The fusion protein SS18-SSX1 employs core Wnt pathway transcription factors to induce a partial Wnt signature in synovial sarcoma

Abstract

Expression of the SS18/SYT-SSX fusion protein is believed to underlie the pathogenesis of synovial sarcoma (SS). Recent evidence suggests that deregulation of the Wnt pathway may play an important role in SS but the mechanisms whereby SS18-SSX might affect Wnt signaling remain to be elucidated. Here, we show that SS18/SSX tightly regulates the elevated expression of the key Wnt target AXIN2 in primary SS. SS18-SSX is shown to interact with TCF/LEF, TLE and HDAC but not β-catenin in vivo and to induce Wnt target gene expression by forming a complex containing promoter-bound TCF/LEF and HDAC but lacking β-catenin. Our observations provide a tumor-specific mechanistic basis for Wnt target gene induction in SS that can occur in the absence of Wnt ligand stimulation.

Figure 1. SS18-SSX regulates mRNA levels in cell context-specific manner.

(A) Q -PCR analysis of AXIN2 in primary human SS, satellite cells, myoblasts and MSC. (B) Q-PCR analysis of AXIN2 and SS18-SSX (indicated) after SS18-SSX depletion in SS cells and SS cell lines. (C,D) Q-PCR analysis of AXIN2 transcripts and SS18-SSX (indicated) in C3H10T1/2 cells. In (C) cells were infected with V5-tagged SS18-SSX1 (SS-V5) or empty pLIVc vector and selected for 10 days in 1 μg/ml puromycin. In (D) selected C3H10T1/2^pLIVc and C3H10T1/2^SS18-SSX1-V5 cells were infected with LV-CRE pLKO.1 or an empty pLKO.1 vector and harvested 96 hours later. Protein expression and knockdown were assessed by Western blot analysis (inset in (C) and lowest panel in D) using mouse anti-V5 and HRP-conjugated goat anti-mouse IgG. Monoclonal mouse anti-tubulin antibody provided the loading control. Q-PCR results are representative of two (B) or three (D) independent experiments. Bar represents the SD of triplicate PCRs; in B the significance (indicated by asterisks) of SS18-SSX and AXIN2 repression was p = 0.014 and p = 0.022, respectively, for SS11, p = 0.0130 and p = 1.03E-5 for SS12, p = 0.000738 and 6.74E-5 for HS-SYII and p = 0.000297 and 0.0103 for FUJI. (E) The effect of SS18-SSX1 expression on mouse AXIN2 promoter activity was measured in STO cells stably expressing V5-tagged SS18-SSX1 protein or empty pLIVc, and in wild type C3H10T1/2, C3H10T1/2^pLIVc, C3H10T1/2^SS18-SSX1-V5 and C3H10T1/2^SS18-SSX1-HA cells. The ratio of firefly to renilla luciferase activity is reported. Results are representative of three independent experiments. Error bars represent the SD of triplicate tests.

Figure 2. SS18-SSX1-dependent AXIN2 promoter activity does not involve the LEF-1 β–catenin activation complex.

(A) (upper panel) Fluorescence microscopy images of C3H10T1/2^pLIVc (a) and C3H10T1/2^SS18-SSX1-V5 (b) cells infected with a lentiviral vector expressing both ΔNLEF-1 and RFP, 200X magnification; (lower panel) AXIN2 message was assessed by Q-PCR after 24 hr stimulation with 100 ng/ml recombinant Wnt3a or PBS. (B) mAXIN2 promoter activity in C3H10T1/2^SS18-SSX1-V5 infected with ΔNLEF-1 or control vector. The ratio of firefly to renilla luciferase activity is reported on a logarithmic scale. Results are representative of three independent experiments. Error bars represent the SD of triplicate tests. (C) (upper panels) Fluorescence and bright field microscope images of cells derived from fresh SS samples infected with a lentiviral vector expressing both ΔNLEF-1 and RFP, 200X magnification; (lower panels) AXIN2 message assessed by Q-PCR. (D) Graphical representation and statistical analysis of PLA using anti-β−catenin (mouse) and anti-LEF-1(goat) antibodies in unstimulated or 24 hr recombinant Wnt3a-stimulated C3H10T1/2^pLIVc (pLIVc) and C3H10T1/2^SS18-SSX1-V5 (SS-V5) cells. PLA signal quantification was performed as described in materials and methods. Bars represent the S.E.M. (E) Luciferase activity after transient transfection with TOP- and FOP-Flash plasmids of resting or 24 hr recombinant Wnt3a-stimulated C3H10T1/2^pLIVc (pLIVc) and C3H10T1/2^SS18-SSX1-V5 (SS-V5) cells. Results are reported as the ratio of TOP-Flash:FOP-Flash activity and are representative of three independent experiments. Error bars represent the SD of triplicate tests.

Figure 3. β−catenin participates in SS18-SSX1-induced AXIN2 promoter activity.

C3H10T1/2^pLIVc (pLIVc) and C3H10T1/2^SS18-SSX1-V5 (SS-V5) cells were depleted of β−catenin using a pool of siRNAs or control siRNA (AllStars). Depletion was verified by Western blot and densitometric analysis using imageJ (Fig. 3A) or by qPCR (Fig. 3B). The effect on AXIN2 transcript levels was measured by Q-PCR (Fig. 3B).

Figure 4. Implication of TCF/LEF family members in regulating AXIN2 expression.

(A) Q-PCR analysis of AXIN2 in unstimulated or Wnt3a-stimulated (0.1 μM, 24 hrs) C3H10T1/2^pLIVc (pLIVc) and C3H10T1/2^SS18-SSX1-V5 (SS-V5) cells depleted of or over-expressing individual TCF members as indicated. Depletion was achieved using a pool of specific or control siRNA (AllStars). (B) (first 3 columns): Depletion of each TCF/LEF (as indicated) was verified in each cell population by Q-PCR. B (right columns): Over-expression of TCF4 was measured by Q –PCR. (C) TCF4 depletion was verified by Western blot analysis using an anti-TCF4 antibody and anti-βactin as loading control; (D) LEF1 depletion was verified by immunofluorescence microscopy using an anti-LEF1 antibody. (E) Over-expression of HA-tagged LEF1 was assessed by Western blot analysis using an anti-HA antibody and anti-tubulin antibody as loading control.

Figure 5. Interactions among SS18-SSX, HDAC1, TCFs, TLE and LEF-1.

(A) Graphical representation and statistical analysis of PLAs using antibodies against the indicated proteins in C3H10T1/2^pLIVc (pLIVc) and C3H10T1/2^SS18-SSX1-V5 (SS-V5) cells. PLA signal quantification was performed as described in materials and methods. Corresponding representative fluorescence microscopy images are shown in Fig. S3. (B) Immunoprecipitation (IP) was done on C3H10T1/2^pLIVc (pLIVc) and C3H10T1/2^SS18-SSX1-V5 (SS-V5) cell lysates using anti-V5 monoclonal antibody (a and b), normal goat serum (NGS) or a goat polyclonal α-LEF-1 antibody (c). Immunoprecipitates were subjected to 10% SDS-PAGE together with total lysates and revealed using anti- TCF4 (a), anti-HDAC1 (b) or monoclonal α-V5 antibody (c).TCF4, HDAC1, SS18-SSX-V5, LEF1, immunoglobulins and protein size markers are indicated. (C) Native gel electrophoresis of C3H10T1/2^pLIVc (pLIVc) and C3H10T1/2^SS18-SSX1-V5 (SS-V5) cell lysates. Equal amounts of each sample were loaded onto four wells on the same 5% gel, subjected to electrophoresis and transferred onto nitrocellulose membranes. After blotting the nitrocellulose membrane was cut and incubated with the indicated antibodies. The arrow indicates the co-migration position of HDAC1, TCF4 and SS18-SSX-V5.

Figure 6. DNA dependence of SS18-SSX interactions.

(A) Graphical representation and statistical analysis of PLAs using antibodies against the indicated proteins. C3H10T1/2^SS18-SSX1-V5 cells were pre-stained with 5 μM DAPI for 30 min and subjected to PLA or subjected to PLA and then counterstained with DAPI. PLA signal quantification was performed as described in materials and methods. (B) ChIP using anti-TCF3/4, anti-V5 and anti-HDAC1 antibodies in C3H10T1/2^pLIVc and C3H10T1/2^SS18-SSX1-V5 cells. Results are expressed as fold enrichment of values obtained with rabbit or mouse IgG immunoprecipitates, after normalization for the total amount of input chromatin. Results are representative of three independent experiments. Error bars represent the SD of triplicate PCR tests.

Figure 7

(A) SS18-SSX expression promotes interaction between β−catenin and HDAC1 in nuclei. Graphical representation with statistical analysis of PLAs using antibodies against the indicated proteins in C3H10T1/2^pLIVc (pLIVc) and C3H10T1/2^SS18-SSX1-V5 (SS-V5) cells. Correspondent representative fluorescence microscopy images are reported in Fig. S7. (B) LEF-1 depletion promotes SS18-SSX-HDAC1 association. Graphical representation of PLA as in (A). (C) TSA treatment inhibits SS18-SSX association with HDAC1. Representative fluorescence microscopy images of PLA using anti-V5 and anti-HDAC1 antibodies in C3H10T1/2^SS18-SSX1-V5 cells stimulated with 250 nM TSA or DMSO for 16hrs. Scale bar: 5 μm. (D,E) Effect of TSA treatment on LEF-1 interactions with HDAC (D) and SS18-SSX (E); graphical representation of PLA as in (A).

Figure 8. SS18-SSX promotes Histone H3K9 acetylation and chromatin accessibility at the AXIN2 promoter.

(A) Histone H3K9^Ac ChIP in C3H10T1/2^pLIVc (pLIVc), C3H10T1/2^SS18-SSX1-V5 (SS-V5) and STO cells infected with SS18-SSX-V5 or empty pLIVc. Cross-linked chromatin was sonicated and immunoprecipitated with an anti Histone H3K9^Ac antibody or rabbit IgG. Co-immunoprecipitated DNA was quantified by real-time PCR using primer pairs annealing to the mouse AXIN2 promoter region at the indicated positions. Results are expressed as fold enrichment of values obtained with rabbit IgG precipitates after normalization for the total amount of input chromatin. Results are representative of three independent experiments. Error bars represent the SD of triplicate PCR tests. (B) Chromatin accessibility tests in C3H10T1/2^pLIVc (pLIVc), C3H10T1/2^SS18-SSX1-V5 (SS-V5) and C3H10T1/2^SS18-SS1-HA (SS-HA) cells. Positions of primers used for amplification of MspI CHART products by qRT-PCR are reported with approximate location of ATGs and transcription factor binding sites. Results are representative of three independent experiments. Error bars represent the SD of triplicate PCR tests.