3 and LIS-4). Cell and tissue extracts (10 μg) were resolved in SDS-PAGE gel (10% acrylamide), followed by immunoblotting analysis with anti-C/EBPβ, anti-C/EBPδ, anti-PPARγ, anti-C/EBPα and anti-actin antibodies. These data are representative of three independent experiments. (C) Western blot analysis of fat cell markers such as aP2 and adiponectin in liposarcomas of FUS-DDIT3 transgenic mice and in human liposarcoma cell lines carrying the translocation t(12;16)(q13;p11). These data are representative of three independent experiments. (D) Expression of the human FUS-DDIT3 oncogene by RT-PCR both in liposarcomas of FUS-DDIT3 transgenic mice and in human liposarcoma cell lines carrying the translocation t(12;16)(q13;p11).

Figure 2. CombitTA-FUS-DDIT3 expression and effect of FUS-DDIT3 on adipocyte differentiaton.

A) Analysis of the tetracycline (Doxycycline) dependent CombitTA-FUS-DDIT3 expression by RT-PCR in the presence (+tet) or in the absence (-tet) of doxycycline in MEF (the time of treatment with doxycycline was 48 hours). Actin was used to check the RNA integrity and loading. B) Adipocyte differentiation in CombitTA-FUS-DDIT3 MEFs after suppression of FUS-DDIT3 expression by tetracycline treatment. CombitTA-FUS-DDIT3 MEFs in the presence (+tet) or in the absence (-tet) of doxycycline were cultered up to confluence and grown in the presence of standard adipose differentiation induction medium. At day 8 after induction of adipocyte differentiation, cells were fixed and stained for neutral lipids with Oil-Red-O and the morphological differentiation is shown (the original magnification is ×20). This experiment was repeated three times using cells prepared from different embryos and similar results were obtained. C) Western blot analyses of PPARγ, and C/EBPα in liposarcoma arisen in CombitTA-FUS-DDIT3 mice in the presence (+tet) or in the absence (−tet) of doxycycline. Doxycycline was given at 4 mg/mL for 4 weeks.

Figure 3. Retroviral-mediated expression of PPARγ2 rescues the impaired adipogenesis of FUS-DDIT3 MEFs.

A) FUS-DDIT3 MEFs were infected with either control retroviral vector or one expressing PPARγ2 (pQCXIP-PPARγ2) and selected for 3 days with 2 μg/ml puromycin. Then, wild-type MEF, FUS-DDIT3 MEF and PPARγ2 expressing FUS-DDIT3-MEF were cultered up to confluence and grown in the presence of standard adipose differentiation induction medium. At day 8 after induction of adipocyte differentiation, cells were fixed and stained for neutral lipids with Oil-Red-O and the morphological differentiation is shown (the original magnification is ×20). This experiment was repeated three times using cells prepared from all lines and from different embryos and similar results were obtained. B) Analysis of the PPARγ2 protein by western-blot in FUS-DDIT3 MEFs infected with either a control retroviral vector (pQCXIP) or one expressing PPARγ2 (pQCXIP- PPARγ2) 4 days after infection.

Figure 4. FUS-DDIT3 represses the PPARγ2 promoter.

A 1 kb proximal promoter region of human PPARγ2 was previously shown to be sufficient to drive the PPARγ2′s expression in reporter assays [34. 35] and it is active in U2OS cells when co-transfected with C/EBPβ expression vectors. To directly assess the ability of FUS-DDIT3 to modulate transcription from DNA sequences present in the PPARγ2 promoter, an expression vector containing either the human FUS-DDIT3 cDNA, the human DDIT3 domain or the human FUS domain were co-transfected into U2OS cells along with the reporter vector containing the PPARγ2 promoter (pGL3-hPPARγ2p1000 vector) and C/EBPβ expression vector (ratC/EBPβ wtpSG5). Luciferase reporter assays demonstrate that FUS-DDIT3 repressed the human PPARg2 reporter in a DDIT3·dependent manner. In all lines, 1 μg of pRL-SV40 (Renilla basal control (PROMEGA) was used for normalization of the results along with 5 μg of pGL3-hPPARγ2p1000 (lines 2–10); 3 μg of ratC/EBPβ wtpSG5 (lines 3–10); 3, 5 and 7 μg of the hFUS-DDIT3 expression vector (lines 4–6, respectively); 3 and 7 μg of the hDDIT3 expression vector (lines 7–8, respectively); 3 and 7 μg the NH2-hFUS expression vector (lines 9–10, respectively); 5 μg of the hFUS-DDIT3 expression vector (line 11). These data are representative of three independent experiments.

Figure 5. C/EBPα does not bypass adipogenesis blockade in FUS-DDIT3 expressing-MEF.

(A) FUS-DDIT3 represses the C/EBPα transactivation induced by C/EBPβ. U2OS cells were cotransfected with 1 μg of pRL-SV40 (Renilla basal control (PROMEGA), lines 1–6) along with: 5 μg of pCEBP1171 (luciferase reporter vector containing 1171 of the rat C/EBPα promoter, samples 2–6); 3 μg ratC/EBPβwtpSG5 (C/EBPβ expressing vector, lines 3–6); 3, 5 and 7 μg pcDNA3-hFUS-DDIT3 (hFUS-DDIT3 expression vector, lines 4–6); 5 μg of the hFUS-DDIT3 expression vector (line 7). These data are representative of three independent experiments. (B) Retroviral expression of C/EBPα does not rescue the adipocyte differentiation blockade in FUS-DDIT3 MEFs. FUS-DDIT3 MEFs were infected with a retroviral vector expressing C/EBPα (pQCXIP-C/EBPα) and selected for 3 days with 2 μg/ml puromycin. Then, wild-type MEF, FUS-DDIT3 MEF and C/EBPα expressing FUS-DDIT3-MEFS were cultered up to confluence and grown in the presence of standard adipose differentiation induction medium. At day 8 after induction of adipocyte differentiation, cells were fixed and stained for neutral lipids with Oil-Red-O and the morphological differentiation is shown (the original magnification is ×20). This experiment was repeated three times using cells prepared from all lines and from different embryos and similar results were obtained. (C) Analysis of C/EBPα (p42-C/EBPα and p30-C/EBPα isoforms) protein expression by western-blot in FUS-DDIT3 MEFs infected with either a control retroviral vector (pQCXIP) or one expressing C/EBPα (pQCXIP- C/EBPα) 4 days after infection.

Figure 6. FUS-DDIT3 upregulates eIF2α and eIF4E.

(A) Western blot analyses of eIF4E and eIF2α expression in wild-type white adipose tissue (WT-WAT), liposarcoma arisen in FUS-DDIT3 mice (Tumor), normal WAT from FUS-DDIT3 mice (FD-WAT), human liposarcomas cell lines expressing FUS-DDIT3 (LIS-3 and LIS-4), human adipose cells (Zen-bio), and in liposarcoma arisen in CombitTA-FUS-DDIT3 mice in the presence (+tet) or in the absence (−tet) of doxycycline (doxycycline was given at 4 mg/mL for 4 weeks). Cell and tissue extract (10 μg) were resolved in SDS-PAGE gel (10% acrylamide), followed by immunoblotting analysis with anti- eIF4E, anti- eIF2α and anti-actin antibodies. (B) Transactivation of the CAT reporter gene linked to mouse eIF4E promoter by FUS-DDIT3. C3H10T1/2 cells were transiently cotransfected with 1 μg of pm4ECAT (CAT reporter vector containing ∼2.5 kb of the mouse eIF4E promoter) together with 5 μg of pcDNA (empty vector, panel 2) or with 5 μg of pcDNA3-hFUS-DDIT3 (hFUS-DDIT3 expression vector, panel 3). The data represent the fold activation with respect to a sample where reporter alone was transfected (panel 1). Data represent an average obtained from three separated experiments. A representative thin layer chromatograph is shown on the right. (C) Transactivation of the CAT reporter gene linked to mouse eIF4E promoter by FUS and DDIT3 domains of FUS-DDIT3 fusion protein. C3H10T1/2 cells were transiently cotransfected with 1 μg of pm4ECAT together with 5 μg of a vector expressing the FUS domain (panel 2) or with 5 μg of a vector expressing the DDIT3 domain (panel 3). The data represent the fold activation with respect to a sample where reporter alone was transfected (panel 1). Data represent an average obtained from three separated experiments.

Figure 7. Model for the adipocyte differentiation arrest produced by FUS-DDIT3 in liposarcoma development.

(A) Scheme of the normal differentiation program in mesenchymal progenitor cells. (B) FUS-DDIT3 blocks the adipocyte differentiation program in mesenchymal cell progenitors by interfering with the PPARγ and C/EBPα activities at the transcriptional level. In addition, FUS-DDIT3 induces the expression of eIF4E, that in turns, is able to inactivate the C/EBPα pathway by shifting the normal isoform ratio towards the truncated p30- C/EBPα isoform, which has a negative effect on adipogenesis.