HMGA1 regulates trabectedin sensitivity in advanced soft-tissue sarcoma (STS): A Spanish Group for Research on Sarcomas (GEIS) study

Abstract

HMGA1 is a structural epigenetic chromatin factor that has been associated with tumor progression and drug resistance. Here, we reported the prognostic/predictive value of HMGA1 for trabectedin in advanced soft-tissue sarcoma (STS) and the effect of inhibiting HMGA1 or the mTOR downstream pathway in trabectedin activity. The prognostic/predictive value of HMGA1 expression was assessed in a cohort of 301 STS patients at mRNA (n = 133) and protein level (n = 272), by HTG EdgeSeq transcriptomics and immunohistochemistry, respectively. The effect of HMGA1 silencing on trabectedin activity and gene expression profiling was measured in leiomyosarcoma cells. The effect of combining mTOR inhibitors with trabectedin was assessed on cell viability in vitro studies, whereas in vivo studies tested the activity of this combination. HMGA1 mRNA and protein expression were significantly associated with worse progression-free survival of trabectedin and worse overall survival in STS. HMGA1 silencing sensitized leiomyosarcoma cells for trabectedin treatment, reducing the spheroid area and increasing cell death. The downregulation of HGMA1 significantly decreased the enrichment of some specific gene sets, including the PI3K/AKT/mTOR pathway. The inhibition of mTOR, sensitized leiomyosarcoma cultures for trabectedin treatment, increasing cell death. In in vivo studies, the combination of rapamycin with trabectedin downregulated HMGA1 expression and stabilized tumor growth of 3-methylcholantrene-induced sarcoma-like models. HMGA1 is an adverse prognostic factor for trabectedin treatment in advanced STS. HMGA1 silencing increases trabectedin efficacy, in part by modulating the mTOR signaling pathway. Trabectedin plus mTOR inhibitors are active in preclinical models of sarcoma, downregulating HMGA1 expression levels and stabilizing tumor growth.

Keywords: HMGA1; Leiomyosarcoma; Soft-tissue sarcoma; Trabectedin; mTOR pathway.

Fig. 1

Univariate progression-free survival (PFS) analysis. A PFS according to HMGA1 gene expression; B PFS according to HMGA2 gene expression; C PFS according to HMGB1 gene expression: D PFS according to HMGB2 gene expression; and E PFS according to HMGB3 gene expression. The groups were defined according to the median of gene expression: above (N = 66) and below (N = 67) the median. Log‐rank test statistical significance was defined at p ≤ 0.05

Fig. 2

Prognostic value of HMGA1 protein levels. A Distribution of HMGA1 and HMGB1 protein expression and intensity; B progression-free survival (PFS) according to HMGA1 protein expression (low expression, N = 224 vs. high expression, N = 48); C PFS according to HMGA1 immunostaining intensity (weak-negative, N = 206 vs. strong, N = 66); D Overall survival (OS) according to HMGA1 protein expression (low expression, N = 224 vs. high expression, N = 48); E OS according to HMGA1 immunostaining intensity (weak-negative, N = 206 vs. strong, N = 66)

Fig. 3

HMGA1 expression levels and effect of HMGA1 downregulation in trabectedin sensitivity in in vitro models of soft-tissue sarcoma (STS). A HMGA1 gene expression levels in STS cell lines; B relative expression of HMGA1 mRNA levels in STS cell lines to an external human RNA pool. C HMGA1 protein expression levels in STS cell lines; D correlation between mRNA and protein expression levels in the panel of STS cell lines used. Pearson’s r-test statistical significance was defined at p ≤ 0.05; E depletion of HMGA1 in CP0024 cells, using two different shRNAs. HMGA1 was knocked down and the level of HMGA1 protein was determined by western blot. F Effect of HMGA1 knockdown on the formation of 3D spheroids of CP0024 cells treated with 10 nM trabectedin for 72 h. G Effect of HMGA1 knockdown in the area of the CP0024 3D spheroid treated with 10 nM trabectedin for 72 h. H Percentage of non-viable cells (apoptotic, early apoptotic, and necrotic cells) measured by flow cytometry after 24 h of treatment with 10 nM trabectedin (n = 3). CP0024 cells were transduced with a shHMGA1 or with a non-targetting shControl, before treatment. Student t-test statistical significance was defined at p ≤ 0.05 (*), p ≤ 0.005 (**) or p ≤ 0.0005 (***). Spheroid areas were determined with ImageJ

Fig. 4

Effect of HGMA1 silencing in gene expression. A Volcano plot with the top significant genes (adjusted p-value < 0.05 and logFC < − 1 or > 1), comparing shHMGA1 vs. shControl; B enrichment analysis in Hallmark gene sets of the effect of HMGA1 knockdown and C effect of HMGA1 silencing in mTOR cell signaling pathway target phospho-S6, by western blot, in CP0024 leiomyosarcoma cells. Protein expression levels were quantified with Image J

Fig. 5

Efficacy of rapamycin plus trabectedin combination in soft-tissue sarcoma preclinical models. A Percentage of proliferating cells in the CP0024 and AA cell lines treated with increased concentrations of trabectedin (10E-11 to 10E-8) for 72 h (red curve) and in CP0024 and AA cells pre-treated with 20 nM rapamycin for 2 h, before trabectedin treatment (10E-11 to 10E-8) for 72 h (black curve). Three biological replicates with three technical replicates were performed. B Protein expression, by western blot, of apoptotic makers (cleaved PARP1 and cleave Caspase 3) after 24 h of treatment with 20 nM rapamycin and 2 nM trabectedin, as single agents or in combination in CP0024 cell line. C Protein expression, by western blot, of apoptotic makers (cleaved PARP1 and cleave Caspase 3) after 24 h of treatment with 20 nM rapamycin and 0.5 nM trabectedin, as single agents or in combination in AA cell line. D Tumor volume (mm³) of mice treated with intraperitoneal rapamycin (0.5 mg/kg) two times a week (q7dx2) and intravenous trabectedin (0.15 mg/kg) once a week (q7dx1), as single agents or in combination. DMSO was used as the control vehicle. E HMGA1 protein expression was analyzed by immunohistochemistry in the tumors collected from mice (200x). F Extension of HMGA1 protein expression levels quantification. Student t-test statistical significance was defined at p ≤ 0.05 (*), p ≤ 0.005 (**) or p ≤ 0.0005 (***)