The TAZ-CAMTA1 Fusion Protein Promotes Tumorigenesis via Connective Tissue Growth Factor and Ras-MAPK Signaling in Epithelioid Hemangioendothelioma

Abstract

Purpose: A consistent genetic alteration in vascular cancer epithelioid hemangioendothelioma (EHE) is the t(1;3)(p36;q25) chromosomal translocation, which generates a WWTR1(TAZ)-CAMTA1 (TC) fusion gene. TC is a transcriptional coactivator that drives EHE. Here, we aimed to identify the TC transcriptional targets and signaling mechanisms that underlie EHE tumorigenesis.

Experimental design: We used NIH3T3 cells transformed with TC (NIH3T3/TC) as a model system to uncover TC-dependent oncogenic signaling. These cells proliferated in an anchorage-independent manner in suspension and soft agar. The findings of the cell-based studies were validated in a xenograft model.

Results: We identified connective tissue growth factor (CTGF) as a tumorigenic transcriptional target of TC. We show that CTGF binds to integrin αIIbβ3, which is essential for sustaining the anchorage-independent proliferation of transformed NIH3T3/TC cells. NIH3T3/TC cells also have enhanced Ras and MAPK signaling, and the activity of these pathways is reduced upon CTGF knockdown, suggesting that CTGF signaling occurs via the Ras-MAPK cascade. Further, pharmacologic inhibition of MAPK signaling through PD 0325901 and trametinib abrogated TC-driven anchorage-independent growth. Likewise, for tumor growth in vivo, NIH3T3/TC cells require CTGF and MAPK signaling. NIH3T3/TC xenograft growth was profoundly reduced upon CTGF knockdown and after trametinib treatment.

Conclusions: Collectively, our results demonstrated that CTGF and the Ras-MAPK signaling cascade are essential for TC-mediated tumorigenesis. These studies provided the preclinical rationale for SARC033 (NCI 10015-NCT03148275), a nonrandomized, open-label, phase II study of trametinib in patients with unresectable or metastatic EHE.

©2022 The Authors; Published by the American Association for Cancer Research.

Figures

Figure 1.

CTGF is a key downstream target of the TC fusion protein. A, Immunoblots to estimate the expression levels of CTGF and CYR61 in NIH3T3 cells transfected with either an EV or FLAG-TC fusion gene construct (TC). Cells were grown in suspension on HEMA-treated culture plates. Cell proliferation assay to measure the rate of anchorage-independent cell growth of NIH3T3/TC cells grown in suspension after the knockdown of CTGF (B) or CYR61 (C). D, Soft agar colonies of NIH3T3/TC cells expressing either control (SHC1/2), CTGF, or CYR61 shRNAs; quantification of the colony numbers is also shown. E, Soft agar assay to observe colony growth after CTGF overexpression. Data represent the mean ± SD of independent wells; experiments were performed at least twice.

Figure 2.

Integrin αIIbβ3 is crucial for the TC signaling pathway. A, Lysates from NIH3T3 cells transfected with FLAG-CTGF were immunoprecipitated with an anti-FLAG antibody. Coimmunoprecipitated integrins were detected in an immunoblot using anti-integrin αIIb, anti-integrin β3, or anti-integrin αV antibodies. B, Immunoblots to detect the efficiency of shRNA-mediated knockdown of integrin β3, αV, or αIIb in NIH3T3/TC cells. SHC1 is an EV, SHC2 is a scrambled control shRNA. For integrin knockdowns, two independent shRNA were used. C, Quantification of the soft agar colony counts after knockdown of the indicated integrins in NIH3T3/TC cells. Silencing of integrin αIIb and integrin β3, but not integrin αV, inhibited the ability of NIH3T3/TC cells to form colonies. Data represent mean and SD of individual wells. IP, immunoprecipitated; IB, immunoblot.

Figure 3.

Ras is an important member of the TC signaling pathway. A, Expression levels of active Ras (Ras-GTP) in NIH3T3/TC and NIH3T3/EV cells that were cultured in an anchorage-independent manner. Active Ras was precipitated using the RBD of Raf-1, and the precipitated Ras levels were determined by immunoblot using an anti-Ras antibody. B, Expression levels of active Ras in NIH3T3/TC cells upon CTGF knockdown using two independent shRNA; the cells were cultured in suspension. C, Immunoblots to estimate active Ras levels and D, the activity of MAPK signaling after sorafenib (Raf inhibitor) treatment. E, Soft agar assay in the presence of vehicle (DMSO) or sorafenib. Inhibiting RAS suppressed the ability of NIH3T3/TC cells to form colonies. Comparison is to DMSO vehicle. Error bars represent one SD, and the plot is representative of two independent experiments.

Figure 4.

TC predominantly activates the MAPK signaling pathway. A, NIH3T3/EV and NIH3T3/TC cells were grown attached to dishes (day 0) or in suspension for 1, 3, 7, 9, 11, or 13 days before harvesting for lysate preparation. Levels of pERK, ERK, pAKT, AKT, pS6, and S6 in cell lysates were detected by immunoblots. B, MAPK activity was estimated using a pERK antibody in MS1 cells after exogenous TC expression. C, NIH3T3 cells expressing a double FLAG-CTGF were grown in suspension for 7 days before collection. Cell lysates were analyzed by immunoblot with antibodies to pERK and ERK. Densitometry data for pERK is also shown. Expression of pERK, ERK, pAKT, and AKT in NIH3T3/TC cells after (D) CTGF or (E) CYR61 knockdown. Actin is used as a loading control. F, Immunoblots to estimate the levels of pERK and ERK between normal and matched human EHE tumor tissue samples. N, normal tissue samples; T, tumor tissue samples.

Figure 5.

Effects of pharmacologic inhibitors of MAPK or PI3K pathway on anchorage-independent growth of NIH3T3/TC cells. A, NIH3T3/TC cells were cultured in the presence of either PD 0325901 or GDC 0941 for 2 days after being grown in suspension for 7 days, lysed, and the cell lysates were analyzed by immunoblotting with the indicated antibodies. GDC 0941 is a PI3K inhibitor, and PD 0325901 is a MEK inhibitor. B, The change in cell number over time of NIH3T3/TC cells grown in suspension in the presence of GDC 0941 or PD 0325901. PD 0325901 effectively inhibited the anchorage-independent proliferation of NIH3T3/TC cells (P < 0.05). C, Estimating the cell number of NIH3T3/TC cells grown in suspension. Administration of trametinib significantly decreased cell proliferation in suspension in a dose-dependent manner compared withto DMSO only (P < 0.0001). D, Soft agar colony numbers of NIH3T3/TC cells in the presence of GDC 0941 or PD 0325901 as compared with vehicle DMSO. E, Quantification of soft agar colonies with NIH3T3/TC cells in the presence of vehicle (DMSO) or trametinib. Error bars represent one SD; experiments were performed at least twice.

Figure 6.

CTGF and MAPK signaling sustain tumor growth in vivo. NIH3T3/TC cells were infected with a control GFP shRNA or a shRNA targeting CTGF and assayed for A,Ctgf mRNA expression by qPCR; data are shown as mean ± SD, samples were normalized to control cells lacking any shRNA (dotted line; n = 5 samples run in triplicate from five independent experiments, **P < 0.01). B, Control or CTGF knockdown cells were injected subcutaneously in NSG mice, and tumor volume was measured twice per week for 85 days. C, Tumors from (B) were isolated and weighed; data are shown as mean ± SD (n = 6 shGFP mice and n = 5 shCTGF-1 mice, *P < 0.05 by unpaired t test). D, NIH3T3 cells stably expressing TAZ-CAMTA1 were injected subcutaneously and allowed to form tumors. Once tumors reached a volume of between 150 to 250 mm3, mice received daily oral treatment with either vehicle or 1 mg/kg trametinib for 14 days. Plots show individual mouse tumor volume normalized to tumor volume at the start of treatment (solid Iines with dots). Solid blue lines represent linear regression with 95% CI (dashed blue lines) shown (n = 6 mice per group); the difference in the slopes of the linear regression was tested by ANOVA; **** P < 0.0001.